From Claudius at sailer-online.de Sat Jan 1 17:44:44 2005 From: Claudius at sailer-online.de (Claudius Sailer) Date: Sat Jan 1 10:45:05 2005 Subject: V4RB: Question about SQL Message-ID: <707EFFF0-5C14-11D9-8195-00039365848C@sailer-online.de> Hi, I have 2 Functions which calculates 2 values out of one table. Is it possible to bring this with valentina 2 in one function in one SQL-Statement? The problem is, that the values in every function could by bigger then RB integer but the diffence of both normally never. IMHO I think following would bring solution, when it will work in V4RB 2.X SELECT ( SELECT SUM(Betrag) FROM Buchungen WHERE Ziel_Konto=KontoRecID AND BDatum>'StartDatum' AND BDatum<='Enddatum' ) - ( SELECT SUM(Betrag) FROM Buchungen WHERE Herk_Konto= KontoRecID AND BDatum>'StartDatum' AND BDatum<='Enddatum' ) But with V4RB 1.10 this wouldn't work or is this wrong? Thanks for help Claudius FUNCTION 1 //Hier werden die Einnahmen ermittelt Dim CurBuchungen as VCursor //Einnahmen SQLString="SELECT SUM(Betrag)" SQLString=SQLString+" FROM Buchungen" SQLString=SQLString+ " WHERE Ziel_Konto= "+KontoRecID if NOT CheckLeer(Startdatum) then SQLString=SQLString+ " AND BDatum>'"+StartDatum+"'" end if if NOT checkLeer(Enddatum) then SQLString=SQLString+ " AND BDatum<='"+Enddatum+"'" end if CurBuchungen=meineDatenbank.SQLselect(SQLString, kV_Server, kV_NoLocks, kV_Random) if CurBuchungen<>NIL then CurBuchungen.CurrentPosition=1 return CurBuchungen.LLongField(1).value else return 0 end if FUNCTION 2 //Hier werden die Ausgaben ermittelt Dim CurBuchungen as VCursor //Ausgaben SQLString="SELECT SUM(Betrag)" SQLString=SQLString+" FROM Buchungen" SQLString=SQLString+ " WHERE Herk_Konto= "+KontoRecID if NOT CheckLeer(Startdatum) then SQLString=SQLString+ " AND BDatum>'"+StartDatum+"'" end if if NOT checkLeer(Enddatum) then SQLString=SQLString+ " AND BDatum<='"+Enddatum+"'" end if CurBuchungen=meineDatenbank.SQLselect(SQLString, kV_Server, kV_NoLocks, kV_Random) if CurBuchungen<>NIL then CurBuchungen.CurrentPosition=1 return CurBuchungen.LLongField(1).value else return 0 end if From sunshine at public.kherson.ua Sat Jan 1 19:59:22 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Sat Jan 1 11:59:38 2005 Subject: V4RB: Question about SQL In-Reply-To: <707EFFF0-5C14-11D9-8195-00039365848C@sailer-online.de> Message-ID: On 1/1/05 6:44 PM, "Claudius Sailer" wrote: > Hi, > > I have 2 Functions which calculates 2 values out of one table. Is it > possible to bring this with valentina 2 in one function in one > SQL-Statement? > The problem is, that the values in every function could by bigger then > RB integer but the diffence of both normally never. > > IMHO I think following would bring solution, when it will work in V4RB > 2.X > > SELECT > ( SELECT SUM(Betrag) > FROM Buchungen > WHERE Ziel_Konto=KontoRecID > AND BDatum>'StartDatum' > AND BDatum<='Enddatum' > ) > - > ( SELECT SUM(Betrag) > FROM Buchungen > WHERE Herk_Konto= KontoRecID > AND BDatum>'StartDatum' > AND BDatum<='Enddatum' > ) Yes this should works. Right Ivan ? Claudius, you have any problems with this query ? > But with V4RB 1.10 this wouldn't work or is this wrong? Yes in 1.10 this will not work. -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From IvanSmahin at public.kherson.ua Sat Jan 1 20:15:00 2005 From: IvanSmahin at public.kherson.ua (Ivan Smahin) Date: Sat Jan 1 12:14:56 2005 Subject: V4RB: Question about SQL In-Reply-To: References: Message-ID: <41D6E8A4.3080103@public.kherson.ua> Ruslan Zasukhin ?????: > On 1/1/05 6:44 PM, "Claudius Sailer" wrote: > > >>Hi, >> >>I have 2 Functions which calculates 2 values out of one table. Is it >>possible to bring this with valentina 2 in one function in one >>SQL-Statement? > > >>The problem is, that the values in every function could by bigger then >>RB integer but the diffence of both normally never. >> >>IMHO I think following would bring solution, when it will work in V4RB >>2.X >> >>SELECT >>( SELECT SUM(Betrag) >> FROM Buchungen >> WHERE Ziel_Konto=KontoRecID >> AND BDatum>'StartDatum' >> AND BDatum<='Enddatum' >> ) >>- >>( SELECT SUM(Betrag) >> FROM Buchungen >> WHERE Herk_Konto= KontoRecID >> AND BDatum>'StartDatum' >> AND BDatum<='Enddatum' >> ) > > > Yes this should works. Right Ivan ? > Seems to be correct. > Claudius, you have any problems with this query ? > > > >>But with V4RB 1.10 this wouldn't work or is this wrong? > > > Yes in 1.10 this will not work. > > -- Best regards, Ivan Smahin. From sunshine at public.kherson.ua Sat Jan 1 20:22:33 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Sat Jan 1 12:22:40 2005 Subject: [V4RB 0000242]: Database.Open crashes for my NEW database In-Reply-To: Message-ID: On 12/31/04 11:22 PM, "Ruslan Zasukhin" wrote: > On 12/31/04 10:31 PM, "vbt@paradigma.ukrcom.kherson.ua" > wrote: > >> The following issue has been SUBMITTED. >> ====================================================================== >> >> ====================================================================== >> Reported By: Frank Schima >> Assigned To: >> ====================================================================== >> Project: V4RB >> Issue ID: 242 >> Category: API >> Reproducibility: always >> Severity: crash >> Priority: normal >> Status: new >> ====================================================================== >> Date Submitted: 12-31-2004 22:31 EET >> Last Modified: 12-31-2004 22:31 EET >> ====================================================================== >> Summary: Database.Open crashes for my NEW database >> Description: >> Attempting to open my newly created database with ViSQL 2 causes a crash. >> I'm sending a project that creates the database to Ruslan. Aha, Frank, We have study this problem, and we found that problem is that you use small db.SegmentSize 2 or 4 KB. If set it 8 or 16 or 32 then all works. I will try fix this bug now. -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From Claudius at sailer-online.de Sat Jan 1 19:26:57 2005 From: Claudius at sailer-online.de (Claudius Sailer) Date: Sat Jan 1 12:27:06 2005 Subject: V4RB: Question about SQL In-Reply-To: References: Message-ID: Hi Ruslan, Am 01. Jan 2005 um 18:59 Uhr schrieb Ruslan Zasukhin: > On 1/1/05 6:44 PM, "Claudius Sailer" wrote: > >> Hi, >> >> I have 2 Functions which calculates 2 values out of one table. Is it >> possible to bring this with valentina 2 in one function in one >> SQL-Statement? > >> The problem is, that the values in every function could by bigger then >> RB integer but the diffence of both normally never. >> >> IMHO I think following would bring solution, when it will work in V4RB >> 2.X >> >> SELECT >> ( SELECT SUM(Betrag) >> FROM Buchungen >> WHERE Ziel_Konto=KontoRecID >> AND BDatum>'StartDatum' >> AND BDatum<='Enddatum' >> ) >> - >> ( SELECT SUM(Betrag) >> FROM Buchungen >> WHERE Herk_Konto= KontoRecID >> AND BDatum>'StartDatum' >> AND BDatum<='Enddatum' >> ) > > Claudius, you have any problems with this query ? Not at the moment: problem is, that I want to wait until I tranform my datgabase to new version until I have a good working converting routine. I was only asking whether this will be possible or not. Normally I would expect that RS will implement 64bit numeric values, but I asked for feature request 12 months or more ago and nothing happen, so I have to look further for other solutions :-(( In my point of view it is faster to get good solutions in Valentina then in RB. I am waiting for final 2.X, because this will make much much SQL-Satements possible and will be faster as my work around in RB .... thnaks a lot for your work and a happy new year to all of you bye Claudius -- G4/733 QS / MacOS X 10.3.5de / RB 5.5.3de/ Valentina 1.10.0 Homepage http://www.ClaSai.de Download finale Versionen, Betaversionen auf der Homepage RealBasic ListBoxes: [ I feel the need...the need for speed!!! ] From sunshine at public.kherson.ua Sat Jan 1 20:32:42 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Sat Jan 1 12:32:45 2005 Subject: V4RB: Question about SQL In-Reply-To: Message-ID: On 1/1/05 8:26 PM, "Claudius Sailer" wrote: Hi Claudius, > Not at the moment: problem is, that I want to wait until I tranform my > datgabase to new version until I have a good working converting > routine. You want write some routine? For what ? > I was only asking whether this will be possible or not. yes > Normally I would expect that RS will implement 64bit numeric values, > but I asked for feature request 12 months or more ago and nothing > happen, so I have to look further for other solutions :-(( > In my point of view it is faster to get good solutions in Valentina > then in RB. I am waiting for final 2.X, because this will make much > much SQL-Satements possible and will be faster as my work around in RB SUM() will return us column of ULLONG or LLONG type I think. You will be able obtain result using s = cursor.field(1).GetString() longV = val( s ) > .... > > thnaks a lot for your work and a happy new year to all of you Thank you Claudius, Happy New Year to everybody! -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From sunshine at public.kherson.ua Sat Jan 1 20:33:24 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Sat Jan 1 12:33:27 2005 Subject: SQL LIKE In-Reply-To: Message-ID: On 12/31/04 8:36 PM, "jda" wrote: > Sorry if this is received twice -- it didn't show up for me the first > time I sent it: > > Looking at the latest SQL docs for V2 (p. 24), it seems that LIKE > should be acceptable for searching (equal to REGEX). But it doesn't > work for me. Should it? Or am I misreading the documentation? Jon, Please note that LIKE description is in the ValentinaSQL.pdf -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From Claudius at sailer-online.de Sat Jan 1 20:29:30 2005 From: Claudius at sailer-online.de (Claudius Sailer) Date: Sat Jan 1 13:29:41 2005 Subject: V4RB: Question about SQL In-Reply-To: References: Message-ID: <7561B9A5-5C2B-11D9-8195-00039365848C@sailer-online.de> Am 01. Jan 2005 um 19:32 Uhr schrieb Ruslan Zasukhin: > Hi Claudius, > >> Not at the moment: problem is, that I want to wait until I tranform my >> datgabase to new version until I have a good working converting >> routine. > > You want write some routine? No, this is a missunderstanding. I have actually less time to test everything, so I have to wait for final version. But I made some easy tests. At the moment my application export database in textfile and is able to import it again. I will use same functionality for moving to 2.X. Export old database with old Application Version Run new Application Generate new database Import exported datas. > For what ? Do you have a converting routine? If yes then I missed this information by the mass of eMails >> Normally I would expect that RS will implement 64bit numeric values, >> but I asked for feature request 12 months or more ago and nothing >> happen, so I have to look further for other solutions :-(( > >> In my point of view it is faster to get good solutions in Valentina >> then in RB. I am waiting for final 2.X, because this will make much >> much SQL-Satements possible and will be faster as my work around in RB > > SUM() will return us column of ULLONG or LLONG type I think. > > You will be able obtain result using > > s = cursor.field(1).GetString() > > longV = val( s ) Thats correct. I am using CurBuchungen=meineDatenbank.SQLselect(SQLString, kV_Server, kV_NoLocks, kV_Random) if CurBuchungen<>NIL then CurBuchungen.CurrentPosition=1 return CurBuchungen.LLongField(1).value else return 0 end if but the returned value could be bigger as it is allowed as integer and I want to return integer because of the more accurate value. Float is not so correct as integer und numeric is not supported by RB.... Thats the reason why I am asking for 64bit integer in RB. bye Claudius From sunshine at public.kherson.ua Sat Jan 1 21:49:47 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Sat Jan 1 13:49:53 2005 Subject: V4RB: Question about SQL In-Reply-To: <7561B9A5-5C2B-11D9-8195-00039365848C@sailer-online.de> Message-ID: On 1/1/05 9:29 PM, "Claudius Sailer" wrote: Hi Claudius, >>> Not at the moment: problem is, that I want to wait until I tranform my >>> datgabase to new version until I have a good working converting >>> routine. >> >> You want write some routine? > > No, this is a missunderstanding. I have actually less time to test > everything, so I have to wait for final version. But I made some easy > tests. At the moment my application export database in textfile and is > able to import it again. I will use same functionality for moving to > 2.X. > Export old database with old Application Version > Run new Application > Generate new database > Import exported datas. > >> For what ? > > Do you have a converting routine? > If yes then I missed this information by the mass of eMails We going to have * kernel 1.x DLL to be in the Vcomponents folder * special function to convert 1 to 2 convertdb_1_2( oldPath, newPath ) -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From jda at his.com Sat Jan 1 15:21:13 2005 From: jda at his.com (jda) Date: Sat Jan 1 14:21:25 2005 Subject: Bug #219 Message-ID: Any chance that this bug will be addressed in the next iteration? It is giving my testers a lot of trouble... Jon From sunshine at public.kherson.ua Sat Jan 1 22:27:04 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Sat Jan 1 14:27:13 2005 Subject: Bug #219 In-Reply-To: Message-ID: On 1/1/05 10:21 PM, "jda" wrote: > Any chance that this bug will be addressed in the next iteration? It > is giving my testers a lot of trouble... Yes Jon, I am going fix * frank bug with Grow FAT * your 2 index reports. -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From jda at his.com Sat Jan 1 15:32:16 2005 From: jda at his.com (jda) Date: Sat Jan 1 14:32:21 2005 Subject: Bug #219 In-Reply-To: References: Message-ID: >On 1/1/05 10:21 PM, "jda" wrote: > >> Any chance that this bug will be addressed in the next iteration? It >> is giving my testers a lot of trouble... > >Yes Jon, > >I am going fix > >* frank bug with Grow FAT >* your 2 index reports. > > :-). Happy New Year (regardless). Jon From frank-list3 at mindstarprods.com Sat Jan 1 15:55:12 2005 From: frank-list3 at mindstarprods.com (Frank Schima) Date: Sat Jan 1 16:55:16 2005 Subject: [V4RB 0000242]: Database.Open crashes for my NEW database In-Reply-To: References: Message-ID: <318CA43D-5C48-11D9-A679-0003939246BC@mindstarprods.com> Hi Ruslan, On Jan 1, 2005, at 11:22 AM, Ruslan Zasukhin wrote: > On 12/31/04 11:22 PM, "Ruslan Zasukhin" > wrote: > >> On 12/31/04 10:31 PM, "vbt@paradigma.ukrcom.kherson.ua" >> wrote: >> >>> The following issue has been SUBMITTED. >>> ===================================================================== >>> = >>> >>> ===================================================================== >>> = >>> Reported By: Frank Schima >>> Assigned To: >>> ===================================================================== >>> = >>> Project: V4RB >>> Issue ID: 242 >>> Category: API >>> Reproducibility: always >>> Severity: crash >>> Priority: normal >>> Status: new >>> ===================================================================== >>> = >>> Date Submitted: 12-31-2004 22:31 EET >>> Last Modified: 12-31-2004 22:31 EET >>> ===================================================================== >>> = >>> Summary: Database.Open crashes for my NEW database >>> Description: >>> Attempting to open my newly created database with ViSQL 2 causes a >>> crash. >>> I'm sending a project that creates the database to Ruslan. > > Aha, Frank, > > We have study this problem, and we found that problem is that you use > small > db.SegmentSize 2 or 4 KB. > > If set it 8 or 16 or 32 then all works. > > I will try fix this bug now. OK, that gives me an easy workaround. Thanks! The real problem I'm trying to track down is that an UPDATE ... WHERE ... SQL is causing a crash in that database and I cannot debug the problem with ViSQL. I'm having trouble trying to replicate the problem though, because it works fine on an empty database (of course nothing is changed in that case, but it does not crash!) and it also works when I add a few hundred dummy records. Maybe a larger segment size will solve this problem too. I will keep studying this and let you know if I can isolate anything. Best regards, Frank Schima From sunshine at public.kherson.ua Sun Jan 2 01:09:24 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Sat Jan 1 17:09:30 2005 Subject: [V4RB 0000242]: Database.Open crashes for my NEW database In-Reply-To: <318CA43D-5C48-11D9-A679-0003939246BC@mindstarprods.com> Message-ID: On 1/2/05 12:55 AM, "Frank Schima" wrote: >> Aha, Frank, >> >> We have study this problem, and we found that problem is that you use >> small >> db.SegmentSize 2 or 4 KB. >> >> If set it 8 or 16 or 32 then all works. >> >> I will try fix this bug now. > > OK, that gives me an easy workaround. Thanks! > > The real problem I'm trying to track down is that an UPDATE ... WHERE > ... SQL is causing a crash in that database and I cannot debug the > problem with ViSQL. I'm having trouble trying to replicate the problem > though, because it works fine on an empty database (of course nothing > is changed in that case, but it does not crash!) and it also works when > I add a few hundred dummy records. Maybe a larger segment size will > solve this problem too. > > I will keep studying this and let you know if I can isolate anything. Frank, When you develop and debug I recommend to all set Valentina.DebugLevel = kLogParams And if you have crash you need set Valentina.FlushEachLog = true So you not loose log of last commands in the FILE BUFFER. Also you may find useful look into files ASSERTS -- this file MUST BE empty if no any bugs SqlTree.txt -- here you can see how SQL query is parsed. small help, but yet... -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From frank-list3 at mindstarprods.com Sat Jan 1 16:16:25 2005 From: frank-list3 at mindstarprods.com (Frank Schima) Date: Sat Jan 1 17:16:16 2005 Subject: V4RB: Question about SQL In-Reply-To: <7561B9A5-5C2B-11D9-8195-00039365848C@sailer-online.de> References: <7561B9A5-5C2B-11D9-8195-00039365848C@sailer-online.de> Message-ID: <288F5A0E-5C4B-11D9-A679-0003939246BC@mindstarprods.com> Hi guys, On Jan 1, 2005, at 12:29 PM, Claudius Sailer wrote: > > Am 01. Jan 2005 um 19:32 Uhr schrieb Ruslan Zasukhin: > >>> Normally I would expect that RS will implement 64bit numeric values, >>> but I asked for feature request 12 months or more ago and nothing >>> happen, so I have to look further for other solutions :-(( >> >>> In my point of view it is faster to get good solutions in Valentina >>> then in RB. I am waiting for final 2.X, because this will make much >>> much SQL-Satements possible and will be faster as my work around in >>> RB >> >> SUM() will return us column of ULLONG or LLONG type I think. >> >> You will be able obtain result using >> >> s = cursor.field(1).GetString() >> >> longV = val( s ) > > Thats correct. > > I am using > > CurBuchungen=meineDatenbank.SQLselect(SQLString, kV_Server, > kV_NoLocks, kV_Random) > if CurBuchungen<>NIL then > CurBuchungen.CurrentPosition=1 > return CurBuchungen.LLongField(1).value > else > return 0 > end if > > but the returned value could be bigger as it is allowed as integer and > I want to return integer because of the more accurate value. Float is > not so correct as integer und numeric is not supported by RB.... > Thats the reason why I am asking for 64bit integer in RB. Another workaround is to use this free plug-in which handles huge ( > 64 bit ) integers: It works great for me, the only "problem" is that it creates a new class, not an intrinsic datatype, and hence you have to use "New" for each variable of this type as well as class methods to add, subtract, etc. I have voted for signed/unsigned 64 bit integers in RB too of course. Everyone please vote for this if you haven't yet! Best regards, Frank Schima From Claudius at sailer-online.de Sun Jan 2 08:40:37 2005 From: Claudius at sailer-online.de (Claudius Sailer) Date: Sun Jan 2 01:40:49 2005 Subject: V4RB: Question about SQL In-Reply-To: <288F5A0E-5C4B-11D9-A679-0003939246BC@mindstarprods.com> References: <7561B9A5-5C2B-11D9-8195-00039365848C@sailer-online.de> <288F5A0E-5C4B-11D9-A679-0003939246BC@mindstarprods.com> Message-ID: <9828F468-5C91-11D9-8195-00039365848C@sailer-online.de> Hi, Am 02. Jan 2005 um 00:16 Uhr schrieb Frank Schima: > Another workaround is to use this free plug-in which handles huge ( > > 64 bit ) integers: > > > It works great for me, the only "problem" is that it creates a new > class, not an intrinsic datatype, and hence you have to use "New" for > each variable of this type as well as class methods to add, subtract, > etc. Can you say something about speed? Is it correct that you can't use it as return value in functions? Or is this a missunderstanding from me? Thanks for the information about that PlugIn!! bye Claudius From frank-list3 at mindstarprods.com Sun Jan 2 08:28:37 2005 From: frank-list3 at mindstarprods.com (Frank Schima) Date: Sun Jan 2 09:28:26 2005 Subject: V4RB: Question about SQL In-Reply-To: <9828F468-5C91-11D9-8195-00039365848C@sailer-online.de> References: <7561B9A5-5C2B-11D9-8195-00039365848C@sailer-online.de> <288F5A0E-5C4B-11D9-A679-0003939246BC@mindstarprods.com> <9828F468-5C91-11D9-8195-00039365848C@sailer-online.de> Message-ID: Hi Claudius, On Jan 2, 2005, at 12:40 AM, Claudius Sailer wrote: > Am 02. Jan 2005 um 00:16 Uhr schrieb Frank Schima: > >> Another workaround is to use this free plug-in which handles huge ( > >> 64 bit ) integers: >> >> >> It works great for me, the only "problem" is that it creates a new >> class, not an intrinsic datatype, and hence you have to use "New" for >> each variable of this type as well as class methods to add, subtract, >> etc. > > Can you say something about speed? I have not run any benchmarks, or used the new classes in any processor intensive situations so I cannot really address this. They work fast enough where I have used them though. > Is it correct that you can't use it as return value in functions? Or > is this a missunderstanding from me? It should work fine for this, although I have not actually done so. Any object type can be a return variable type for functions/methods AFAIK. > Thanks for the information about that PlugIn!! Happy coding! Best regards, Frank Schima From frank-list3 at mindstarprods.com Sun Jan 2 09:53:23 2005 From: frank-list3 at mindstarprods.com (Frank Schima) Date: Sun Jan 2 10:53:11 2005 Subject: [V4RB 0000242]: Database.Open crashes for my NEW database In-Reply-To: References: Message-ID: On Jan 1, 2005, at 4:09 PM, Ruslan Zasukhin wrote: > On 1/2/05 12:55 AM, "Frank Schima" > wrote: > >> The real problem I'm trying to track down is that an UPDATE ... WHERE >> ... SQL is causing a crash in that database and I cannot debug the >> problem with ViSQL. I'm having trouble trying to replicate the problem >> though, because it works fine on an empty database (of course nothing >> is changed in that case, but it does not crash!) and it also works >> when >> I add a few hundred dummy records. Maybe a larger segment size will >> solve this problem too. >> >> I will keep studying this and let you know if I can isolate anything. > > When you develop and debug I recommend to all set > > Valentina.DebugLevel = kLogParams > > And if you have crash you need set > > Valentina.FlushEachLog = true > > So you not loose log of last commands in the FILE BUFFER. > > > Also you may find useful look into files > > ASSERTS -- this file MUST BE empty if no any bugs > > SqlTree.txt -- here you can see how SQL query is parsed. > small help, but yet... I am still having a crash with my UPDATE SQL statement even with the larger segment size. So I set the debug level and FlushEachLog = True as you suggested after the Valentina.Init statement but I cannot find either of those files anywhere on my computer after doing a search. There was nothing special in the RB project folder either. Where should those files appear? I guess I don't know where any of the debugging information is supposed to appear. Where should I see the debug log? There is no application crash log file in this case. The problematic SQL statement is pretty simple: UPDATE MyTable SET MyVarCharField = 'Some Text' WHERE MyVarCharField is Null Coincidently, this is the same SQL statement that V4RB 1.11b (the latest build of 1.11) fixed - it was broken in the initial build of 1.11. But it simply failed to update without an error in that case. It is causing a crash for me in V4RB 2b33. Best regards, Frank From sunshine at public.kherson.ua Sun Jan 2 19:57:42 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Sun Jan 2 11:57:49 2005 Subject: [V4RB 0000242]: Database.Open crashes for my NEW database In-Reply-To: Message-ID: On 1/2/05 6:53 PM, "Frank Schima" wrote: > I am still having a crash with my UPDATE SQL statement even with the > larger segment size. So I set the debug level and FlushEachLog = True > as you suggested after the Valentina.Init statement but I cannot find > either of those files anywhere on my computer after doing a search. > There was nothing special in the RB project folder either. > > Where should those files appear? In the same folder. Hmm, Aha, we upload RELEASE version now, So you do not see them. -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From sunshine at public.kherson.ua Sun Jan 2 20:05:23 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Sun Jan 2 12:05:37 2005 Subject: [V4RB 0000242]: Database.Open crashes for my NEW database In-Reply-To: Message-ID: On 1/2/05 6:53 PM, "Frank Schima" wrote: > I am still having a crash with my UPDATE SQL statement even with the > larger segment size. Segment size must resolve problem of open in the viSQL > So I set the debug level and FlushEachLog = True > as you suggested after the Valentina.Init statement but I cannot find > either of those files anywhere on my computer after doing a search. > There was nothing special in the RB project folder either. > > Where should those files appear? > > I guess I don't know where any of the debugging information is supposed > to appear. Where should I see the debug log? There is no application > crash log file in this case. Frank, But you should see in the same folder where is app, the file V4RB_Log.txt Do you see it ? > The problematic SQL statement is pretty simple: > UPDATE MyTable SET MyVarCharField = 'Some Text' WHERE MyVarCharField > is Null > > Coincidently, this is the same SQL statement that V4RB 1.11b (the > latest build of 1.11) fixed - it was broken in the initial build of > 1.11. But it simply failed to update without an error in that case. It > is causing a crash for me in V4RB 2b33. Can you send us db ? We will make query in viSQL. -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From frank-list3 at mindstarprods.com Sun Jan 2 13:47:53 2005 From: frank-list3 at mindstarprods.com (Frank Schima) Date: Sun Jan 2 14:47:42 2005 Subject: [V4RB 0000242]: Database.Open crashes for my NEW database In-Reply-To: References: Message-ID: <92F18487-5CFF-11D9-AFC3-0003939246BC@mindstarprods.com> On Jan 2, 2005, at 11:05 AM, Ruslan Zasukhin wrote: > On 1/2/05 6:53 PM, "Frank Schima" > wrote: > >> I am still having a crash with my UPDATE SQL statement even with the >> larger segment size. > > Segment size must resolve problem of open in the viSQL > >> So I set the debug level and FlushEachLog = True >> as you suggested after the Valentina.Init statement but I cannot find >> either of those files anywhere on my computer after doing a search. >> There was nothing special in the RB project folder either. >> >> Where should those files appear? >> >> I guess I don't know where any of the debugging information is >> supposed >> to appear. Where should I see the debug log? There is no application >> crash log file in this case. > > Frank, > > But you should see in the same folder where is app, the file > > V4RB_Log.txt > > Do you see it ? No, I do not see it. Also, I have another problem. I cannot open the empty database with ViSQL even with a segment size of 32 * 1024 (32 kB). I'm not sure what you are doing differently. >> The problematic SQL statement is pretty simple: >> UPDATE MyTable SET MyVarCharField = 'Some Text' WHERE >> MyVarCharField >> is Null >> >> Coincidently, this is the same SQL statement that V4RB 1.11b (the >> latest build of 1.11) fixed - it was broken in the initial build of >> 1.11. But it simply failed to update without an error in that case. It >> is causing a crash for me in V4RB 2b33. > > Can you send us db ? > We will make query in viSQL. For various reasons, I do not want to send it if possible. For one thing is over 20 MB (surely less when compressed though) and I am on modem these days. But also because it contains sensitive data. I will keep trying to replicate the problem with dummy data. Best regards, Frank Schima From sunshine at public.kherson.ua Sun Jan 2 23:02:29 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Sun Jan 2 15:02:35 2005 Subject: [V4RB 0000242]: Database.Open crashes for my NEW database In-Reply-To: <92F18487-5CFF-11D9-AFC3-0003939246BC@mindstarprods.com> Message-ID: On 1/2/05 10:47 PM, "Frank Schima" wrote: >> But you should see in the same folder where is app, the file >> >> V4RB_Log.txt >> >> Do you see it ? > > No, I do not see it. > > Also, I have another problem. I cannot open the empty database with > ViSQL even with a segment size of 32 * 1024 (32 kB). I'm not sure what > you are doing differently. Strange, but we work on this bug right now. Let's wait for its fix. -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From sunshine at public.kherson.ua Sun Jan 2 23:03:17 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Sun Jan 2 15:03:21 2005 Subject: [V4RB 0000242]: Database.Open crashes for my NEW database In-Reply-To: <92F18487-5CFF-11D9-AFC3-0003939246BC@mindstarprods.com> Message-ID: On 1/2/05 10:47 PM, "Frank Schima" wrote: >> >> Can you send us db ? >> We will make query in viSQL. > > For various reasons, I do not want to send it if possible. For one > thing is over 20 MB (surely less when compressed though) and I am on > modem these days. But also because it contains sensitive data. I will > keep trying to replicate the problem with dummy data. In fact I think it is NEVER mind how many records you have! I think we can get this crash even on empty db. If this is parser fault. You see? -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From sunshine at public.kherson.ua Sun Jan 2 23:03:51 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Sun Jan 2 15:03:55 2005 Subject: [V4RB 0000242]: Database.Open crashes for my NEW database In-Reply-To: <92F18487-5CFF-11D9-AFC3-0003939246BC@mindstarprods.com> Message-ID: On 1/2/05 10:47 PM, "Frank Schima" wrote: >> Frank, >> >> But you should see in the same folder where is app, the file >> >> V4RB_Log.txt >> >> Do you see it ? > > No, I do not see it. Hmm, then I doubt you have set ON debug Level. Please copy your code in letter -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From frank-list3 at mindstarprods.com Sun Jan 2 19:28:47 2005 From: frank-list3 at mindstarprods.com (Frank Schima) Date: Sun Jan 2 20:28:40 2005 Subject: [V4RB 0000242]: Database.Open crashes for my NEW database In-Reply-To: References: Message-ID: <32A92FEC-5D2F-11D9-AFC3-0003939246BC@mindstarprods.com> On Jan 2, 2005, at 2:03 PM, Ruslan Zasukhin wrote: > On 1/2/05 10:47 PM, "Frank Schima" > wrote: > >>> Frank, >>> >>> But you should see in the same folder where is app, the file >>> >>> V4RB_Log.txt >>> >>> Do you see it ? >> >> No, I do not see it. > > Hmm, then I doubt you have set ON debug Level. > Please copy your code in letter Here is the code including the Init: Valentina.Init(50 * 1024 *1024, "[serial number deleted]", "" ) Valentina.DebugLevel = EVDebugLevel.kLogParams Valentina.FlushEachLog = True Best regards, Frank Schima From frank-list3 at mindstarprods.com Sun Jan 2 22:43:38 2005 From: frank-list3 at mindstarprods.com (Frank Schima) Date: Sun Jan 2 23:43:37 2005 Subject: [V4RB 0000242]: Database.Open crashes for my NEW database In-Reply-To: References: Message-ID: <6AB04F3D-5D4A-11D9-A992-0003939246BC@mindstarprods.com> On Jan 2, 2005, at 2:03 PM, Ruslan Zasukhin wrote: > On 1/2/05 10:47 PM, "Frank Schima" > wrote: > >>> >>> Can you send us db ? >>> We will make query in viSQL. >> >> For various reasons, I do not want to send it if possible. For one >> thing is over 20 MB (surely less when compressed though) and I am on >> modem these days. But also because it contains sensitive data. I will >> keep trying to replicate the problem with dummy data. > > In fact I think it is NEVER mind how many records you have! > > I think we can get this crash even on empty db. > If this is parser fault. > > You see? Yes. But it does not crash with an empty database. And it also does not crash with some dummy records I added. So far, it only happens with my real data. Let's see what happens with the next beta version. Best regards, Frank Schima From sunshine at public.kherson.ua Mon Jan 3 08:58:24 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Mon Jan 3 00:58:34 2005 Subject: [V4RB 0000242]: V4RB_Log.txt, Jon? In-Reply-To: <32A92FEC-5D2F-11D9-AFC3-0003939246BC@mindstarprods.com> Message-ID: On 1/3/05 4:28 AM, "Frank Schima" wrote: >>>> But you should see in the same folder where is app, the file >>>> >>>> V4RB_Log.txt >>>> >>>> Do you see it ? >>> >>> No, I do not see it. >> >> Hmm, then I doubt you have set ON debug Level. >> Please copy your code in letter > > Here is the code including the Init: > > Valentina.Init(50 * 1024 *1024, "[serial number deleted]", "" ) > Valentina.DebugLevel = EVDebugLevel.kLogParams > Valentina.FlushEachLog = True Very strange. Please try to run any example from V4RB. They all have debug level ON. And I always see this file. Jon do you see this file on your side ? -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From jda at his.com Mon Jan 3 08:44:56 2005 From: jda at his.com (jda) Date: Mon Jan 3 07:45:17 2005 Subject: [V4RB 0000242]: V4RB_Log.txt, Jon? In-Reply-To: References: Message-ID: > > > Valentina.Init(50 * 1024 *1024, "[serial number deleted]", "" ) > > Valentina.DebugLevel = EVDebugLevel.kLogParams > > Valentina.FlushEachLog = True > >Very strange. > >Please try to run any example from V4RB. >They all have debug level ON. > >And I always see this file. >Jon do you see this file on your side ? > Yes, I see a file V4RB_Log.txt in the same folder as RB and my project. Jon From frank-list3 at mindstarprods.com Mon Jan 3 08:53:59 2005 From: frank-list3 at mindstarprods.com (Frank Schima) Date: Mon Jan 3 09:53:48 2005 Subject: [V4RB 0000242]: V4RB_Log.txt, Jon? In-Reply-To: References: Message-ID: Hi guys, I think I found the problem. I have been compiling for Mach-O, but when I changed to PEF, then the file appeared! Ruslan, shall I log this as a bug? Best regards, Frank On Jan 3, 2005, at 6:44 AM, jda wrote: >> >> > Valentina.Init(50 * 1024 *1024, "[serial number deleted]", "" ) >> > Valentina.DebugLevel = EVDebugLevel.kLogParams >> > Valentina.FlushEachLog = True >> >> Very strange. >> >> Please try to run any example from V4RB. >> They all have debug level ON. >> >> And I always see this file. >> Jon do you see this file on your side ? >> > > Yes, I see a file V4RB_Log.txt in the same folder as RB and my project. > > Jon > _______________________________________________ > Valentina-beta mailing list > Valentina-beta@lists.macserve.net > http://lists.macserve.net/mailman/listinfo/valentina-beta > From vidal_olivier at yahoo.fr Mon Jan 3 16:55:44 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Mon Jan 3 09:56:45 2005 Subject: valentina 2 Message-ID: Hello and happy New Year to all, some small questions about valentina 2. Valentina 2 is it much faster than valentina 1 (for a local application, always by using SQL) ? Or it is the about similar but offering more features ? Is the cache memory of valentina now optimized beyond 20 Mb (50, 100, 200 Mb...) ? By using API of valentina (binary link..) rather than SQL, will the application be much faster (for a local application) ? Significantly ? Because in a previous e-mail Ruslan had said "SQL is slower because you need spend time to build strings" but it is not only to gain (with API) some milliseconds ? thank you olivier From sunshine at public.kherson.ua Mon Jan 3 17:57:03 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Mon Jan 3 09:57:10 2005 Subject: [V4RB 0000242]: V4RB_Log.txt, Jon? In-Reply-To: Message-ID: On 1/3/05 5:53 PM, "Frank Schima" wrote: > Hi guys, > > > I think I found the problem. I have been compiling for Mach-O, but when > I changed to PEF, then the file appeared! > > Ruslan, shall I log this as a bug? Aha, no this is not a bug, glitch :-) In your case you also have this file, just you need open the package of application and go to folder MacOS. May be it is possible in plugin define that it work in the MACHO env And put file 3 levels UP. -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From jda at his.com Mon Jan 3 11:04:06 2005 From: jda at his.com (jda) Date: Mon Jan 3 10:04:16 2005 Subject: valentina 2 In-Reply-To: References: Message-ID: >some small questions about valentina 2. > >Valentina 2 is it much faster than valentina 1 (for a local >application, always by using SQL) ? >Or it is the about similar but offering more features ? I don't know how search speed compares (haven't spec'ed that), but it is *very* fast. Some things are much slower, though (in my case, seems to be related to creating/filling records and, perhaps, indexing). I think there is a substantial amount of fine-tuning that remains to be done. Jon From sunshine at public.kherson.ua Mon Jan 3 18:06:35 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Mon Jan 3 10:06:39 2005 Subject: valentina 2 In-Reply-To: Message-ID: On 1/3/05 5:55 PM, "Olivier" wrote: Hi Oliver, > Hello and happy New Year to all, > > some small questions about valentina 2. > > Valentina 2 is it much faster than valentina 1 (for a local > application, always by using SQL) ? Right now no. I see that 2.0 now is 2 times slower on addRecord. Don't know about rest operations. But we yet in debug mode for indexes for example. > Or it is the about similar but offering more features ? I think that AFTER we spend a month or so on optimzation and tunning, We will make 2.0 faster of 1.x. In the same time note, that if you will use UTF16, this make strings 2 times bigger, so strings itself may work 2 times slower. On the other hand for search and sort that use binary search and qsort slow down will be NOT significant I think > Is the cache memory of valentina now optimized beyond 20 Mb (50, 100, > 200 Mb...) ? Yes. Cache in 2.0 is 300-500 times faster of 1.x. The bigger cache the bigger difference to 1.x > By using API of valentina (binary link..) rather than SQL, will the > application be much faster (for a local application) ? A LOTS!!! More correct to say 5-10 times easy. SQL adds significant overhead. > Significantly ? Because in a previous e-mail Ruslan had said "SQL is > slower because you need spend time to build strings" but it is not only > to gain (with API) some milliseconds ? But if you add in loop 100,000 records, You need make 100,000 INSERT strings. Valentina 2.0 offer here for SQL lovers -- binding, This feature can speed up a lots such things. -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From vidal_olivier at yahoo.fr Mon Jan 3 18:01:43 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Mon Jan 3 11:02:39 2005 Subject: valentina 2 In-Reply-To: References: Message-ID: <24CCD684-5DA9-11D9-A02F-000D93C35C66@yahoo.fr> hmmm if I indeed understands, valentina 2 is much slower for the moment than valentina 1 by using SQL but about so fast by using API ? As a lot of people I think, I hesitate to use Valentina 2 from now on. I began a project with Valentina 1. This project must be absolutely ended by March 20th. I am very enthusiastic to use valentina 2 (by API) for this project now because if the version 1.0 of my software will be already based on valentina 2, I can probably make much more quickly updates for the next versions. My software will be built well from the beginning. On the other hand, if my software 1.0 is not stable, finally not much more than with valentina 1 and if I lose a lot of time in the corrections of (valentina 2) bugs, I hesitate and I wonder if I am not going to wait for my next project to use valentina 2. As customer and user of valentina, of whom do you think of it ? PS : it is damage that there not in the documentation some pages to explain "philosophy" and paradigmes of valentina 2. olivier Le 3 janv. 05, ? 17:06, Ruslan Zasukhin a ?crit : > On 1/3/05 5:55 PM, "Olivier" wrote: > > Hi Oliver, > >> Hello and happy New Year to all, >> >> some small questions about valentina 2. >> >> Valentina 2 is it much faster than valentina 1 (for a local >> application, always by using SQL) ? > > Right now no. I see that 2.0 now is 2 times slower on addRecord. > Don't know about rest operations. > But we yet in debug mode for indexes for example. > >> Or it is the about similar but offering more features ? > > I think that AFTER we spend a month or so on optimzation and tunning, > We will make 2.0 faster of 1.x. > > In the same time note, that if you will use UTF16, this make strings 2 > times > bigger, so strings itself may work 2 times slower. On the other hand > for > search and sort that use binary search and qsort slow down will be NOT > significant I think > >> Is the cache memory of valentina now optimized beyond 20 Mb (50, 100, >> 200 Mb...) ? > > Yes. > > Cache in 2.0 is 300-500 times faster of 1.x. > The bigger cache the bigger difference to 1.x > >> By using API of valentina (binary link..) rather than SQL, will the >> application be much faster (for a local application) ? > > A LOTS!!! > > More correct to say 5-10 times easy. > > SQL adds significant overhead. > >> Significantly ? Because in a previous e-mail Ruslan had said "SQL is >> slower because you need spend time to build strings" but it is not >> only >> to gain (with API) some milliseconds ? > > But if you add in loop 100,000 records, > You need make 100,000 INSERT strings. > > Valentina 2.0 offer here for SQL lovers -- binding, > This feature can speed up a lots such things. > > > -- > Best regards, > Ruslan Zasukhin [ I feel the need...the need for speed ] > ------------------------------------------------------------- > e-mail: ruslan@paradigmasoft.com > web: http://www.paradigmasoft.com > > To subscribe to the Valentina mail list go to: > http://lists.macserve.net/mailman/listinfo/valentina > ------------------------------------------------------------- > > > _______________________________________________ > Valentina-beta mailing list > Valentina-beta@lists.macserve.net > http://lists.macserve.net/mailman/listinfo/valentina-beta > From jda at his.com Mon Jan 3 12:21:43 2005 From: jda at his.com (jda) Date: Mon Jan 3 11:21:52 2005 Subject: valentina 2 In-Reply-To: <24CCD684-5DA9-11D9-A02F-000D93C35C66@yahoo.fr> References: <24CCD684-5DA9-11D9-A02F-000D93C35C66@yahoo.fr> Message-ID: >hmmm if I indeed understands, valentina 2 is much slower for the >moment than valentina 1 by using SQL but about so fast by using API ? > >As a lot of people I think, I hesitate to use Valentina 2 from now >on. I began a project with Valentina 1. This project must be >absolutely ended by March 20th. I am very enthusiastic to use >valentina 2 (by API) for this project now because if the version 1.0 >of my software will be already based on valentina 2, I can probably >make much more quickly updates for the next versions. My software >will be built well from the beginning. > >On the other hand, if my software 1.0 is not stable, finally not >much more than with valentina 1 and if I lose a lot of time in the >corrections of (valentina 2) bugs, I hesitate and I wonder if I am >not going to wait for my next project to use valentina 2. > >As customer and user of valentina, of whom do you think of it ? > I wouldn't say that V2 is *much* slower than V1 right now, at least for me...the decrease in speed of addrecord and indexing is bothersome, but Ruslan has said it will be improved, and I believe him. Your more fundamental question is whether a stable V2 is on a schedule that fits your (and my) needs. This has troubled me quite a lot (and Ruslan is tired of hearing from me, I know). Initially Unicode was going to be added to V1, and that is all I really need. But that intention was changed for financial reasons, and it was decided that Unicode was only going to be included in V2. V2 has been delayed, delayed, delayed, which has caused me a lot of grief. I remember on the Valentina list in 2003 (!) when Lynn assured me that V2 would be out *before* RB 5.5. Gosh, was that wrong. When the V2 beta was first posted (May 11, 2004 I believe) we were told it would take perhaps 2 months to get to release (i.e. mid-Summer). Then as recently as two months or so ago we were told it would be released by the end of 2004. Now the guess if for the end of January. Well, I keep making plans, and those plans are constantly being changed, which has cost me dearly. I know Ruslan (and Ivan, Serg, and the rest) are working very diligently to get V2 out. And if it takes time, well, that's the way it is. But I do wish the process was more transparent and that forecasting was more accurate. The market for V2 is developers, and we are all well aware of the pitfalls and delays that can be encountered. But we have to make plans, too, and would all be better served if realistic predictions were made and we were kept in the loop. I am now aiming to release my V2-based app at the end of January. Will that happen -- based on past experience, I'd have to say no. A stable and fast release by March 20 for you? I won't guess. Jon From lfredricks at proactive-intl.com Mon Jan 3 09:33:26 2005 From: lfredricks at proactive-intl.com (Lynn Fredricks) Date: Mon Jan 3 11:33:38 2005 Subject: valentina 2 In-Reply-To: Message-ID: <20050103173325.A308C17D03D@spatula.dreamhost.com> > Initially Unicode was going to be added to V1, and that is > all I really need. But that intention was changed for > financial reasons, and it was decided that Unicode was only > going to be included in V2. > > V2 has been delayed, delayed, delayed, which has caused me a > lot of grief. I remember on the Valentina list in 2003 (!) > when Lynn assured me that V2 would be out *before* RB 5.5. > Gosh, was that wrong. When the V2 beta was first posted (May > 11, 2004 I believe) we were told it would take perhaps 2 > months to get to release (i.e. mid-Summer). Then as recently > as two months or so ago we were told it would be released by > the end of 2004. Now the guess if for the end of January. > Well, I keep making plans, and those plans are constantly > being changed, which has cost me dearly. It has slipped from the end of 2004 after much discussion and debate internally. The wait for at least some products is almost over. > I know Ruslan (and Ivan, Serg, and the rest) are working very > diligently to get V2 out. And if it takes time, well, that's > the way it is. But I do wish the process was more transparent > and that forecasting was more accurate. The market for V2 is > developers, and we are all well aware of the pitfalls and > delays that can be encountered. But we have to make plans, > too, and would all be better served if realistic predictions > were made and we were kept in the loop. Unfortunately, the forecasts have not been very accurate to date. Im hoping we will be able to change that very soon. That is one reason why, as you noticed above, we've added additional staff to take care of individual sub-systems of Valentina 2. > I am now aiming to release my V2-based app at the end of January. > Will that happen -- based on past experience, I'd have to say > no. A stable and fast release by March 20 for you? I won't guess. I hope we can prove you wrong on this one, Jon. We had originally planned on releasing updates to all V1 based products all at once. Instead, we are going to concentrate on getting the most popular ones out first, with the rest trailing by a few months. For the most popular products, we are very, very close to release. As Ruslan said, we are going to have some additional optimizations. Its going to be possible for you to ship products without those optimizations, but you will benefit from them when they do arrive. In fact, there are several companies who have already purchased V2-based products in their current state and have been shipping their products for quite some time. Best regards, Lynn Fredricks President Proactive International, LLC - Because it is about who you know.(tm) From sunshine at public.kherson.ua Mon Jan 3 20:40:54 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Mon Jan 3 12:41:01 2005 Subject: valentina 2 In-Reply-To: Message-ID: On 1/3/05 7:21 PM, "jda" wrote: > Your more fundamental question is whether a stable V2 is on a > schedule that fits your (and my) needs. This has troubled me quite a > lot (and Ruslan is tired of hearing from me, I know). >From all this I want comment just stability. We have now about 20,000 tests DONE, and only 30 FAILs. Fails are related to sort of join table query. So we have great sure that starting from 2.0 Valentina will be as stable as never before. And we have sure that RELEASE 2.0 will be very stable from start. -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From jda at his.com Mon Jan 3 14:34:37 2005 From: jda at his.com (jda) Date: Mon Jan 3 13:34:46 2005 Subject: valentina 2 In-Reply-To: References: Message-ID: > >From all this I want comment just stability. > >We have now about 20,000 tests DONE, and only 30 FAILs. >Fails are related to sort of join table query. > >So we have great sure that starting from 2.0 Valentina will >be as stable as never before. > >And we have sure that RELEASE 2.0 will be very stable from start. > If you want stability, you should start converting to V2 *now*. Very few people outside of Paradigma are actually reporting bugs...the more that use it the more bugs will be uncovered and squashed before release. Jon From sunshine at public.kherson.ua Mon Jan 3 21:51:55 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Mon Jan 3 13:52:01 2005 Subject: valentina 2 In-Reply-To: Message-ID: On 1/3/05 9:34 PM, "jda" wrote: >>> From all this I want comment just stability. >> >> We have now about 20,000 tests DONE, and only 30 FAILs. >> Fails are related to sort of join table query. >> >> So we have great sure that starting from 2.0 Valentina will >> be as stable as never before. >> >> And we have sure that RELEASE 2.0 will be very stable from start. >> > > If you want stability, you should start converting to V2 *now*. Very > few people outside of Paradigma are actually reporting bugs...the > more that use it the more bugs will be uncovered and squashed before > release. Jon, but this will not speed up nothing. Assume we fix tomorrow all YOUR bug reports and of Frank. We still have a work todo V4MD MAC Vserver MAC Docs More examples. During docs development yes, we will need as many bug reports as possible. -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From frank-list3 at mindstarprods.com Mon Jan 3 12:53:59 2005 From: frank-list3 at mindstarprods.com (Frank Schima) Date: Mon Jan 3 13:53:52 2005 Subject: [V4RB] API Equivalent of UPDATE WHERE Message-ID: <35D16E86-5DC1-11D9-A907-0003939246BC@mindstarprods.com> Hi all, With all this talk about the API way being faster than SQL, I was wondering what is the equivalent code for UPDATE... WHERE... with the new V4RB API? UpdateRecord will update a single record, and UpdateAllRecords will update all the records of a table/cursor. So how can I update a subset of records without looping through them all and using UpdateRecord? Basically I'm looking for an API workaround for that UPDATE SQL statement that is crashing for me. Namely: UPDATE MyTable Set MyVarcharField = 'some text' WHERE MyVarcharField is Null Pseudo-code please? :^) Best regards, Frank From sunshine at public.kherson.ua Mon Jan 3 22:15:21 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Mon Jan 3 14:15:29 2005 Subject: [V4RB] API Equivalent of UPDATE WHERE In-Reply-To: <35D16E86-5DC1-11D9-A907-0003939246BC@mindstarprods.com> Message-ID: On 1/3/05 9:53 PM, "Frank Schima" wrote: Hi Frank, > With all this talk about the API way being faster than SQL, I was > wondering what is the equivalent code for UPDATE... WHERE... with the > new V4RB API? > > UpdateRecord will update a single record, and UpdateAllRecords will > update all the records of a table/cursor. So how can I update a subset > of records without looping through them all and using UpdateRecord? > > Basically I'm looking for an API workaround for that UPDATE SQL > statement that is crashing for me. Namely: > > UPDATE MyTable Set MyVarcharField = 'some text' WHERE MyVarcharField > is Null > > Pseudo-code please? :^) * But what is bad in looping ? if this is local db then no problems in fact, RB is fast on this. * API way will require more coding. This is a RULE. so the pure API way can be: set = fldVarChar.FindNulls() Iter = set.MakeNewIterator recid = Iter.first do tbl.RecID = recid fldVarChar.value = "some text" tbl.UpdateRecord() while recid = iter.next * Also your task CAN be made with UpdateAllRecords() as I see. curs = db.SqlSelect( "SELECT MyVarcharField WHERE MyVarcharField IS NULL" ) curs.VarCharField(x).valule = "someText" curs.UpdateAllRecords() * keep in mind that API way cannot do JOINS, GROUP BY and so on. this is RELATIONAL Model features, mirrored in the SQL. -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From jda at his.com Mon Jan 3 15:29:02 2005 From: jda at his.com (jda) Date: Mon Jan 3 14:29:06 2005 Subject: valentina 2 In-Reply-To: References: Message-ID: > > If you want stability, you should start converting to V2 *now*. Very >> few people outside of Paradigma are actually reporting bugs...the >> more that use it the more bugs will be uncovered and squashed before >> release. > >Jon, but this will not speed up nothing. > >Assume we fix tomorrow all YOUR bug reports and of Frank. Yes, let's assume it. I like that. :-> I didn't say it would speed things up -- I said it would make it more *stable*. >We still have a work todo > >V4MD MAC >Vserver MAC >Docs >More examples. UTF-8 storage, V1 to V2 converter, compression... Jon From frank-list3 at mindstarprods.com Mon Jan 3 13:35:17 2005 From: frank-list3 at mindstarprods.com (Frank Schima) Date: Mon Jan 3 14:35:06 2005 Subject: [V4RB] API Equivalent of UPDATE WHERE In-Reply-To: References: Message-ID: On Jan 3, 2005, at 1:15 PM, Ruslan Zasukhin wrote: > On 1/3/05 9:53 PM, "Frank Schima" > wrote: > > Hi Frank, > >> With all this talk about the API way being faster than SQL, I was >> wondering what is the equivalent code for UPDATE... WHERE... with the >> new V4RB API? >> >> UpdateRecord will update a single record, and UpdateAllRecords will >> update all the records of a table/cursor. So how can I update a subset >> of records without looping through them all and using UpdateRecord? >> >> Basically I'm looking for an API workaround for that UPDATE SQL >> statement that is crashing for me. Namely: >> >> UPDATE MyTable Set MyVarcharField = 'some text' WHERE >> MyVarcharField >> is Null >> >> Pseudo-code please? :^) > > * But what is bad in looping ? > if this is local db then no problems in fact, > RB is fast on this. I thought it would be slower than SQL. > * API way will require more coding. This is a RULE. > > so the pure API way can be: > > set = fldVarChar.FindNulls() > > Iter = set.MakeNewIterator > recid = Iter.first > > do > tbl.RecID = recid > fldVarChar.value = "some text" > tbl.UpdateRecord() > > while recid = iter.next I will try this. > * Also your task CAN be made with UpdateAllRecords() as I see. > > curs = db.SqlSelect( > "SELECT MyVarcharField WHERE MyVarcharField IS NULL" ) > > curs.VarCharField(x).valule = "someText" > > curs.UpdateAllRecords() Well i have some records with text that i want to preserve because it's different than 'some text'. I only want the NULL ones changed. > * keep in mind that API way cannot do JOINS, GROUP BY and so on. > this is RELATIONAL Model features, mirrored in the SQL. So SQL will still be needed then for some tasks then right? Best regards, Frank Schima From sunshine at public.kherson.ua Mon Jan 3 23:24:37 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Mon Jan 3 15:24:43 2005 Subject: valentina 2 In-Reply-To: Message-ID: On 1/3/05 10:29 PM, "jda" wrote: >>> If you want stability, you should start converting to V2 *now*. Very >>> few people outside of Paradigma are actually reporting bugs...the >>> more that use it the more bugs will be uncovered and squashed before >>> release. >> >> Jon, but this will not speed up nothing. >> >> Assume we fix tomorrow all YOUR bug reports and of Frank. > > Yes, let's assume it. I like that. :-> > > I didn't say it would speed things up -- I said it would make it more > *stable*. > >> We still have a work todo >> >> V4MD MAC >> Vserver MAC >> Docs >> More examples. > > UTF-8 storage, V1 to V2 converter, compression... + Byteswap between platforms :-)) You see how many cool tasks you want from us! Each task you list here -- one developer week. So may be we need ship truncated 2.0 release? ;-) Or better wait for complete and full one? -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From sunshine at public.kherson.ua Mon Jan 3 23:30:14 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Mon Jan 3 15:30:18 2005 Subject: [V4RB] API Equivalent of UPDATE WHERE In-Reply-To: Message-ID: On 1/3/05 10:35 PM, "Frank Schima" wrote: >> * But what is bad in looping ? >> if this is local db then no problems in fact, >> RB is fast on this. > > I thought it would be slower than SQL. Frank, all depend on task and environment. In C++ loops will be NEVER slower of SQL, because SQL is parser + the same loops. >> * API way will require more coding. This is a RULE. >> >> so the pure API way can be: >> >> set = fldVarChar.FindNulls() >> >> Iter = set.MakeNewIterator >> recid = Iter.first >> >> do >> tbl.RecID = recid >> fldVarChar.value = "some text" >> tbl.UpdateRecord() >> >> while recid = iter.next > > I will try this. BTW, may be more simple to use set = fldVarChar.FindNulls() fldVarChar.value = "some text" Table.UpdateAllRecords( set ) >> * Also your task CAN be made with UpdateAllRecords() as I see. >> >> curs = db.SqlSelect( >> "SELECT MyVarcharField WHERE MyVarcharField IS NULL" ) >> >> curs.VarCharField(x).valule = "someText" >> >> curs.UpdateAllRecords() > > Well i have some records with text that i want to preserve because it's > different than 'some text'. I only want the NULL ones changed. Frank! But only NULL records will be selected in the cursor. So only they will be updated. >> * keep in mind that API way cannot do JOINS, GROUP BY and so on. >> this is RELATIONAL Model features, mirrored in the SQL. > > So SQL will still be needed then for some tasks then right? Of course! SQL is simply a choice if you work with remote computer. Or if you want realtional way, Or if you need do GROUP BY. On the other hand, many tasks can be made with Navigational model, And even more effective, because you as developer always can write better code that SQL optimizer. -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From jda at his.com Mon Jan 3 16:40:14 2005 From: jda at his.com (jda) Date: Mon Jan 3 15:40:20 2005 Subject: valentina 2 In-Reply-To: References: Message-ID: > >> We still have a work todo >>> >>> V4MD MAC >>> Vserver MAC >>> Docs >>> More examples. >> >> UTF-8 storage, V1 to V2 converter, compression... >+ Byteswap between platforms > >:-)) > >You see how many cool tasks you want from us! >Each task you list here -- one developer week. > >So may be we need ship truncated 2.0 release? ;-) >Or better wait for complete and full one? > :-) There is no point to a truncated release...by which I mean a release that would not automatically take advantage of refinements as they were rolled into V2.0.1, 2.0.2, etc. In my particular case, UTF-8 and the V1/2 converter are necessary on release (of *my* app). Compression could come later (again, as long as my app could use it without making older V2.x databases incompatible). Ruslan, you know that I've been waiting on this for one-and-a-half years. A week or two difference now isn't going to kill me (but 3 months would!!!! ;->). Jon From frank-list3 at mindstarprods.com Mon Jan 3 16:47:10 2005 From: frank-list3 at mindstarprods.com (Frank Schima) Date: Mon Jan 3 17:47:04 2005 Subject: [V4RB] API Equivalent of UPDATE WHERE In-Reply-To: References: Message-ID: On Jan 3, 2005, at 2:30 PM, Ruslan Zasukhin wrote: >>> * Also your task CAN be made with UpdateAllRecords() as I see. >>> >>> curs = db.SqlSelect( >>> "SELECT MyVarcharField WHERE MyVarcharField IS NULL" >>> ) >>> >>> curs.VarCharField(x).valule = "someText" >>> >>> curs.UpdateAllRecords() >> >> Well i have some records with text that i want to preserve because >> it's >> different than 'some text'. I only want the NULL ones changed. > > Frank! > > But only NULL records will be selected in the cursor. > So only they will be updated. [Slaps head] Yes, of course - I get it now! Sorry for being dense (i.e. stupid). I'm so used to thinking in SQL terms and I'm trying to learn to use the nice new Valentina API more in place of SQL where possible for faster code. :^) Best regards, Frank Schima From sunshine at public.kherson.ua Tue Jan 4 02:05:09 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Mon Jan 3 18:05:15 2005 Subject: [V4RB] API Equivalent of UPDATE WHERE In-Reply-To: Message-ID: On 1/4/05 1:47 AM, "Frank Schima" wrote: >>>> * Also your task CAN be made with UpdateAllRecords() as I see. >>>> >>>> curs = db.SqlSelect( >>>> "SELECT MyVarcharField WHERE MyVarcharField IS NULL" >>>> ) >>>> >>>> curs.VarCharField(x).valule = "someText" >>>> >>>> curs.UpdateAllRecords() >>> >>> Well i have some records with text that i want to preserve because >>> it's >>> different than 'some text'. I only want the NULL ones changed. >> >> Frank! >> >> But only NULL records will be selected in the cursor. >> So only they will be updated. > > [Slaps head] Yes, of course - I get it now! Sorry for being dense (i.e. > stupid). I'm so used to thinking in SQL terms and I'm trying to learn > to use the nice new Valentina API more in place of SQL where possible > for faster code. Please note that THIS WAY (cursor.UpdateAll) will NOT add speed I think Just it is little other SQL way based on the API of Vcursor class. UPDATE WHERE is more powerful command although, because I can do with it SET f = f + 1 WHERE f > 10 Curs.UpdateAll() can update all recs of cursor by THE SAME value. Curs.UpdateAll() was in the 1.x also btw. -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From jda at his.com Mon Jan 3 19:21:19 2005 From: jda at his.com (jda) Date: Mon Jan 3 18:21:33 2005 Subject: [V4RB] API Equivalent of UPDATE WHERE In-Reply-To: References: Message-ID: > > >UPDATE WHERE is more powerful command although, because I can do with it > > SET f = f + 1 WHERE f > 10 > Hm, is there a fast way to set the values of a numeric field to a sequence without looping through each record in a cursor? Say if there is 1000 records in the cursor, setting field id to 1 2 3 4 ... 1000 I do that now, but as I said, I loop through each record in a cursor and set the value. It works, but gets kind of slow when there are 5000-10000 records. Jon From vidal_olivier at yahoo.fr Tue Jan 4 06:12:38 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Mon Jan 3 23:13:45 2005 Subject: valentina 2 In-Reply-To: References: <24CCD684-5DA9-11D9-A02F-000D93C35C66@yahoo.fr> Message-ID: <408426EA-5E0F-11D9-8F37-000D93C35C66@yahoo.fr> Thank you very much Jon for all this precisions. It is important to know of what think the customers who have already used valentina 2. Olivier Le 3 janv. 05, ? 18:21, jda a ?crit : >> > > I wouldn't say that V2 is *much* slower than V1 right now, at least > for me...the decrease in speed of addrecord and indexing is > bothersome, but Ruslan has said it will be improved, and I believe > him. > > Your more fundamental question is whether a stable V2 is on a schedule > that fits your (and my) needs. This has troubled me quite a lot (and > Ruslan is tired of hearing from me, I know). > > Initially Unicode was going to be added to V1, and that is all I > really need. But that intention was changed for financial reasons, and > it was decided that Unicode was only going to be included in V2. > > V2 has been delayed, delayed, delayed, which has caused me a lot of > grief. I remember on the Valentina list in 2003 (!) when Lynn assured > me that V2 would be out *before* RB 5.5. Gosh, was that wrong. When > the V2 beta was first posted (May 11, 2004 I believe) we were told it > would take perhaps 2 months to get to release (i.e. mid-Summer). Then > as recently as two months or so ago we were told it would be released > by the end of 2004. Now the guess if for the end of January. Well, I > keep making plans, and those plans are constantly being changed, which > has cost me dearly. > > I know Ruslan (and Ivan, Serg, and the rest) are working very > diligently to get V2 out. And if it takes time, well, that's the way > it is. But I do wish the process was more transparent and that > forecasting was more accurate. The market for V2 is developers, and we > are all well aware of the pitfalls and delays that can be encountered. > But we have to make plans, too, and would all be better served if > realistic predictions were made and we were kept in the loop. > > I am now aiming to release my V2-based app at the end of January. Will > that happen -- based on past experience, I'd have to say no. A stable > and fast release by March 20 for you? I won't guess. > > Jon > _______________________________________________ > Valentina-beta mailing list > Valentina-beta@lists.macserve.net > http://lists.macserve.net/mailman/listinfo/valentina-beta > From vidal_olivier at yahoo.fr Tue Jan 4 06:15:50 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Mon Jan 3 23:16:42 2005 Subject: valentina 2 In-Reply-To: <24CCD684-5DA9-11D9-A02F-000D93C35C66@yahoo.fr> References: <24CCD684-5DA9-11D9-A02F-000D93C35C66@yahoo.fr> Message-ID: Hi Ruslan, that think of : > hmmm if I indeed understands, valentina 2 is much slower for the > moment than valentina 1 by using SQL but about so fast by using API ? > > As a lot of people I think, I hesitate to use Valentina 2 from now on. > I began a project with Valentina 1. This project must be absolutely > ended by March 20th. I am very enthusiastic to use valentina 2 (by > API) for this project now because if the version 1.0 of my software > will be already based on valentina 2, I can probably make much more > quickly updates for the next versions. My software will be built well > from the beginning. > > On the other hand, if my software 1.0 is not stable, finally not much > more speed than with valentina 1 and if I lose a lot of time in the > corrections of (valentina 2) bugs, I hesitate and I wonder if I am not > going to wait for my next project to use valentina 2. > > As customer and user of valentina, of whom do you think of it ? > > PS : it is damage that there not in the documentation some pages to > explain "philosophy" and paradigmes of valentina 2. > > olivier > thank you olivier From vidal_olivier at yahoo.fr Tue Jan 4 06:24:28 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Mon Jan 3 23:25:31 2005 Subject: valentina 2 In-Reply-To: References: Message-ID: Ruslan, why to have chosen UTF-16 while RB (and Mac) uses everywhere UTF-8 ? of what consists this converter V1/2 ? it is a tool which will put in the format 2 databases 1 ? or it is a tool which will also help to clean the code SQL / Valentina in RB ? thank you olivier Le 3 janv. 05, ? 22:40, jda a ?crit : > UTF-8 and the V1/2 converter are necessary on release (of *my* app) From vidal_olivier at yahoo.fr Tue Jan 4 07:02:11 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Tue Jan 4 00:03:39 2005 Subject: doc Message-ID: <2CA78E46-5E16-11D9-8F37-000D93C35C66@yahoo.fr> Hi, in "V4RB_reference_en.pdf" : page 17, method "create" : it is indicated (in the text and the examples) that the function returns a boolean while apparently it is not any more the case. page 18, method "close" : similar (in the example) page 18, method "open" : similar (in the example and the title) olivier From vidal_olivier at yahoo.fr Tue Jan 4 07:25:33 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Tue Jan 4 00:26:35 2005 Subject: doc Message-ID: <7046B3A6-5E19-11D9-8F37-000D93C35C66@yahoo.fr> in "V4RB_reference_en.pdf" : page 73, "Vstring/Vvarchar" : in constructor, it is indicated the parameter "inLanguage" while apparently it is not any more used. From vidal_olivier at yahoo.fr Tue Jan 4 08:30:27 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Tue Jan 4 01:31:44 2005 Subject: doc vfield Message-ID: <81772FF4-5E22-11D9-8F37-000D93C35C66@yahoo.fr> in "V4RB_reference_en.pdf" : page 57, "isIndexed" : in example, the old "indexed" is used on the place of "isIndexed" From vidal_olivier at yahoo.fr Tue Jan 4 08:33:26 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Tue Jan 4 01:34:34 2005 Subject: doc vfield Message-ID: in "V4RB_reference_en.pdf" : page 57, "isUnique" : in example, the old "unique" is used on the place of "isUnique". From vidal_olivier at yahoo.fr Tue Jan 4 09:20:49 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Tue Jan 4 02:22:06 2005 Subject: doc vTable - addrecord Message-ID: <8AB1019E-5E29-11D9-8F37-000D93C35C66@yahoo.fr> The method "addRecord" returns now an integer. In the documentation, page 39 (V4RB reference), the description of the method "addrecord" does not indicate of what consists this integer. In the example, "Addrecord" does not return value. From vidal_olivier at yahoo.fr Tue Jan 4 09:49:16 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Tue Jan 4 02:50:36 2005 Subject: doc Vcursor Importtext + Addrecord Message-ID: <83EFCDF0-5E2D-11D9-8F37-000D93C35C66@yahoo.fr> in V4RB reference, page 110, "Vcursor", the description of method "Import_Text" indicate the old "ImportText" on the place of new "Import_Text". Same problem in the example. In the description of the parameters, the new parameter "inHasColumnHeader" is not put in brackets to indicate that it is optional while the others parameters (delimiters) are now, them, indispensable. page 107, the description of the method "addRecord" does not indicate of what consists the returned integer. in the example, it is old "add" what is used on the place of "addrecord". From sunshine at public.kherson.ua Tue Jan 4 10:58:56 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Tue Jan 4 02:59:16 2005 Subject: [V4RB] API Equivalent of UPDATE WHERE In-Reply-To: Message-ID: On 1/4/05 2:21 AM, "jda" wrote: >> UPDATE WHERE is more powerful command although, because I can do with it >> >> SET f = f + 1 WHERE f > 10 >> > > Hm, is there a fast way to set the values of a numeric field to a > sequence without looping through each record in a cursor? Say if > there is 1000 records in the cursor, setting field id to > 1 > 2 > 3 > 4 > ... > 1000 > > I do that now, but as I said, I loop through each record in a cursor > and set the value. It works, but gets kind of slow when there are > 5000-10000 records. In normal dbs there is no such way. You must loop. Or engine must loop. Jon, IMPORTANT: * if you change MANY records, you may consider remove index. not-indexed field give linear time for adding. Btw, I have see that import of 50 records with PUB format, You make with indexes ON. Remove indexes, and I think import of 4000 records will be done not in 12 minutes but in few seconds. -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From vidal_olivier at yahoo.fr Tue Jan 4 09:58:53 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Tue Jan 4 03:00:24 2005 Subject: doc Vcursor UpdateRecord + UpdateAllRecord Message-ID: in doc "V4RB reference", page 108, "vCursor", the description (and title) of method "updateRecord" indicates that a boolean returned while it is not any more the case. In the example, it is the old ".update()" which is used on the place of the new ".updateRecord". Same problems for "UpdateAllRecord". From sunshine at public.kherson.ua Tue Jan 4 11:02:44 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Tue Jan 4 03:02:50 2005 Subject: valentina 2 In-Reply-To: Message-ID: On 1/4/05 7:24 AM, "Olivier" wrote: Hi Oliver, > Ruslan, > > why to have chosen UTF-16 while RB (and Mac) uses everywhere UTF-8 ? Because it is better. And this is why we get so hard task. RB have go by simple way > of what consists this converter V1/2 ? It will be integrated > it is a tool which will put in the format 2 databases 1 ? Converter only. > or it is a tool which will also help to clean the code SQL / Valentina > in RB ? Code? No. We will make just a list of advices instructions. -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From vidal_olivier at yahoo.fr Tue Jan 4 10:44:10 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Tue Jan 4 03:45:49 2005 Subject: vTable - deleteRecord() Message-ID: <2F34B8E3-5E35-11D9-8F37-000D93C35C66@yahoo.fr> It is strange that methods as "deleteRecord()" (vTable) contain brackets while they are not functions and what they do not return value. From sunshine at public.kherson.ua Tue Jan 4 12:13:18 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Tue Jan 4 04:13:33 2005 Subject: vTable - deleteRecord() In-Reply-To: <2F34B8E3-5E35-11D9-8F37-000D93C35C66@yahoo.fr> Message-ID: On 1/4/05 11:44 AM, "Olivier" wrote: > It is strange that methods as "deleteRecord()" (vTable) contain > brackets while they are not functions and what they do not return > value. In fact this is MY special action. In C++ all functions have () Pascal like REALbasic, allows to skip (). I do not like this personally. obj.Something -- is this property of procedure? You cannot answer. I prefer have obj.Property obj.Proc() Why? Because procedure can be something big, It can CHANGE some other objects. And so on. Property always -- this is just get its value. So I prefer explicitly underline that this is PROCEDURE. Something potentially big. -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From vidal_olivier at yahoo.fr Tue Jan 4 11:36:10 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Tue Jan 4 04:37:33 2005 Subject: vTable - deleteRecord() In-Reply-To: References: Message-ID: <7328EB39-5E3C-11D9-8F37-000D93C35C66@yahoo.fr> > > obj.Something -- is this property of procedure? > > You cannot answer. > if it's a property, always : obj.something.set.... or obj.something.get... or obj.something=.... if it's method : obj.something if it's function : rs=obj.something() It is the functioning of RB. It is always better if plug-in have the same functioning I think. If valentina has a different behavior, ok, but it would be necessary, in that case, whether it is for all its methods. For example "cur.updateRecord" has no brackets. Why ? All this would have to be homogeneous. Another example. Now the parameter "indexed" of a Vfield is "isIndexed" ("unique" is now "isUnique"...) while "indexedByWord" is always "indexedByWord". It is a detail naturally and it is not very important but it would indeed be good if they had all the same logic. > I prefer have > > obj.Property > obj.Proc() > > Why? Because procedure can be something big, > It can CHANGE some other objects. And so on. > > Property always -- this is just get its value. > > So I prefer explicitly underline that this is PROCEDURE. > Something potentially big. > > > -- > Best regards, > Ruslan Zasukhin [ I feel the need...the need for speed ] > ------------------------------------------------------------- > e-mail: ruslan@paradigmasoft.com > web: http://www.paradigmasoft.com > > To subscribe to the Valentina mail list go to: > http://lists.macserve.net/mailman/listinfo/valentina > ------------------------------------------------------------- > > > _______________________________________________ > Valentina-beta mailing list > Valentina-beta@lists.macserve.net > http://lists.macserve.net/mailman/listinfo/valentina-beta > From sunshine at public.kherson.ua Tue Jan 4 12:50:16 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Tue Jan 4 04:50:24 2005 Subject: vTable - deleteRecord() In-Reply-To: <7328EB39-5E3C-11D9-8F37-000D93C35C66@yahoo.fr> Message-ID: On 1/4/05 12:36 PM, "Olivier" wrote: >> obj.Something -- is this property of procedure? >> >> You cannot answer. >> > if it's a property, always : > > obj.something.set.... > or > obj.something.get... > or > obj.something=.... > > if it's method : > > obj.something > > if it's function : > > rs=obj.something() So res = obj.something() // function obj.something // proc res = obj.something // property Well may be... > It is the functioning of RB. > It is always better if plug-in have the same functioning I think. But RB do allow us obj.something() // proc Right? Then what problems? > If valentina has a different behavior, ok, but it would be necessary, > in that case, whether it is for all its methods. For example > "cur.updateRecord" has no brackets. Why ? Oliver, this is not different behavior! This is just syntax. You can use () for proc, and you can NOT use them. Up to you. REALbasic allow both ways. Or I am wrong? And RB will complain on such syntax ? obj.something() > All this would have to be homogeneous. > > Another example. > Now the parameter "indexed" of a Vfield is "isIndexed" ("unique" is now > "isUnique"...) while "indexedByWord" is always "indexedByWord". It is a > detail naturally and it is not very important but it would indeed be > good if they had all the same logic. So you have idea how to make good English for IsIndexByWords ? -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From vidal_olivier at yahoo.fr Tue Jan 4 13:15:03 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Tue Jan 4 06:16:00 2005 Subject: vTable - deleteRecord() In-Reply-To: References: Message-ID: <43275CD2-5E4A-11D9-8F37-000D93C35C66@yahoo.fr> Le 4 janv. 05, ? 11:50, Ruslan Zasukhin a ?crit : > This is just syntax. From vidal_olivier at yahoo.fr Tue Jan 4 12:58:14 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Tue Jan 4 06:16:02 2005 Subject: vTable - deleteRecord() In-Reply-To: References: Message-ID: ok Ruslan. Le 4 janv. 05, ? 11:50, Ruslan Zasukhin a ?crit : > This is just syntax. From jda at his.com Tue Jan 4 07:41:39 2005 From: jda at his.com (jda) Date: Tue Jan 4 06:41:50 2005 Subject: [V4RB] API Equivalent of UPDATE WHERE In-Reply-To: References: Message-ID: >IMPORTANT: > >* if you change MANY records, you may consider remove index. > > not-indexed field give linear time for adding. > >Btw, I have see that import of 50 records with PUB format, >You make with indexes ON. > >Remove indexes, and I think import of 4000 records will be done not in 12 >minutes but in few seconds. > Actually, I tried that. Importing was only *slightly* faster for the first several thousand records (it did maintain speed better when it got to 10000 records, though)! And in V1.x (to which I compare V2), importing was much faster with indexing ON. The problem with turning indexing OFF is that my app doesn't know how may records are going to be imported beforehand. Therefore, if only one is being imported (quite common), when a large db (say 30000 records) is *reindexed* after importing that one record it is going to take a long time. And having the user decide whether to have indexing on or off during a particular import is too techie for a consumer app. So I leave it ON all the time, which as I said, worked well for V1. Jon From sunshine at public.kherson.ua Tue Jan 4 14:56:03 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Tue Jan 4 06:56:16 2005 Subject: speed of importing In-Reply-To: Message-ID: On 1/4/05 2:41 PM, "jda" wrote: >> IMPORTANT: >> >> * if you change MANY records, you may consider remove index. >> >> not-indexed field give linear time for adding. >> >> Btw, I have see that import of 50 records with PUB format, >> You make with indexes ON. >> >> Remove indexes, and I think import of 4000 records will be done not in 12 >> minutes but in few seconds. >> > > Actually, I tried that. Importing was only *slightly* faster for the > first several thousand records (it did maintain speed better when it > got to 10000 records, though)! And in V1.x (to which I compare V2), > importing was much faster with indexing ON. No, this can't be. What size of data you import? What size of cache ? What size of final db ? > The problem with turning indexing OFF is that my app doesn't know how > may records are going to be imported beforehand. Therefore, if only > one is being imported (quite common), when a large db (say 30000 > records) is *reindexed* after importing that one record it is going > to take a long time. And having the user decide whether to have > indexing on or off during a particular import is too techie for a > consumer app. So I leave it ON all the time, which as I said, worked > well for V1. Number of records to import you can estimate by the size of file. -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From jda at his.com Tue Jan 4 08:00:54 2005 From: jda at his.com (jda) Date: Tue Jan 4 07:00:58 2005 Subject: speed of importing In-Reply-To: References: Message-ID: > > Actually, I tried that. Importing was only *slightly* faster for the >> first several thousand records (it did maintain speed better when it >> got to 10000 records, though)! And in V1.x (to which I compare V2), >> importing was much faster with indexing ON. > >No, this can't be. > >What size of data you import? >What size of cache ? >What size of final db ? > I'll do a benchmark and let you know. Jon From sunshine at public.kherson.ua Tue Jan 4 15:03:37 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Tue Jan 4 07:03:45 2005 Subject: speed of importing In-Reply-To: Message-ID: On 1/4/05 3:00 PM, "jda" wrote: >>> Actually, I tried that. Importing was only *slightly* faster for the >>> first several thousand records (it did maintain speed better when it >>> got to 10000 records, though)! And in V1.x (to which I compare V2), >>> importing was much faster with indexing ON. >> >> No, this can't be. >> >> What size of data you import? >> What size of cache ? >> What size of final db ? Sense in benchamrk? In 2.0 indexes are in debug mode and not linked to cache yet -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From jda at his.com Tue Jan 4 08:19:29 2005 From: jda at his.com (jda) Date: Tue Jan 4 07:19:38 2005 Subject: speed of importing In-Reply-To: References: Message-ID: > >> No, this can't be. >>> >>> What size of data you import? >>> What size of cache ? >>> What size of final db ? > >Sense in benchamrk? > >In 2.0 indexes are in debug mode and not linked to cache yet > Then why ask these questions? I told you that turning off indexing during import had only a small affect on import speed. What more do you want, then? Jon From vidal_olivier at yahoo.fr Tue Jan 4 14:27:42 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Tue Jan 4 07:28:49 2005 Subject: doc Vcursor UpdateRecord + UpdateAllRecord In-Reply-To: References: Message-ID: <696113A2-5E54-11D9-8F37-000D93C35C66@yahoo.fr> Several functions (as " updateRecord ") do not return not any more a boolean. It was practical to verify if the operation was OK. Why we do not have more boolean returned? RB 5.5.4, valentina 2 (december 18 version : with the finder, We cannot see the number of version. It would indeed be good to indicate it) Le 4 janv. 05, ? 09:58, Olivier a ?crit : > in doc "V4RB reference", page 108, "vCursor", the description (and > title) of method "updateRecord" indicates that a boolean returned > while it is not any more the case. > From sunshine at public.kherson.ua Tue Jan 4 15:41:18 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Tue Jan 4 07:41:26 2005 Subject: doc Vcursor UpdateRecord + UpdateAllRecord In-Reply-To: <696113A2-5E54-11D9-8F37-000D93C35C66@yahoo.fr> Message-ID: On 1/4/05 3:27 PM, "Olivier" wrote: > Several functions (as " updateRecord ") do not return not any more a > boolean. > It was practical to verify if the operation was OK. > Why we do not have more boolean returned? Because now in case of any problems will be throw exception -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From rjb at robelko.com Tue Jan 4 14:47:24 2005 From: rjb at robelko.com (Robert Brenstein) Date: Tue Jan 4 08:01:15 2005 Subject: valentina 2 Message-ID: > >PS : it is damage that there not in the documentation some pages to >explain "philosophy" and paradigmes of valentina 2. > Ruslan, this brings back my suggestion from last year: you may want to reconsider reworking that article that never appeared in MacTech as a white (philosophy) paper for Valentinas. There are many different database products and it may be helpful for potential and new customers to understand where Valentina is positioned within the crowd, particularly since the OO databases is a somewhat wishy-washy area in terms what term really means what to whom. Lynn, wouldn't you agree? Robert Brenstein From sunshine at public.kherson.ua Tue Jan 4 16:15:35 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Tue Jan 4 08:15:41 2005 Subject: speed of importing In-Reply-To: Message-ID: On 1/4/05 3:19 PM, "jda" wrote: >>>> No, this can't be. >>>> >>>> What size of data you import? >>>> What size of cache ? >>>> What size of final db ? >> >> Sense in benchamrk? >> >> In 2.0 indexes are in debug mode and not linked to cache yet >> > > Then why ask these questions? I told you that turning off indexing > during import had only a small affect on import speed. What more do > you want, then? Because I want to see size of FILE, CACHE and DB -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From sunshine at public.kherson.ua Tue Jan 4 16:21:50 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Tue Jan 4 08:21:55 2005 Subject: valentina 2 In-Reply-To: Message-ID: On 1/4/05 3:47 PM, "Robert Brenstein" wrote: >> >> PS : it is damage that there not in the documentation some pages to >> explain "philosophy" and paradigmes of valentina 2. >> > > Ruslan, this brings back my suggestion from last year: you may want > to reconsider reworking that article that never appeared in MacTech > as a white (philosophy) paper for Valentinas. There are many > different database products and it may be helpful for potential and > new customers to understand where Valentina is positioned within the > crowd, particularly since the OO databases is a somewhat wishy-washy > area in terms what term really means what to whom. Lynn, wouldn't you > agree? Yes, Robert, we plan such texts. Because Valentina 2.0 this is: * cross Relational + OR + Navigational models * cross disk + RAM based db * cross local + client/server * cross different API * cross platfrom * cross langugaes We try to join areas that nobody else join. Other products usually work only in ONE OF this areas. P.S. Valentina is NOT OO-DBMS in terms of the ODMG standard. I have told I do not like their terms and not agree. in particular I not agree that OO DBMS must work only with Java, C++ and Smalltalk. -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From jda at his.com Tue Jan 4 09:22:41 2005 From: jda at his.com (jda) Date: Tue Jan 4 08:22:53 2005 Subject: speed of importing In-Reply-To: References: Message-ID: > >Because I want to see size of FILE, CACHE and DB > It doesn't really matter, since I see this with any large file. But FYI, I made a file that I use to test with 1000 records (932 KB), use a cache of 8 MB, and the db size after import is 5.7 MB. Jon From sunshine at public.kherson.ua Tue Jan 4 16:27:53 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Tue Jan 4 08:27:59 2005 Subject: speed of importing In-Reply-To: Message-ID: On 1/4/05 4:22 PM, "jda" wrote: > Because I want to see size of FILE, CACHE and DB >> > > It doesn't really matter, since I see this with any large file. But > FYI, I made a file that I use to test with 1000 records (932 KB), use > a cache of 8 MB, and the db size after import is 5.7 MB. So problem NOT in the cache overhead. Hmm. This db must be able created in 1 second ideally. With indexing longer of course. You have big texts. Each record means adding about 1000 points into indexes. So it looks like you have db in 400,000 records. But still... -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From lfredricks at proactive-intl.com Tue Jan 4 06:37:10 2005 From: lfredricks at proactive-intl.com (Lynn Fredricks) Date: Tue Jan 4 08:37:29 2005 Subject: valentina 2 In-Reply-To: Message-ID: <20050104143713.6BD9117D024@spatula.dreamhost.com> > > why to have chosen UTF-16 while RB (and Mac) uses everywhere UTF-8 ? > > Because it is better. > > And this is why we get so hard task. > > RB have go by simple way UTF-16 allows for more accurate sorting of Asian languages like Japanese and Chinese. UTF-8 is a design decision. My understanding is that both MacOS X and Windows XP use UTF-16 and UTF-8 at different levels http://developer.apple.com/documentation/MacOSX/Conceptual/SystemOverview/In ternational/chapter_10_section_2.html Not supporting UTF-8 was a design decision of REAL Software, not Paradigma Software ;-) Best regards, Lynn Fredricks President Proactive International, LLC - Because it is about who you know.(tm) From jda at his.com Tue Jan 4 09:38:55 2005 From: jda at his.com (jda) Date: Tue Jan 4 08:39:00 2005 Subject: speed of importing In-Reply-To: References: Message-ID: >So problem NOT in the cache overhead. > > >Hmm. This db must be able created in 1 second ideally. > >With indexing longer of course. You have big texts. >Each record means adding about 1000 points into indexes. > >So it looks like you have db in 400,000 records. > >But still... > I turned OFF indexing and imported these 1000 records -> ~ 120 seconds With indexing ON -> ~ 112 seconds (times are approximate, so I'd say there was no difference). Jon From jda at his.com Tue Jan 4 09:41:20 2005 From: jda at his.com (jda) Date: Tue Jan 4 08:41:26 2005 Subject: valentina 2 In-Reply-To: <20050104143713.6BD9117D024@spatula.dreamhost.com> References: <20050104143713.6BD9117D024@spatula.dreamhost.com> Message-ID: > > > why to have chosen UTF-16 while RB (and Mac) uses everywhere UTF-8 ? >> >> Because it is better. >> >> And this is why we get so hard task. >> >> RB have go by simple way > >UTF-16 allows for more accurate sorting of Asian languages like Japanese and >Chinese. UTF-8 is a design decision. My understanding is that both MacOS X >and Windows XP use UTF-16 and UTF-8 at different levels > >http://developer.apple.com/documentation/MacOSX/Conceptual/SystemOverview/In >ternational/chapter_10_section_2.html > >Not supporting UTF-8 was a design decision of REAL Software, not Paradigma >Software ;-) > I can't let this pass -- UTF-16 and UTF-8 are both good, and each is better *at different things*. For Roman languages with the occasional accented character, UTF-8 is much more efficient! For Asian languages UTF-16 is more efficient. As a developer, you know what is best for you and your users, and you should be able to chose accordingly. For me, that's UTF-8. Jon From sunshine at public.kherson.ua Tue Jan 4 16:42:18 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Tue Jan 4 08:42:26 2005 Subject: valentina 2 In-Reply-To: <20050104143713.6BD9117D024@spatula.dreamhost.com> Message-ID: On 1/4/05 4:37 PM, "Lynn Fredricks" wrote: >>> why to have chosen UTF-16 while RB (and Mac) uses everywhere UTF-8 ? Btw, only now have to see. MAC use UTF16 internally. OS X uses in the background the same IBM ICU library that Valentina 2.0 does. Windows also internally works with utf16 ---------- With UTF8 work only software products that do not have time or power to switch to UTF16. Just C++ app much more simple hack to to UTF8 support then to UTF16. -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From sunshine at public.kherson.ua Tue Jan 4 16:53:46 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Tue Jan 4 08:53:56 2005 Subject: [V4RB 2.0 b34] uploaded Message-ID: kernel - 2.0 b34 ================ - 0000251: [Indexes] IndexStyle does not save properties like IgnoreLength to the schema. (Ivan Smahin) - 0000248: [SysTables] IndexByWords flag is lost after Close/Open() (Ruslan Zasukhin) - 0000247: [Cursors] There is a crash if db is closed with alive cursor over temporary table. (Ivan Smahin) - 0000246: [Volumes] Crash if external reference exists after destroying of Volume. (Ruslan Zasukhin) V4RB - 2.0 b34 ============== - 0000242: [API] Database.Open crashes for my NEW database (Igor Gomon) - 0000219: [API] Corrupted index (Ruslan Zasukhin) -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- -------------- next part -------------- An HTML attachment was scrubbed... URL: http://lists.macserve.net/pipermail/valentina-beta/attachments/20050104/0e663272/attachment.html From sunshine at public.kherson.ua Tue Jan 4 16:55:04 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Tue Jan 4 08:55:10 2005 Subject: speed of importing In-Reply-To: Message-ID: On 1/4/05 4:38 PM, "jda" wrote: >> So problem NOT in the cache overhead. >> >> >> Hmm. This db must be able created in 1 second ideally. >> >> With indexing longer of course. You have big texts. >> Each record means adding about 1000 points into indexes. >> >> So it looks like you have db in 400,000 records. >> >> But still... >> > > I turned OFF indexing and imported these 1000 records -> ~ 120 seconds > > With indexing ON -> ~ 112 seconds > > (times are approximate, so I'd say there was no difference). I think that Valentina self set them ON on some reason. I do not see reason why. Only if a field is marked unique -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From sunshine at public.kherson.ua Tue Jan 4 16:56:10 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Tue Jan 4 08:56:16 2005 Subject: valentina 2 In-Reply-To: Message-ID: On 1/4/05 4:41 PM, "jda" wrote: > I can't let this pass -- UTF-16 and UTF-8 are both good, and each is > better *at different things*. For Roman languages with the occasional > accented character, UTF-8 is much more efficient! For storage, Jon. But not for algorithms. I can ensure you that REALBasic send UTF8 strings to OS X. OS X convert them to UTF16, do required operation and return result. > For Asian languages > UTF-16 is more efficient. As a developer, you know what is best for > you and your users, and you should be able to chose accordingly. For > me, that's UTF-8. -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From lfredricks at proactive-intl.com Tue Jan 4 07:12:44 2005 From: lfredricks at proactive-intl.com (Lynn Fredricks) Date: Tue Jan 4 09:12:52 2005 Subject: valentina 2 In-Reply-To: Message-ID: <20050104151246.660DB17D026@spatula.dreamhost.com> > >PS : it is damage that there not in the documentation some pages to > >explain "philosophy" and paradigmes of valentina 2. > > > > Ruslan, this brings back my suggestion from last year: you > may want to reconsider reworking that article that never > appeared in MacTech as a white (philosophy) paper for > Valentinas. There are many different database products and it > may be helpful for potential and new customers to understand > where Valentina is positioned within the crowd, particularly > since the OO databases is a somewhat wishy-washy area in > terms what term really means what to whom. Lynn, wouldn't you agree? I agree. I don't think Ive seen this article though. Best regards, Lynn Fredricks President Proactive International, LLC - Because it is about who you know.(tm) From lfredricks at proactive-intl.com Tue Jan 4 07:18:13 2005 From: lfredricks at proactive-intl.com (Lynn Fredricks) Date: Tue Jan 4 09:18:19 2005 Subject: valentina 2 In-Reply-To: Message-ID: <20050104151814.DEC2F17D014@spatula.dreamhost.com> > I can't let this pass -- UTF-16 and UTF-8 are both good, and > each is better *at different things*. For Roman languages > with the occasional accented character, UTF-8 is much more > efficient! For Asian languages > UTF-16 is more efficient. As a developer, you know what is > best for you and your users, and you should be able to chose > accordingly. For me, that's UTF-8. That's why UTF-8 is an optional format for Valentina 2 when it is released. You arent going to be forced to use UTF-16 for everything. Best regards, Lynn Fredricks President Proactive International, LLC - Because it is about who you know.(tm) From vidal_olivier at yahoo.fr Tue Jan 4 16:23:37 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Tue Jan 4 09:24:37 2005 Subject: doc Vcursor UpdateRecord + UpdateAllRecord In-Reply-To: References: Message-ID: <9B2C2B28-5E64-11D9-8F37-000D93C35C66@yahoo.fr> Vexception is not in the documentation. Are the numbers of errors of "Vexception.errorNumber" the same that in Valentina 1? In my application (that I began with Valentina 1), in every "addRecord" and " updateRecord ", I looked boolean returned to verify if the recording had been made or not. I have many "UNIQUE" fields. If boolean returned was " false ", that allowed me to know that an identical recording ALREADY existed and I could thus manage it. In that case handbook, what type of error is going to return Vexception? Le 4 janv. 05, ? 14:41, Ruslan Zasukhin a ?crit : > On 1/4/05 3:27 PM, "Olivier" wrote: > >> Several functions (as " updateRecord ") do not return not any more a >> boolean. >> It was practical to verify if the operation was OK. >> Why we do not have more boolean returned? > > Because now in case of any problems will be throw exception > > -- > Best regards, > Ruslan Zasukhin [ I feel the need...the need for speed ] > ------------------------------------------------------------- > e-mail: ruslan@paradigmasoft.com > web: http://www.paradigmasoft.com > > To subscribe to the Valentina mail list go to: > http://lists.macserve.net/mailman/listinfo/valentina > ------------------------------------------------------------- > > > _______________________________________________ > Valentina-beta mailing list > Valentina-beta@lists.macserve.net > http://lists.macserve.net/mailman/listinfo/valentina-beta > From sunshine at public.kherson.ua Tue Jan 4 17:28:06 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Tue Jan 4 09:28:13 2005 Subject: doc Vcursor UpdateRecord + UpdateAllRecord In-Reply-To: <9B2C2B28-5E64-11D9-8F37-000D93C35C66@yahoo.fr> Message-ID: On 1/4/05 5:23 PM, "Olivier" wrote: > Vexception is not in the documentation. Not yet > Are the numbers of errors of "Vexception.errorNumber" the same that in > Valentina 1? on Look into Vcomponents/Resources/English/*.xml files Use MS IE for this. Or other XML browser. Here everything. > In my application (that I began with Valentina 1), in every "addRecord" > and " updateRecord ", I looked boolean returned to verify if the > recording had been made or not. I have many "UNIQUE" fields. If boolean > returned was " false ", that allowed me to know that an identical > recording ALREADY existed and I could thus manage it. In that case > handbook, what type of error is going to return Vexception? Look into our examples. exists EVErrors it seems, with full list of error constants. We have example on unique catch. As far as I remember -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From vidal_olivier at yahoo.fr Tue Jan 4 16:28:33 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Tue Jan 4 09:29:28 2005 Subject: valentina 2 In-Reply-To: <20050104151246.660DB17D026@spatula.dreamhost.com> References: <20050104151246.660DB17D026@spatula.dreamhost.com> Message-ID: <4BBAF1ED-5E65-11D9-8F37-000D93C35C66@yahoo.fr> yes ! it's very important... Le 4 janv. 05, ? 16:12, Lynn Fredricks a ?crit : >> WHITE PAPERS FOR VALENTINA. From vidal_olivier at yahoo.fr Tue Jan 4 16:49:36 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Tue Jan 4 09:50:31 2005 Subject: valentina 2 In-Reply-To: References: Message-ID: <3C289A76-5E68-11D9-8F37-000D93C35C66@yahoo.fr> Le 4 janv. 05, ? 15:42, Ruslan Zasukhin a ?crit : > With UTF8 work only software products that do not have time or power to > switch to UTF16. Strange. In Matt Neuburg's article in " Realbasic developer " N?2.2, page 24, Matt presents us the various Unicode (32,16 and 8). The UTF8 is clearly presented as being the best: "UTF-8 is still more complicated, but it?s also very ingeniously designed to be compact and easily machine-parsable. In contrast to UTF-16, where every string is at least twice as long as its ASCII counterpart, in UTF-8 the 128 ASCII values appear as themselves, a single byte, so that they take up no extra space; ASCII strings are completely interchangeable with UTF-8 strings consisting of just ASCII values. Other characters are represented by two, three, or four bytes, in accordance with a clever mathematical formula such that (1) you know instantly from the first byte of a UTF-8 character how many bytes it consists of, and (2) if you start with a byte in the middle of a UTF-8 string, you know instantly whether this is the first byte of a character and, if not, you can easily find the first byte. The UTF-8 representation of Spanish enye is C3B1, but note that this is just a way of writing bytes; it still represents (by way of the mathematical formula) the numeric value F1." In practice, how does it take place between Valentina and does realbasic? If for the moment Valentina uses only utf-16 and realbasic utf-8, it is necessary for every writing / reading to make a conversion? As before? -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: text/enriched Size: 1849 bytes Desc: not available Url : http://lists.macserve.net/pipermail/valentina-beta/attachments/20050104/cae951b4/attachment.bin From sunshine at public.kherson.ua Tue Jan 4 17:56:12 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Tue Jan 4 09:56:21 2005 Subject: Frank, Message-ID: Hi Frank, b34 must open your db in viSQL. So we can concentrate on crash in UPDATE. Please try again. Since you say problem happens on your big db, but not on small dbs, I afraid bug can be in the VarChar field. Since you make only one table UPDATE, I think you can REMOVE all other tables and left only this one to reproduce problem. Okay ? -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From lfredricks at proactive-intl.com Tue Jan 4 08:06:56 2005 From: lfredricks at proactive-intl.com (Lynn Fredricks) Date: Tue Jan 4 10:07:09 2005 Subject: valentina 2 In-Reply-To: <3C289A76-5E68-11D9-8F37-000D93C35C66@yahoo.fr> Message-ID: <20050104160657.9E63617D028@spatula.dreamhost.com> In Matt's quote below, he is talking about storage as it applies to Western languages -- as Ruslan said, its better for storage. Many western developers do not appreciate the value of UTF-16 because the value is outside their cultural understanding. UTF-8 does everything most western developers want, and much of what "eastern" developers want -- but it isnt really good enough if you need accurate and reliable sorting of Asian character data. Valentina 2 is truely an international product. REALbasic is mostly there, but could be better. REALbasic's internationalization is quite good. Unfortunately, making REALbasic as international as Valentina wont happen for quite some time as there are many priorities for REALbasic which impact all markets where REALbasic is sold. Best regards, Lynn Fredricks President Proactive International, LLC - Because it is about who you know.(tm) _____ From: valentina-beta-bounces@lists.macserve.net [mailto:valentina-beta-bounces@lists.macserve.net] On Behalf Of Olivier Sent: Tuesday, January 04, 2005 7:50 AM To: Valentina Beta Subject: Re: valentina 2 Le 4 janv. 05, ? 15:42, Ruslan Zasukhin a ?crit : With UTF8 work only software products that do not have time or power to switch to UTF16. Strange. In Matt Neuburg's article in " Realbasic developer " N?2.2, page 24, Matt presents us the various Unicode (32,16 and 8). The UTF8 is clearly presented as being the best: "UTF-8 is still more complicated, but it?s also very ingeniously designed to be compact and easily machine-parsable. In contrast to UTF-16, where every string is at least twice as long as its ASCII counterpart, in UTF-8 the 128 ASCII values appear as themselves, a single byte, so that they take up no extra space; ASCII strings are completely interchangeable with UTF-8 strings consisting of just ASCII values. Other characters are represented by two, three, or four bytes, in accordance with a clever mathematical formula such that (1) you know instantly from the first byte of a UTF-8 character how many bytes it consists of, and (2) if you start with a byte in the middle of a UTF-8 string, you know instantly whether this is the first byte of a character and, if not, you can easily find the first byte. The UTF-8 representation of Spanish enye is C3B1, but note that this is just a way of writing bytes; it still represents (by way of the mathematical formula) the numeric value F1." In practice, how does it take place between Valentina and does realbasic? If for the moment Valentina uses only utf-16 and realbasic utf-8, it is necessary for every writing / reading to make a conversion? As before? -------------- next part -------------- An HTML attachment was scrubbed... URL: http://lists.macserve.net/pipermail/valentina-beta/attachments/20050104/26d288b3/attachment.html From jda at his.com Tue Jan 4 11:28:55 2005 From: jda at his.com (jda) Date: Tue Jan 4 10:29:08 2005 Subject: speed of importing In-Reply-To: References: Message-ID: > > >I think that Valentina self set them ON on some reason. >I do not see reason why. > >Only if a field is marked unique > I have one unique field in the record. Jon From jda at his.com Tue Jan 4 11:30:14 2005 From: jda at his.com (jda) Date: Tue Jan 4 10:30:20 2005 Subject: valentina 2 In-Reply-To: References: Message-ID: >For storage, Jon. But not for algorithms. > >I can ensure you that REALBasic send UTF8 strings to OS X. >OS X convert them to UTF16, do required operation and return result. > Yes, I agree. For storage. I have written my own Pascal routines to handle unicode in OS X, and everything must be converted to/from UTF-16. But the overhead is small, and the savings in storage space for me is enormous, so I prefer UTF-8. Jon From sunshine at public.kherson.ua Tue Jan 4 18:57:05 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Tue Jan 4 10:57:12 2005 Subject: valentina 2 In-Reply-To: <3C289A76-5E68-11D9-8F37-000D93C35C66@yahoo.fr> Message-ID: On 1/4/05 5:49 PM, "Olivier" wrote: > In practice, how does it take place between Valentina and does realbasic? If > for the moment Valentina uses only utf-16 and realbasic utf-8, it is necessary > for every writing / reading to make a conversion? As before? No, Oliver, In 2.0 you do not do nothing manually. REALbasic and Valentina talk between self -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From sunshine at public.kherson.ua Tue Jan 4 18:57:42 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Tue Jan 4 10:57:49 2005 Subject: speed of importing In-Reply-To: Message-ID: On 1/4/05 6:28 PM, "jda" wrote: >> I think that Valentina self set them ON on some reason. >> I do not see reason why. >> >> Only if a field is marked unique >> > > I have one unique field in the record. Closer to finish, I will wish to profile your import -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From frank-list3 at mindstarprods.com Tue Jan 4 12:07:09 2005 From: frank-list3 at mindstarprods.com (Frank Schima) Date: Tue Jan 4 13:07:09 2005 Subject: doc Vcursor Importtext + Addrecord In-Reply-To: <83EFCDF0-5E2D-11D9-8F37-000D93C35C66@yahoo.fr> References: <83EFCDF0-5E2D-11D9-8F37-000D93C35C66@yahoo.fr> Message-ID: Hi Olivier, I think it would be much more appropriate to submit these documentation errors into the Mantis Bug Reporter. Thanks! Frank Schima On Jan 4, 2005, at 1:49 AM, Olivier wrote: > in V4RB reference, page 110, "Vcursor", the description of method > "Import_Text" indicate the old "ImportText" on the place of new > "Import_Text". > > Same problem in the example. > > In the description of the parameters, the new parameter > "inHasColumnHeader" is not put in brackets to indicate that it is > optional while the others parameters (delimiters) are now, them, > indispensable. > > page 107, the description of the method "addRecord" does not indicate > of what consists the returned integer. > > in the example, it is old "add" what is used on the place of > "addrecord". > > _______________________________________________ > Valentina-beta mailing list > Valentina-beta@lists.macserve.net > http://lists.macserve.net/mailman/listinfo/valentina-beta > From sunshine at public.kherson.ua Tue Jan 4 21:12:16 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Tue Jan 4 13:12:22 2005 Subject: doc Vcursor Importtext + Addrecord In-Reply-To: Message-ID: On 1/4/05 9:07 PM, "Frank Schima" wrote: > Hi Olivier, > > > I think it would be much more appropriate to submit these documentation > errors into the Mantis Bug Reporter. > This is true. I have move some of them to Mantis. V4RB project (and rest) have sections "Docs". Btw, Oliver, thank you very much for this docs notes! -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From frank-list3 at mindstarprods.com Tue Jan 4 13:10:17 2005 From: frank-list3 at mindstarprods.com (Frank Schima) Date: Tue Jan 4 14:10:07 2005 Subject: Open Database and Update Problems In-Reply-To: References: Message-ID: Hi Ruslan, On Jan 4, 2005, at 8:56 AM, Ruslan Zasukhin wrote: > b34 must open your db in viSQL. Argh. It still crashes for me. I've tried 32 kB segment size and 8 kB. I'm sure I recompiled viSQL with b34 too. Do you still have my sample project? The database it creates crashes for me too. I'm not sure what I'm doing wrong. To be clear, my project (as well as the stripped down sample project that I mailed you) creates a brand new database with b34 which crashes in ViSQL b34. And this is without importing any data into it. > So we can concentrate on crash in UPDATE. > Please try again. > > Since you say problem happens on your big db, but not on small dbs, > I afraid bug can be in the VarChar field. > > Since you make only one table UPDATE, > I think you can REMOVE all other tables and left only this one to > reproduce > problem. > > Okay ? Maybe, but that one table still has sensitive data. Plus I have many object pointers in it. I will keep working on this and see if that one table isolates the problem though. Also, I tried 3 other methods for Updating the table. All 3 eventually crash. 1. All API way using a Set, .FindNulls and .UpdateAllRecords. Crash occurs at .UpdateAllRecords. 2. SQL and API way. SELECT into a cursor and .UpdateAllRecords in the cursor. Crash occurs at .UpdateAllRecords. 3. Loop way. Using a Set, SetIterator and .UpdateRecord. Eventually crashes after an unknown number of iterations greater than 3 (I stepped through the code only that far :^)). Maybe the Varchar field does eventually get corrupted and that causes the crash. Shall I create a bug report on this problem? Best regards, Frank From sunshine at public.kherson.ua Tue Jan 4 22:20:31 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Tue Jan 4 14:20:37 2005 Subject: Open Database and Update Problems In-Reply-To: Message-ID: On 1/4/05 10:10 PM, "Frank Schima" wrote: > Hi Ruslan, > > > On Jan 4, 2005, at 8:56 AM, Ruslan Zasukhin wrote: > >> b34 must open your db in viSQL. > > Argh. It still crashes for me. I've tried 32 kB segment size and 8 kB. > I'm sure I recompiled viSQL with b34 too. Do you still have my sample > project? The database it creates crashes for me too. I'm not sure what > I'm doing wrong. > > To be clear, my project (as well as the stripped down sample project > that I mailed you) creates a brand new database with b34 which crashes > in ViSQL b34. And this is without importing any data into it. Strange. Let me check again. -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From sunshine at public.kherson.ua Tue Jan 4 22:22:14 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Tue Jan 4 14:22:20 2005 Subject: Open Database and Update Problems In-Reply-To: Message-ID: On 1/4/05 10:10 PM, "Frank Schima" wrote: >> Okay ? > > Maybe, but that one table still has sensitive data. Plus I have many > object pointers in it. I will keep working on this and see if that one > table isolates the problem though. > > Also, I tried 3 other methods for Updating the table. All 3 eventually > crash. This make me think about VarChar or Cache bugs. Bad. May be it is possible left only one column ?! > 1. All API way using a Set, .FindNulls and .UpdateAllRecords. Crash > occurs at .UpdateAllRecords. > 2. SQL and API way. SELECT into a cursor and .UpdateAllRecords in the > cursor. Crash occurs at .UpdateAllRecords. > 3. Loop way. Using a Set, SetIterator and .UpdateRecord. Eventually > crashes after an unknown number of iterations greater than 3 (I stepped > through the code only that far :^)). > > Maybe the Varchar field does eventually get corrupted and that causes > the crash. I also think about this. Try SQL command DIAGNOSE DATABASE Not sure if it is enabled. > Shall I create a bug report on this problem? yes -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From sunshine at public.kherson.ua Tue Jan 4 22:26:45 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Tue Jan 4 14:26:48 2005 Subject: Update Problems In-Reply-To: Message-ID: On 1/4/05 10:10 PM, "Frank Schima" wrote: >> Okay ? > > Maybe, but that one table still has sensitive data. Plus I have many > object pointers in it. I will keep working on this and see if that one > table isolates the problem though. > > Also, I tried 3 other methods for Updating the table. All 3 eventually > crash. Frank, May be it is possible to make one table with one VarChar field, In 1.x you EXPORT only this one column Then import it and now try do UPDATE. ? In this way we get as simple case as possible -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From vidal_olivier at yahoo.fr Wed Jan 5 08:43:36 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Wed Jan 5 01:44:41 2005 Subject: doc Vcursor Importtext + Addrecord In-Reply-To: References: <83EFCDF0-5E2D-11D9-8F37-000D93C35C66@yahoo.fr> Message-ID: <8214089E-5EED-11D9-8B82-000D93C35C66@yahoo.fr> ah ok ! Le 4 janv. 05, ? 20:07, Frank Schima a ?crit : > Hi Olivier, > > > I think it would be much more appropriate to submit these > documentation errors into the Mantis Bug Reporter. > > > > Thanks! > Frank Schima > > On Jan 4, 2005, at 1:49 AM, Olivier wrote: > >> in V4RB reference, page 110, "Vcursor", the description of method >> "Import_Text" indicate the old "ImportText" on the place of new >> "Import_Text". >> >> Same problem in the example. >> >> In the description of the parameters, the new parameter >> "inHasColumnHeader" is not put in brackets to indicate that it is >> optional while the others parameters (delimiters) are now, them, >> indispensable. >> >> page 107, the description of the method "addRecord" does not indicate >> of what consists the returned integer. >> >> in the example, it is old "add" what is used on the place of >> "addrecord". >> >> _______________________________________________ >> Valentina-beta mailing list >> Valentina-beta@lists.macserve.net >> http://lists.macserve.net/mailman/listinfo/valentina-beta >> > > _______________________________________________ > Valentina-beta mailing list > Valentina-beta@lists.macserve.net > http://lists.macserve.net/mailman/listinfo/valentina-beta > From vidal_olivier at yahoo.fr Wed Jan 5 10:58:27 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Wed Jan 5 03:59:37 2005 Subject: vException Message-ID: <5879F482-5F00-11D9-8B82-000D93C35C66@yahoo.fr> Hi Ruslan and list, To try Valentina 2, I begin to clean the code of my application which I had begun with Valentina 1. When I launch my application I have Vexception which I do not manage to catch. This Vexception occurs in app.open, later debuglevel, init and a first creation of DB (" ctrlPref.base.create (nf, 1, 32 *1024) "), free from problems. But later, in the creation of the second DB (" ctrlDico.base.create (nf2, 1, 32 *1024) "), me obtains a "vException". At the end of app.open I have: Exception exc have runtimeException Msgbox str (exc. ErrorNumber) + " " +exc. Message Exception exc have vException Msgbox str (exc. ErrorNumber) + " " +exc. Message Nevertheless, I have a crash and I does not obtain the msgbox. It would be practical to have the numbers of exception in the log. Olivier From vidal_olivier at yahoo.fr Wed Jan 5 11:33:13 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Wed Jan 5 04:34:10 2005 Subject: V4RB- vException, constructor of vTable Message-ID: <34163B14-5F05-11D9-8B82-000D93C35C66@yahoo.fr> I have a vException (not catched) in my creation of DB because in the constructor of one of the tables, I had "stringField.isUnique=true" placed later "stringField.isIndexed=true". If I place " isUnique " before " isIndexed ", I have no exception. If the field is " isUnique ", an index is automatically built, true? We have thus not need in that case to put the field " isIndexed " I suppose? But in Valentina 1, that did not activate error. Olivier From frost at public.kherson.ua Wed Jan 5 05:46:51 2005 From: frost at public.kherson.ua (frost@public.kherson.ua) Date: Wed Jan 5 05:47:38 2005 Subject: test Message-ID: <20050105114651.323622A9323@edison.macserve.net> just test From sunshine at public.kherson.ua Wed Jan 5 13:58:01 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Wed Jan 5 05:58:07 2005 Subject: test Message-ID: -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From sunshine at public.kherson.ua Wed Jan 5 13:58:29 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Wed Jan 5 05:58:33 2005 Subject: vException In-Reply-To: <5879F482-5F00-11D9-8B82-000D93C35C66@yahoo.fr> Message-ID: On 1/5/05 11:58 AM, "Olivier" wrote: > Hi Ruslan and list, > > To try Valentina 2, I begin to clean the code of my application which I > had begun with Valentina 1. > > When I launch my application I have Vexception which I do not manage to > catch. > This Vexception occurs in app.open, later debuglevel, init and a first > creation of DB (" ctrlPref.base.create (nf, 1, 32 *1024) "), free from > problems. Valentina .Init MUST BE the first -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From sunshine at public.kherson.ua Wed Jan 5 13:58:34 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Wed Jan 5 05:58:40 2005 Subject: V4RB- vException, constructor of vTable In-Reply-To: <34163B14-5F05-11D9-8B82-000D93C35C66@yahoo.fr> Message-ID: On 1/5/05 12:33 PM, "Olivier" wrote: > I have a vException (not catched) in my creation of DB because in the > constructor of one of the tables, I had "stringField.isUnique=true" > placed later "stringField.isIndexed=true". > If I place " isUnique " before " isIndexed ", I have no exception. > > If the field is " isUnique ", an index is automatically built, true? > We have thus not need in that case to put the field " isIndexed " I > suppose? But in Valentina 1, that did not activate error. This is not expected. We must check this. What exactly error was thrown ? -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From vidal_olivier at yahoo.fr Wed Jan 5 13:53:21 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Wed Jan 5 06:54:28 2005 Subject: vException In-Reply-To: References: Message-ID: Le 5 janv. 05, ? 12:58, Ruslan Zasukhin a ?crit : > On 1/5/05 11:58 AM, "Olivier" wrote: > >> Hi Ruslan and list, >> >> To try Valentina 2, I begin to clean the code of my application which >> I >> had begun with Valentina 1. >> >> When I launch my application I have Vexception which I do not manage >> to >> catch. >> This Vexception occurs in app.open, later debuglevel, init and a first >> creation of DB (" ctrlPref.base.create (nf, 1, 32 *1024) "), free from >> problems. > > Valentina .Init > > MUST BE the first > ah ok. but it's change nothing. From vidal_olivier at yahoo.fr Wed Jan 5 13:58:29 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Wed Jan 5 06:59:23 2005 Subject: V4RB- vException, constructor of vTable In-Reply-To: References: Message-ID: <7F1D64E7-5F19-11D9-8B82-000D93C35C66@yahoo.fr> Le 5 janv. 05, ? 12:58, Ruslan Zasukhin a ?crit : > This is not expected. > We must check this. > > What exactly error was thrown ? I do not know. The capture of vException does not work. I just have a crash and RB who indicates " vException " to the line " db.create... " olivier From vidal_olivier at yahoo.fr Wed Jan 5 14:19:32 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Wed Jan 5 07:20:29 2005 Subject: vException In-Reply-To: References: Message-ID: <7027A394-5F1C-11D9-8B82-000D93C35C66@yahoo.fr> Do the captures of vException work? In my project, that does not apparently work. The example "errors_handling" does not work either. If I click " throw exception ", the exception will not be caught but I shall have a crash. RB 5.5.4, valentina 2 version of december 18. Le 5 janv. 05, ? 13:53, Olivier a ?crit : > > Le 5 janv. 05, ? 12:58, Ruslan Zasukhin a ?crit : > >> On 1/5/05 11:58 AM, "Olivier" wrote: >> >>> Hi Ruslan and list, >>> >>> To try Valentina 2, I begin to clean the code of my application >>> which I >>> had begun with Valentina 1. >>> >>> When I launch my application I have Vexception which I do not manage >>> to >>> catch. >>> This Vexception occurs in app.open, later debuglevel, init and a >>> first >>> creation of DB (" ctrlPref.base.create (nf, 1, 32 *1024) "), free >>> from >>> problems. >> >> Valentina .Init >> >> MUST BE the first >> > ah ok. > but it's change nothing. > > _______________________________________________ > Valentina-beta mailing list > Valentina-beta@lists.macserve.net > http://lists.macserve.net/mailman/listinfo/valentina-beta > From sunshine at public.kherson.ua Wed Jan 5 15:39:12 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Wed Jan 5 07:39:17 2005 Subject: V4RB- vException, constructor of vTable In-Reply-To: <7F1D64E7-5F19-11D9-8B82-000D93C35C66@yahoo.fr> Message-ID: On 1/5/05 2:58 PM, "Olivier" wrote: > > Le 5 janv. 05, ? 12:58, Ruslan Zasukhin a ?crit : > >> This is not expected. >> We must check this. >> >> What exactly error was thrown ? > > I do not know. The capture of vException does not work. > I just have a crash and RB who indicates " vException " to the line " > db.create... " But have you wrap it with try catch ? -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From sunshine at public.kherson.ua Wed Jan 5 15:40:28 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Wed Jan 5 07:40:33 2005 Subject: vException In-Reply-To: <7027A394-5F1C-11D9-8B82-000D93C35C66@yahoo.fr> Message-ID: On 1/5/05 3:19 PM, "Olivier" wrote: > Do the captures of vException work? Yes. Oliver. First of all I think you need RUN each example that we ship. PLEASE do this. You will fiond example that show how catch exceptions, You will see how it works and that it really works > In my project, that does not apparently work. > > The example "errors_handling" does not work either. > If I click " throw exception ", the exception will not be caught but I > shall have a crash. > > RB 5.5.4, valentina 2 version of december 18. Hmm. I will check example right now. -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From sunshine at public.kherson.ua Wed Jan 5 15:42:01 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Wed Jan 5 07:42:05 2005 Subject: vException In-Reply-To: <7027A394-5F1C-11D9-8B82-000D93C35C66@yahoo.fr> Message-ID: On 1/5/05 3:19 PM, "Olivier" wrote: > Do the captures of vException work? > > In my project, that does not apparently work. > > The example "errors_handling" does not work either. > If I click " throw exception ", the exception will not be caught but I > shall have a crash. > > RB 5.5.4, valentina 2 version of december 18. 1) this example DO WORKS for now. 2) Dec 18 ? I see that latest is b34 at Jan 4 -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From vidal_olivier at yahoo.fr Wed Jan 5 17:17:06 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Wed Jan 5 10:18:11 2005 Subject: vException In-Reply-To: References: Message-ID: <3E0DF3A6-5F35-11D9-AA1C-000D93C35C66@yahoo.fr> > I am now in the version Valentina of January 4th. (vEngine and V4rb) Effectively, there are things which change. The methods vTable.getRecId and vTable.gotoRecId do not exist any more! (While they are always in doc). Why ? RB rejects them to the compilation and we do not have them any more in the predictive type. On my Mac, os X 10.3.5, RB 5.5.4, the example " errors-handling " does not still work (the same problem indicated previously). > 2) Dec 18 ? > I see that latest is b34 at Jan 4 > > > -- > Best regards, > Ruslan Zasukhin [ I feel the need...the need for speed ] > ------------------------------------------------------------- > e-mail: ruslan@paradigmasoft.com > web: http://www.paradigmasoft.com > > To subscribe to the Valentina mail list go to: > http://lists.macserve.net/mailman/listinfo/valentina > ------------------------------------------------------------- > > > _______________________________________________ > Valentina-beta mailing list > Valentina-beta@lists.macserve.net > http://lists.macserve.net/mailman/listinfo/valentina-beta > From sunshine at public.kherson.ua Wed Jan 5 18:26:48 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Wed Jan 5 10:26:54 2005 Subject: vException In-Reply-To: <3E0DF3A6-5F35-11D9-AA1C-000D93C35C66@yahoo.fr> Message-ID: On 1/5/05 6:17 PM, "Olivier" wrote: > I am now in the version Valentina of January 4th. (vEngine and V4rb) > Effectively, there are things which change. > The methods vTable.getRecId and vTable.gotoRecId do not exist any more! > (While they are always in doc). Why ? Hmm, it seems we remove them from docs already. Now instead we have single property VTable.RecID > RB rejects them to the compilation and we do not have them any more in > the predictive type. > > On my Mac, os X 10.3.5, RB 5.5.4, the example " errors-handling " does > not still work (the same problem indicated previously). How this can be ??? I run project Press step1 then step2 and I see error code in the right bottom corner. If I click checkbox and step2 again then I see Error in the LEFT bottom. Jon, Frank, anybody else. Can you also run project example Common/Error_Handling ? -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From jda at his.com Wed Jan 5 11:29:05 2005 From: jda at his.com (jda) Date: Wed Jan 5 10:29:19 2005 Subject: vException In-Reply-To: References: Message-ID: >How this can be ??? > >I run project >Press step1 then step2 and I see error code in the right bottom corner. > >If I click checkbox and step2 again then I see >Error in the LEFT bottom. > >Jon, Frank, anybody else. >Can you also run project example Common/Error_Handling ? > It works correctly for me (the exception is caught). Jon From sunshine at public.kherson.ua Wed Jan 5 18:31:23 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Wed Jan 5 10:31:34 2005 Subject: vException // example " errors-handling In-Reply-To: Message-ID: On 1/5/05 6:26 PM, "Ruslan Zasukhin" wrote: Hi Oliver, I have put plugin from b34 archive, run example. Everything works here. I also use RB 5.5.4 ---------------- Please look into folder of example, You should see file V4RB_Log This is how it looks here ??Valentina_SetDebugLevel... return Database_Ctor... return Database_Ctor1... return Database_Create... return Valentina_GetCacheSize... return (int) AllocatedSize = 8392704 Database_CreateTable... Table_Ctor... return return object = 26576172 Database_CreateTable... ERROR: 0x70506 Table name "Person" is not unique. return object = 0 -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From sunshine at public.kherson.ua Wed Jan 5 18:35:23 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Wed Jan 5 10:35:27 2005 Subject: vException In-Reply-To: Message-ID: On 1/5/05 6:29 PM, "jda" wrote: >> How this can be ??? >> >> I run project >> Press step1 then step2 and I see error code in the right bottom corner. >> >> If I click checkbox and step2 again then I see >> Error in the LEFT bottom. >> >> Jon, Frank, anybody else. >> Can you also run project example Common/Error_Handling ? >> > > It works correctly for me (the exception is caught). Oliver, then why you think that it not works on your side? What you see? You work in debugger or simply run ? RB have in menu DEBUG option Stop on Exception, But this notthing change, this is only for debugging -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From sunshine at public.kherson.ua Wed Jan 5 18:36:17 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Wed Jan 5 10:36:19 2005 Subject: Jon, index bugs fixed! Message-ID: Hi Jon, Well, both your index bugs fixed now. Have you test b34 for the first fix with page corruption ? -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From jda at his.com Wed Jan 5 11:38:26 2005 From: jda at his.com (jda) Date: Wed Jan 5 10:38:32 2005 Subject: Jon, index bugs fixed! In-Reply-To: References: Message-ID: >Well, both your index bugs fixed now. Thank you. > >Have you test b34 for the first fix with page corruption ? > Yes. It is fixed for me, too. Jon From vidal_olivier at yahoo.fr Wed Jan 5 17:45:19 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Wed Jan 5 10:46:14 2005 Subject: vException In-Reply-To: References: Message-ID: <2F867400-5F39-11D9-AA1C-000D93C35C66@yahoo.fr> Le 5 janv. 05, ? 17:35, Ruslan Zasukhin a ?crit : > RB have in menu DEBUG option Stop on Exception, > But this notthing change, this is only for debugging aahh...What an idiot!....really really sorry.. it's that option... it's ok now... From valentina-list at vermontsoftworks.com Wed Jan 5 12:24:54 2005 From: valentina-list at vermontsoftworks.com (Erik Mueller-Harder) Date: Wed Jan 5 11:25:11 2005 Subject: [V4RB] VTable.GoToRecID() In-Reply-To: References: Message-ID: Hi, Ruslan -- On Dec 31, 2004, at 5:33, Ruslan Zasukhin wrote: > V4RB - 2.0 b33 > ============== > - 0000238: [Change] Vtable.GoToRecId() and GetRecID() made as > property .RecID (Ruslan Zasukhin) I'm just getting back to work after traveling and hosting guests over our various holidays. Does this change mean that the table method for going to a particular RecID should now be: i = myTargetRecID if not myTable.RecID(i) then // report error end if If so, it's not working properly for me. I get a "This is not an array but you are using it as one" error in Rb. I hope you have a good Christmas! -- Erik From sunshine at public.kherson.ua Wed Jan 5 21:53:21 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Wed Jan 5 13:53:24 2005 Subject: [V4RB] VTable.GoToRecID() In-Reply-To: Message-ID: On 1/5/05 7:24 PM, "Erik Mueller-Harder" wrote: >> V4RB - 2.0 b33 >> ============== >> - 0000238: [Change] Vtable.GoToRecId() and GetRecID() made as >> property .RecID (Ruslan Zasukhin) > > I'm just getting back to work after traveling and hosting guests over > our various holidays. > > Does this change mean that the table method for going to a particular > RecID should now be: > > i = myTargetRecID > > if not myTable.RecID(i) then > // report error > end if > > If so, it's not working properly for me. I get a "This is not an array > but you are using it as one" error in Rb. Argh, I also did think about this problem. GoToRecord() as Boolean Have return us info. Property do not return it. In fact we still can resolve this using new method Exists if myTable.RecordExists( I ) Table.RecID = i end if Ops, I see that I have not add yet to V4RB .RecordExists property. I will do this for next beta. Now okay, Erik ? > I hope you have a good Christmas! Yes, thank you. -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From valentina-list at vermontsoftworks.com Wed Jan 5 15:00:54 2005 From: valentina-list at vermontsoftworks.com (Erik Mueller-Harder) Date: Wed Jan 5 14:01:07 2005 Subject: [V4RB] VTable.GoToRecID() In-Reply-To: References: Message-ID: <82188AEC-5F54-11D9-9409-000D936D2E58@vermontsoftworks.com> On Jan 5, 2005, at 14:53, Ruslan Zasukhin wrote: > GoToRecord() as Boolean > > Have return us info. Property do not return it. > In fact we still can resolve this using new method Exists > > if myTable.RecordExists( I ) > Table.RecID = i > end if > > Ops, I see that I have not add yet to V4RB .RecordExists property. > I will do this for next beta. > > Now okay, Erik ? Yes, thank you! -- Erik From sunshine at public.kherson.ua Thu Jan 6 01:10:18 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Wed Jan 5 17:10:25 2005 Subject: [V4RB] 2.0 b35 UPLOADED Message-ID: kernel - 2.0 b35 ================ - 0000268: [SysTables] Incorrect saving to schema "staticaly" created objects. (Ivan Smahin) V4RB - 2.0 b35 ============== - 0000267: [Tables] Added VTable.RecordExists( inRecID as Integer ) as Boolean method. (Ruslan Zasukhin) - 0000243: [API] Search on indexed method flaky (Ruslan Zasukhin) -------- Frank, now your db will work in viSQL. -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- -------------- next part -------------- An HTML attachment was scrubbed... URL: http://lists.macserve.net/pipermail/valentina-beta/attachments/20050106/0251db31/attachment.html From macsforever2000 at axint.net Wed Jan 5 18:54:17 2005 From: macsforever2000 at axint.net (Frank Schima) Date: Wed Jan 5 19:54:06 2005 Subject: [V4RB] 2.0 b35 UPLOADED In-Reply-To: References: Message-ID: Hi Ruslan, On Jan 5, 2005, at 4:10 PM, Ruslan Zasukhin wrote: > Frank, now your db will work in viSQL. Nope. Still crashes for me. Please try my sample project again, it can replicate the issue still. Best regards, Frank Schima From jda at his.com Wed Jan 5 22:15:02 2005 From: jda at his.com (jda) Date: Wed Jan 5 21:15:12 2005 Subject: [V4RB] 2.0 b35 UPLOADED In-Reply-To: References: Message-ID: > >- 0000243: [API] Search on indexed method flaky (Ruslan Zasukhin) > Hi Ruslan, I'm afraid I'm still seeing bizarre behavior searching the method field. I'll upload another set of steps to take on Mantis and send you a file to reproduce. Jon From vidal_olivier at yahoo.fr Thu Jan 6 05:18:45 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Wed Jan 5 22:19:00 2005 Subject: [V4RB] VTable.GoToRecID() In-Reply-To: References: Message-ID: <0E741050-5F9A-11D9-AA1C-000D93C35C66@yahoo.fr> Le 5 janv. 05, ? 18:24, Erik Mueller-Harder a ?crit : > Hi, Ruslan -- > > On Dec 31, 2004, at 5:33, Ruslan Zasukhin wrote: > >> V4RB - 2.0 b33 >> ============== >> - 0000238: [Change] Vtable.GoToRecId() and GetRecID() made as >> property .RecID (Ruslan Zasukhin) > > I'm just getting back to work after traveling and hosting guests over > our various holidays. > > Does this change mean that the table method for going to a particular > RecID should now be: > > i = myTargetRecID > > if not myTable.RecID(i) then > if not mytable.recId=i then ? > // report error > end if > > If so, it's not working properly for me. I get a "This is not an > array but you are using it as one" error in Rb. > > I hope you have a good Christmas! > > -- Erik > > _______________________________________________ > Valentina-beta mailing list > Valentina-beta@lists.macserve.net > http://lists.macserve.net/mailman/listinfo/valentina-beta > From vidal_olivier at yahoo.fr Thu Jan 6 06:20:52 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Wed Jan 5 23:20:46 2005 Subject: vCursor.import_text Message-ID: Hi Ruslan and list, "Cursor.import_text" work? I have a crash in its use. Cur.import_Text (f, "\t ", " \r") // =crash Cur.import_Text ( f ) // =not accepted Cur.import_Text (f, "\t ", " \r", false) // =not accepted I saw in the mantis that Franck had indicated (ID 185 of December 12th, assigned to Igor, not resolved) a crash on import_text when there were several fields. (but it's not in cursor ?) But I tried with a single field and five recordings and I have the same crash. dim cur As vcursor dim f As folderItem f=getOpenFolderItem("text/plain") ctrlDico.base.cpville.DeleteAllRecords ctrlDico.base.cpVille.flush cur=new vcursor(ctrlDico.base,"SELECT insee,ville,xpays,ypays FROM cpville WHERE recid=0",2,3,2) // ---- OK cur.import_Text(f,"\t","\r") // ----------------------- CRASH ----------------------- ctrlDico.base.cpVille.flush cur=nil exception exc as vexception msgBox exc.Message From vidal_olivier at yahoo.fr Thu Jan 6 06:23:11 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Wed Jan 5 23:23:00 2005 Subject: [V4RB] VTable.GoToRecID() In-Reply-To: <0E741050-5F9A-11D9-AA1C-000D93C35C66@yahoo.fr> References: <0E741050-5F9A-11D9-AA1C-000D93C35C66@yahoo.fr> Message-ID: <0EF77ED0-5FA3-11D9-AA1C-000D93C35C66@yahoo.fr> Le 6 janv. 05, ? 05:18, Olivier a ?crit : > if not mytable.recId=i then ? Sorry, I had had no answer of Ruslan when I made this answer. From macsforever2000 at axint.net Wed Jan 5 23:13:40 2005 From: macsforever2000 at axint.net (Frank Schima) Date: Thu Jan 6 00:13:25 2005 Subject: vCursor.import_text In-Reply-To: References: Message-ID: <1C012856-5FAA-11D9-B97F-0003939246BC@axint.net> Hi Olivier, On Jan 5, 2005, at 10:20 PM, Olivier wrote: > Hi Ruslan and list, > > "Cursor.import_text" work? Here is what I use to import with V 2: theImportCursor.Import_Text(theImportFile, chr(9), chr(13)) > I have a crash in its use. > > Cur.import_Text (f, "\t ", " \r") // =crash > Cur.import_Text ( f ) // =not accepted > Cur.import_Text (f, "\t ", " \r", false) // =not accepted > > I saw in the mantis that Franck had indicated (ID 185 of December > 12th, assigned to Igor, not resolved) a crash on import_text when > there were several fields. (but it's not in cursor ?) > But I tried with a single field and five recordings and I have the > same crash. My bug report is about how picky V 2 importing is compared to V 1. In V 2, the number of fields in the text file must match *exactly* with the number of fields in the SELECT of the cursor. If they do not match in *any* single line of text, it will crash. V 1 was much more forgiving about filling in Null or blank values if there were less text fields than SELECT fields in the cursor - or ignoring the extra text fields if there were too many. So check your import file to make sure the number of fields of every line matches with the SQL statement. BBedit with "Show Invisibles" set ON is very helpful here. Best regards, Frank Schima From sunshine at public.kherson.ua Thu Jan 6 10:59:31 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Thu Jan 6 02:59:41 2005 Subject: doc Vcursor Importtext + Addrecord In-Reply-To: <83EFCDF0-5E2D-11D9-8F37-000D93C35C66@yahoo.fr> Message-ID: On 1/4/05 10:49 AM, "Olivier" wrote: > in V4RB reference, page 110, "Vcursor", the description of method > "Import_Text" indicate the old "ImportText" on the place of new > "Import_Text". I have change not docs but V4RB. So now will be ImportText > Same problem in the example. > > In the description of the parameters, the new parameter > "inHasColumnHeader" is not put in brackets to indicate that it is > optional while the others parameters (delimiters) are now, them, > indispensable. > > page 107, the description of the method "addRecord" does not indicate > of what consists the returned integer. > > in the example, it is old "add" what is used on the place of > "addrecord". -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From vidal_olivier at yahoo.fr Thu Jan 6 10:15:09 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Thu Jan 6 03:15:07 2005 Subject: vCursor.import_text In-Reply-To: <1C012856-5FAA-11D9-B97F-0003939246BC@axint.net> References: <1C012856-5FAA-11D9-B97F-0003939246BC@axint.net> Message-ID: <7686390C-5FC3-11D9-AA1C-000D93C35C66@yahoo.fr> Hi Frank, thank you for your answer. > theImportCursor.Import_Text(theImportFile, chr(9), chr(13)) I tried it, that does not work. But I wonder if the problem is not for the creation of my DBs. Because on three DBs whom I create by launching my application, only the most small and the simplest are then legible on vStudio 1.0a75 and ViSQL 2.0 (version of January 3rd). Two others more complicated crash as soon as I open them with vStudio or ViSQL. Nevertheless I have no crash in my application and the DBs is create bugfree. >> I have a crash in its use. >> >> Cur.import_Text (f, "\t ", " \r") // =crash >> Cur.import_Text ( f ) // =not accepted >> Cur.import_Text (f, "\t ", " \r", false) // =not accepted > > My bug report is about how picky V 2 importing is compared to V 1. In > V 2, the number of fields in the text file must match *exactly* with > the number of fields in the SELECT of the cursor. If they do not match > in *any* single line of text, it will crash. > > V 1 was much more forgiving about filling in Null or blank values if > there were less text fields than SELECT fields in the cursor - or > ignoring the extra text fields if there were too many. > ah ok. What means that if a field in a SINGLE line is not filled, there is a crash? not good... > So check your import file to make sure the number of fields of every > line matches with the SQL statement. BBedit with "Show Invisibles" set > ON is very helpful here. > > > Best regards, > Frank Schima > > _______________________________________________ > Valentina-beta mailing list > Valentina-beta@lists.macserve.net > http://lists.macserve.net/mailman/listinfo/valentina-beta > From sunshine at public.kherson.ua Thu Jan 6 11:47:48 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Thu Jan 6 03:47:53 2005 Subject: vCursor.import_text In-Reply-To: <7686390C-5FC3-11D9-AA1C-000D93C35C66@yahoo.fr> Message-ID: On 1/6/05 11:15 AM, "Olivier" wrote: Igor, try now to fix bug reported by Frank > Hi Frank, > > thank you for your answer. > > >> theImportCursor.Import_Text(theImportFile, chr(9), chr(13)) > > I tried it, that does not work. > > But I wonder if the problem is not for the creation of my DBs. > Because on three DBs whom I create by launching my application, only > the most small and the simplest are then legible on vStudio 1.0a75 and > ViSQL 2.0 (version of January 3rd). Two others more complicated crash > as soon as I open them with vStudio or ViSQL. Nevertheless I have no > crash in my application and the DBs is create bugfree. -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From vidal_olivier at yahoo.fr Thu Jan 6 10:48:15 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Thu Jan 6 03:48:09 2005 Subject: SQL request in vCursor : "or" Message-ID: <16558F1E-5FC8-11D9-AA1C-000D93C35C66@yahoo.fr> Hi, Why this query it works more in Valentina 2 ? t="string" cur=new vcursor(ctrlAdres.base,"SELECT nom FROM personne WHERE corbeille=false and (rep1="+t+" or rep2="+t+" or rep3="+t+" or rep4="+t+")",2,2,2) If I remove "OR", it works : cur=new vcursor(ctrlAdres.base,"SELECT nom FROM personne WHERE corbeille=false and (rep1="+t+")",2,2,2) thank you olivier From sunshine at public.kherson.ua Thu Jan 6 11:56:01 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Thu Jan 6 03:56:12 2005 Subject: SQL request in vCursor : "or" In-Reply-To: <16558F1E-5FC8-11D9-AA1C-000D93C35C66@yahoo.fr> Message-ID: On 1/6/05 11:48 AM, "Olivier" wrote: Hi Oliver, 1) I think Valentina turn SOME error code and SOME error string, yes? Please always check them. > Why this query it works more in Valentina 2 ? > > t="string" > cur=new vcursor(ctrlAdres.base,"SELECT nom FROM personne WHERE > corbeille=false and (rep1="+t+" or rep2="+t+" or rep3="+t+" or > rep4="+t+")",2,2,2) It will be more useful to see the SQL String which you get BTW, you can EASY see SQL string you build in the V4RB_Log file. So as far as I see you get ctrlAdres.base,"SELECT nom FROM personne WHERE corbeille=false and (rep1 = string or rep2 = string .... I see mistake. You have forget single quotes around 'string' > If I remove "OR", it works : > cur=new vcursor(ctrlAdres.base,"SELECT nom FROM personne WHERE > corbeille=false and (rep1="+t+")",2,2,2) -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From vidal_olivier at yahoo.fr Thu Jan 6 11:06:42 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Thu Jan 6 04:06:31 2005 Subject: SQL request in vCursor : "or" In-Reply-To: References: Message-ID: > > It will be more useful to see the SQL String which you get > BTW, you can EASY see SQL string you build in the V4RB_Log file. > LOG : Cursor_Ctor... PARAM: instance = 27059408 PARAM: inDB = 26180540 PARAM: inSQLstr = "SELECT nom FROM personne WHERE corbeille=false and (rep1=1 or rep2=1 or rep3=1 or rep4=1)" PARAM: inCursorLocation = 2 PARAM: inCursorType = 2 PARAM: inDirection = 2 Rep1, Rep2, Rep3, Rep4 are VBYTE. > So as far as I see you get > > ctrlAdres.base,"SELECT nom FROM personne WHERE > corbeille=false and (rep1 = string or rep2 = string .... > > I see mistake. You have forget single quotes around 'string' > Rep1.. are vByte > >> If I remove "OR", it works : >> cur=new vcursor(ctrlAdres.base,"SELECT nom FROM personne WHERE >> corbeille=false and (rep1="+t+")",2,2,2) > > -- > Best regards, > Ruslan Zasukhin [ I feel the need...the need for speed ] > ------------------------------------------------------------- > e-mail: ruslan@paradigmasoft.com > web: http://www.paradigmasoft.com > > To subscribe to the Valentina mail list go to: > http://lists.macserve.net/mailman/listinfo/valentina > ------------------------------------------------------------- > > > _______________________________________________ > Valentina-beta mailing list > Valentina-beta@lists.macserve.net > http://lists.macserve.net/mailman/listinfo/valentina-beta > From sunshine at public.kherson.ua Thu Jan 6 12:14:09 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Thu Jan 6 04:14:17 2005 Subject: SQL request in vCursor : "or" In-Reply-To: Message-ID: On 1/6/05 12:06 PM, "Olivier" wrote: >> It will be more useful to see the SQL String which you get >> BTW, you can EASY see SQL string you build in the V4RB_Log file. >> > LOG : > > Cursor_Ctor... > PARAM: instance = 27059408 > PARAM: inDB = 26180540 > PARAM: inSQLstr = "SELECT nom FROM personne WHERE corbeille=false and > (rep1=1 or rep2=1 or rep3=1 or rep4=1)" > PARAM: inCursorLocation = 2 > PARAM: inCursorType = 2 > PARAM: inDirection = 2 > > Rep1, Rep2, Rep3, Rep4 are VBYTE. Aha. So what happens? Crash ? How it notworks? Does log show any error info ? What viSQL show if do this query again this db ? If viSQL also show problem send me db and we test here -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From vidal_olivier at yahoo.fr Thu Jan 6 11:44:38 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Thu Jan 6 04:44:33 2005 Subject: SQL request in vCursor : "or" In-Reply-To: References: Message-ID: Le 6 janv. 05, ? 11:14, Ruslan Zasukhin a ?crit : > On 1/6/05 12:06 PM, "Olivier" wrote: > >>> It will be more useful to see the SQL String which you get >>> BTW, you can EASY see SQL string you build in the V4RB_Log file. >>> >> LOG : >> >> Cursor_Ctor... >> PARAM: instance = 27059408 >> PARAM: inDB = 26180540 >> PARAM: inSQLstr = "SELECT nom FROM personne WHERE corbeille=false and >> (rep1=1 or rep2=1 or rep3=1 or rep4=1)" >> PARAM: inCursorLocation = 2 >> PARAM: inCursorType = 2 >> PARAM: inDirection = 2 >> >> Rep1, Rep2, Rep3, Rep4 are VBYTE. > > Aha. > > So what happens? Crash ? yes > How it notworks? > > Does log show any error info ? no also the exception not caught. > What viSQL show if do this query again this db ? > If viSQL also show problem send me db and we test here Impossible to open the DB in ViSQL or VStudio. There is a crash from the opening of the DB. the DB is empty. I send you log and DB. Olivier > > From vidal_olivier at yahoo.fr Thu Jan 6 14:08:14 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Thu Jan 6 07:08:19 2005 Subject: SQL request in vCursor : "or" In-Reply-To: <16558F1E-5FC8-11D9-AA1C-000D93C35C66@yahoo.fr> References: <16558F1E-5FC8-11D9-AA1C-000D93C35C66@yahoo.fr> Message-ID: <0663ABD7-5FE4-11D9-AA1C-000D93C35C66@yahoo.fr> Ruslan, Apparently, there is a crash because there is no records. (If I add records before the query, there are no crashes) But it is strange because this query (without "OR" nor "AND") does not provoke crash : > cur=new vcursor(ctrlAdres.base,"SELECT nom FROM personne WHERE > corbeille=false and (rep1="+t+")",2,2,2) Olivier Le 6 janv. 05, ? 10:48, Olivier a ?crit : > Hi, > > Why this query it works more in Valentina 2 ? > > t="string" > cur=new vcursor(ctrlAdres.base,"SELECT nom FROM personne WHERE > corbeille=false and (rep1="+t+" or rep2="+t+" or rep3="+t+" or > rep4="+t+")",2,2,2) > > If I remove "OR", it works : > cur=new vcursor(ctrlAdres.base,"SELECT nom FROM personne WHERE > corbeille=false and (rep1="+t+")",2,2,2) > > thank you > > olivier > > _______________________________________________ > Valentina-beta mailing list > Valentina-beta@lists.macserve.net > http://lists.macserve.net/mailman/listinfo/valentina-beta > From sunshine at public.kherson.ua Thu Jan 6 15:12:44 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Thu Jan 6 07:13:01 2005 Subject: SQL request in vCursor : "or" In-Reply-To: <0663ABD7-5FE4-11D9-AA1C-000D93C35C66@yahoo.fr> Message-ID: On 1/6/05 3:08 PM, "Olivier" wrote: > Ruslan, > > Apparently, there is a crash because there is no records. > (If I add records before the query, there are no crashes) Aha! So crash on empty table?! This should be easy to reproduce. Ivan, please check. > But it is strange because this query (without "OR" nor "AND") does not > provoke crash : >> cur=new vcursor(ctrlAdres.base,"SELECT nom FROM personne WHERE >> corbeille=false and (rep1="+t+")",2,2,2) > > Olivier -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From vidal_olivier at yahoo.fr Thu Jan 6 14:15:46 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Thu Jan 6 07:15:35 2005 Subject: SQL request in vCursor : "or" In-Reply-To: References: Message-ID: <13890612-5FE5-11D9-AA1C-000D93C35C66@yahoo.fr> Le 6 janv. 05, ? 14:12, Ruslan Zasukhin a ?crit : > On 1/6/05 3:08 PM, "Olivier" wrote: > >> Ruslan, >> >> Apparently, there is a crash because there is no records. >> (If I add records before the query, there are no crashes) > > Aha! So crash on empty table?! > yes > This should be easy to reproduce. > Ivan, please check. > > >> But it is strange because this query (without "OR" nor "AND") does not >> provoke crash : >>> cur=new vcursor(ctrlAdres.base,"SELECT nom FROM personne WHERE >>> corbeille=false and (rep1="+t+")",2,2,2) >> >> Olivier > > -- > Best regards, > Ruslan Zasukhin [ I feel the need...the need for speed ] > ------------------------------------------------------------- > e-mail: ruslan@paradigmasoft.com > web: http://www.paradigmasoft.com > > To subscribe to the Valentina mail list go to: > http://lists.macserve.net/mailman/listinfo/valentina > ------------------------------------------------------------- > > > _______________________________________________ > Valentina-beta mailing list > Valentina-beta@lists.macserve.net > http://lists.macserve.net/mailman/listinfo/valentina-beta > From sunshine at public.kherson.ua Thu Jan 6 15:19:29 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Thu Jan 6 07:19:33 2005 Subject: SQL request in vCursor : "or" In-Reply-To: <13890612-5FE5-11D9-AA1C-000D93C35C66@yahoo.fr> Message-ID: On 1/6/05 3:15 PM, "Olivier" wrote: > > Le 6 janv. 05, ? 14:12, Ruslan Zasukhin a ?crit : > >> On 1/6/05 3:08 PM, "Olivier" wrote: >> >>> Ruslan, >>> >>> Apparently, there is a crash because there is no records. >>> (If I add records before the query, there are no crashes) >> >> Aha! So crash on empty table?! >> > yes Ivan have reproduce! -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From vidal_olivier at yahoo.fr Thu Jan 6 14:22:07 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Thu Jan 6 07:21:57 2005 Subject: SQL request in vCursor : "or" + ViSQL In-Reply-To: References: Message-ID: Le 6 janv. 05, ? 11:44, Olivier a ?crit : > Impossible to open the DB in ViSQL or VStudio. > There is a crash from the opening of the DB. it's strange : ViSQL and vStudio crashes from the opening of my DB. Ruslan, Can you open it ? With the new Vengine and V4RB (yesterday), there is numerous methods in VISQL to correct. ( Cursor.currentPosition, link.leftType, vField.GetMethod.........) From vidal_olivier at yahoo.fr Thu Jan 6 14:22:35 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Thu Jan 6 07:22:22 2005 Subject: SQL request in vCursor : "or" In-Reply-To: References: Message-ID: <07501F7F-5FE6-11D9-AA1C-000D93C35C66@yahoo.fr> > > Ivan have reproduce! > ok ! From sunshine at public.kherson.ua Thu Jan 6 15:26:36 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Thu Jan 6 07:26:41 2005 Subject: SQL request in vCursor : "or" + ViSQL In-Reply-To: Message-ID: On 1/6/05 3:22 PM, "Olivier" wrote: >> Impossible to open the DB in ViSQL or VStudio. >> There is a crash from the opening of the DB. > > it's strange : ViSQL and vStudio crashes from the opening of my DB. > Ruslan, Can you open it ? I have not try yet Do you use Classes ? Db of Frank also cannot be opened in viSQL for now. I think the same bug. I hope today I will work on it. > With the new Vengine and V4RB (yesterday), there is numerous methods in > VISQL to correct. > ( Cursor.currentPosition, link.leftType, vField.GetMethod.........) Yes. Wait. I have correct viSQL. It works for me in latest beta... -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From vidal_olivier at yahoo.fr Thu Jan 6 14:30:31 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Thu Jan 6 07:30:23 2005 Subject: SQL request in vCursor : "or" + ViSQL In-Reply-To: References: Message-ID: <234A7A20-5FE7-11D9-AA1C-000D93C35C66@yahoo.fr> > > Do you use Classes ? > Yes > Db of Frank also cannot be opened in viSQL for now. > I think the same bug. > > I hope today I will work on it. > ok, perfect ! From vidal_olivier at yahoo.fr Thu Jan 6 16:50:21 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Thu Jan 6 09:50:23 2005 Subject: Cursor.addrecord Message-ID: Has what corresponds the integer returned by addrecord ? From sunshine at public.kherson.ua Thu Jan 6 17:54:55 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Thu Jan 6 09:55:01 2005 Subject: Cursor.addrecord In-Reply-To: Message-ID: On 1/6/05 5:50 PM, "Olivier" wrote: > Has what corresponds the integer returned by addrecord ? Table.AddRecord() returns RecId of new record. Cursor.AddRecord() Hmm, not sure that it returns useful info. probably position of new record in the cursor, which is LAST always. In most cases you can ignore it. To ignore use syntax call curs.AddRecord() -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From vidal_olivier at yahoo.fr Fri Jan 7 08:52:46 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Fri Jan 7 01:53:17 2005 Subject: Exceptions and old error handling Message-ID: <1ED906E1-6081-11D9-B974-000D93C35C66@yahoo.fr> Hi list, Sorry for this problem which has to seem to you very simple, but I shall like knowing how you would use the Exceptions in the following case. I open a cursor which contains one field (String). This String is modified by RB methods and recorded again in the DB. If the update fails (unique field: doubloon), the recording is deleted. With the old error handling, it is simple, tests it is made IN the loop. With the management of the exceptions, what is the most elegant solution? dim a,i as integer dim s1 as string cur=new vcursor(DB,"SELECT string1 FROM Table1",2,3,2) a=cur.recordCount i=1 if a>0 then do until i=a+1 cur.position=i s1=Method1(cur.stringField("string1").value) if len(s1)>2 then cur.stringField("string1").value=s1 cur.updaterecord //rs=cur.update() //if rs=false then //cur.deleteRecord() //i=i-1 //a=cur.recordCount //end if end if i=i+1 loop end if thank you olivier -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: text/enriched Size: 1144 bytes Desc: not available Url : http://lists.macserve.net/pipermail/valentina-beta/attachments/20050107/323c28c4/attachment.bin From vidal_olivier at yahoo.fr Fri Jan 7 09:01:50 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Fri Jan 7 02:01:42 2005 Subject: Exceptions and old error handling In-Reply-To: <1ED906E1-6081-11D9-B974-000D93C35C66@yahoo.fr> References: <1ED906E1-6081-11D9-B974-000D93C35C66@yahoo.fr> Message-ID: <630D31CC-6082-11D9-B974-000D93C35C66@yahoo.fr> I think that I have to appeal in my loop to an independent method to be able to process the update and possibly the error. But that makes me change many methods in my application. I go to see that. olivier Le 7 janv. 05, ? 08:52, Olivier a ?crit : > Hi list, > > Sorry for this problem which has to seem to you very simple, > but I shall like knowing how you would use the Exceptions in the > following case. > > I open a cursor which contains one field (String). > This String is modified by RB methods and recorded again in the DB. > If the update fails (unique field: doubloon), the recording is deleted. > With the old error handling, it is simple, tests it is made IN the > loop. > With the management of the exceptions, what is the most elegant > solution? > > > dim a,i as integer > dim s1 as string > cur=new vcursor(DB,"SELECT string1 FROM Table1",2,3,2) > a=cur.recordCount > i=1 > > if a>0 then > > do until i=a+1 > cur.position=i > s1=Method1(cur.stringField("string1").value) > > if len(s1)>2 then > cur.stringField("string1").value=s1 > cur.updaterecord //rs=cur.update() > //if rs=false then > //cur.deleteRecord() > //i=i-1 > //a=cur.recordCount > //end if > end if > > i=i+1 > loop > > end if > > thank you > > olivier_______________________________________________ > Valentina-beta mailing list > Valentina-beta@lists.macserve.net > http://lists.macserve.net/mailman/listinfo/valentina-beta -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: text/enriched Size: 1590 bytes Desc: not available Url : http://lists.macserve.net/pipermail/valentina-beta/attachments/20050107/c5010667/attachment.bin From jda at his.com Fri Jan 7 07:42:51 2005 From: jda at his.com (jda) Date: Fri Jan 7 06:43:00 2005 Subject: Exceptions and old error handling In-Reply-To: <1ED906E1-6081-11D9-B974-000D93C35C66@yahoo.fr> References: <1ED906E1-6081-11D9-B974-000D93C35C66@yahoo.fr> Message-ID: >I open a cursor which contains one field (String). >This String is modified by RB methods and recorded again in the DB. >If the update fails (unique field: doubloon), the recording is deleted. >With the old error handling, it is simple, tests it is made IN the loop. >With the management of the exceptions, what is the most elegant solution? > > If you are using RB, use a try/catch block. Here's is some pseudocode for adding a record with a unique integer field: x = generateAUniqueID do try myCursor.LongField("uniqueID").SetString(str(x)) cursPosition = myCursor.addRecord success = true //this only occurs if the addRecord was successful catch theError as VException if theError.ErrorNumber <> EVError.ConstraintUniqueViolation then //not a duplicate uniqueID raise theError end if x = generateAUniqueID end try loop until success Jon From vidal_olivier at yahoo.fr Fri Jan 7 14:11:31 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Fri Jan 7 07:11:32 2005 Subject: Exceptions and old error handling In-Reply-To: References: <1ED906E1-6081-11D9-B974-000D93C35C66@yahoo.fr> Message-ID: perfect, thank you very much Jon ! olivier Le 7 janv. 05, ? 13:42, jda a ?crit : > If you are using RB, use a try/catch block. Here's is some pseudocode > for adding a record with a unique integer field: > > x = generateAUniqueID > do > try > myCursor.LongField("uniqueID").SetString(str(x)) > cursPosition = myCursor.addRecord > success = true //this only occurs if the addRecord was successful > catch theError as VException > if theError.ErrorNumber <> EVError.ConstraintUniqueViolation > then //not a duplicate uniqueID > raise theError > end if > x = generateAUniqueID > end try > loop until success From vidal_olivier at yahoo.fr Sat Jan 8 07:47:42 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Sat Jan 8 00:47:36 2005 Subject: valentina 2 In-Reply-To: References: Message-ID: <322A36BB-6141-11D9-9A27-000D93C35C66@yahoo.fr> Hi Ruslan and list, > Right now no. I see that 2.0 now is 2 times slower on addRecord. > Don't know about rest operations. > But we yet in debug mode for indexes for example. > > I think that AFTER we spend a month or so on optimzation and tunning, > We will make 2.0 faster of 1.x. Effectively, addRecord is MUCH slower. Far too much. Example : AddRecord of records of 4 fields (3 strings, 1 boolean). The 3 strings (lenght 38) are indexed. one unique. one indexedByWord. Valentina 1 : cache 20 Mo. Valentina 2 : cache 50 Mo. RB 5.5.4 Add x records Valentina 1 Valentina 2 Add 100 2 secondes 4 secondes Add 600 2 sec 11 sec Add 3000 3 sec 50 sec Add 5000 3 sec 86 sec Add 8000 4 sec 147 sec Add 11000 5 sec 212 sec Add 17000 6 sec 355 sec Add 25000 8 sec 547 sec Add 36000 10 sec 875 sec Even if three strings are NOT indexed and unique, on the addrecord of 8000 lines, Valentina 2 puts 114 seconds. The gain is modest. Valentina 2 is UTF16 but in Valentina 1 EVERY string is TRANSFORM of UTF8 in ASCII-FRENCH by a RB METHOD. Are you SURE that you can optimize and accelerate significantly Valentina 2.0 ? That you will reach at least the same speed as Valentina 1? Because the distance is really very important. thank you olivier From sunshine at public.kherson.ua Sat Jan 8 11:58:07 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Sat Jan 8 03:58:31 2005 Subject: Linux support In-Reply-To: <41DF1084.6020107@digi-net.com> Message-ID: On 1/8/05 12:43 AM, "Ken Jordan" wrote: Hi Ken, I will answer self in few hours, For now I resend your letter to Valentina list and Valentina beta list. Ken, are you subscribed to these lists ? If now, please subscribe. Here you can get many help. Also if you just start develop new project I recommend you take a look on Valentina 2.0 betas. > Ruslan, > > Thank you for the reply - great to know that Linux will also be > supported soon. > > Based on that, we have decided to go ahead with testing Valentina to see > if it will meet our needs. I am using the evaluation version of > Valentina for REALbasic 1.11, REALbasic 5 format. OS is Win XP. So far > it seems to work very well, but I'm experiencing a couple of problems > that I haven't been able to figure out - maybe you can help? > > 1) I am trying to use Valentina to store e-mail messages (broken down > into parts). One of the fields stored is the Message-ID - a String that > can be from 1 to 70 characters. I'm using VVarChar to store the string - > created in the BaseObject subclass using > > Dim MessageID As VVarChar > ... > MessageID = new VVarChar("message_id", 70, "ASCII", kV_Indexed + kV_Unique) > > I create a VCursor using > cursor = db.SQLselect("SELECT message_id, (...) FROM inbox") and display > the results. When one of the messages is chosen I create another VCursor > using > cursor.CurrentPosition = (selected record) > cursor2 = db.SQLselect("SELECT (...) FROM inbox WHERE message_id = " + > cursor.VarCharField("message_id").Value) > > Sometimes I get the record requested, sometimes a null return. It is > very consistent which will work and which will not. Investigation shows: > > these id strings will not be found - > "<41C313E7.60203@dummy.com>" > "" > "<110C77E416.290691699@dummy.com>" > "" > > these id strings will be found - > "<41931397.60203@dummy.com>" > "<90328470.yy3nizzrvfze@a31fa80>" > "<1108776416-290691699@dummy.com>" > "<110C77E416-290691699@dummy.com>" > "<80039468203@dummy.com>" > > It appears that if any non-digits appear before the '.' or before the > '@' if no '.' is present, the string (obtained FROM the db) is not found > in the db, but if only digits are there then the string is found. It > doesn't seem to matter if there are all non-digits before a '-'. > > The same thing happens if VString is used instead of VVarChar. > > Is this some strange behavior caused by the way Valentina handles > strings, or is there something that I'm missing? > > 2) I can't seem to get data back from a VText. I've tried both > VTextObject.WriteRawData("string") and VTextObject.Value = "string" - in > both cases the data is in the VText object in the BaseObject before the > AddRecord() call and table.Flush() call. However, neither > cursor2.TextField("text_field").ReadRawData() nor > cursor2.TextField("text_field").Value return the data - just an empty > string. What might I be missing? > > Thank you for any help you can provide, > > Ken Jordan -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From sunshine at public.kherson.ua Sat Jan 8 12:00:25 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Sat Jan 8 04:00:35 2005 Subject: V4MD 2.0 beta expected to be introduced today. Message-ID: Hi All Director Developers. I expect that today we will introduce V4MD 2.0 beta for MAC. And for Windows I think. So if you are interested, please subscribe to Valentina beta list >From our support page. -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From vidal_olivier at yahoo.fr Sat Jan 8 11:17:10 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Sat Jan 8 04:17:05 2005 Subject: open DB Message-ID: <7563F628-615E-11D9-9A27-000D93C35C66@yahoo.fr> Hi Ruslan and list, I you indicated that I did not manage to open my DB with ViSQL and Vstudio. I cannot also open it with my application ! Nevertheless I create it free from problems. In my application, I add records, launch SQL queries. That works. I quit cleanly my application without crash. But then impossible to open my DB with my application ("DB.open(nf)" ) or viSQL or Vstudio. My DB is created with classes. Olivier RB 5.5.4 Mac os X 10.3.5 Valentina 2, january 6th version From sunshine at public.kherson.ua Sat Jan 8 14:27:38 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Sat Jan 8 06:27:50 2005 Subject: open DB In-Reply-To: <7563F628-615E-11D9-9A27-000D93C35C66@yahoo.fr> Message-ID: On 1/8/05 12:17 PM, "Olivier" wrote: > Hi Ruslan and list, > > I you indicated that I did not manage to open my DB with ViSQL and > Vstudio. VSTUDIO is not yet ready for real use with Valentina 2.0 > I cannot also open it with my application ! I will check this today evening I think. > Nevertheless I create it free from problems. > In my application, I add records, launch SQL queries. > That works. > I quit cleanly my application without crash. > > But then impossible to open my DB with my application ("DB.open(nf)" ) > or viSQL or Vstudio. > My DB is created with classes. Yes Frank have the same problem. Although he can open db by his own app > Olivier > > RB 5.5.4 > Mac os X 10.3.5 > Valentina 2, january 6th version -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From jda at his.com Sat Jan 8 07:33:32 2005 From: jda at his.com (jda) Date: Sat Jan 8 06:33:44 2005 Subject: valentina 2 In-Reply-To: <322A36BB-6141-11D9-9A27-000D93C35C66@yahoo.fr> References: <322A36BB-6141-11D9-9A27-000D93C35C66@yahoo.fr> Message-ID: > >Effectively, addRecord is MUCH slower. Far too much. > Hi Olivier, You might want to add your comments as a note to my feature request #245 on Mantis: http://paradigma.ukrcom.kherson.ua/bt/login_page.php Jon From vidal_olivier at yahoo.fr Sat Jan 8 15:58:17 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Sat Jan 8 08:58:26 2005 Subject: valentina 2 addRecord by API In-Reply-To: <322A36BB-6141-11D9-9A27-000D93C35C66@yahoo.fr> References: <322A36BB-6141-11D9-9A27-000D93C35C66@yahoo.fr> Message-ID: I continued my tests by trying the addRecord of the API. The first time it was with a cursor. It is faster but it is slow even MUCH MORE than with Valentina 1. The tests are made with at most 36000 simple lines of only 4 small fields without links. But with a lot more lines and complex recordings... Example : AddRecord of records of 4 fields (3 strings, 1 boolean). The 3 strings (lenght 38) are indexed. one unique. one indexedByWord. Valentina 1 : cache 20 Mo. Valentina 2 : cache 50 Mo. I tried 20 or 90 Mb, it is the same thing. RB 5.5.4 Add x records Valentina 1 Valentina 2 CURSOR Valentina 2 API Add 100 2 secondes 4 secondes 3 secondes Add 600 2 sec 11 sec 7 sec Add 3000 3 sec 50 sec 26 sec Add 5000 3 sec 86 sec 46 sec Add 8000 4 sec 147 sec 83 sec Add 11000 5 sec 212 sec 119 sec Add 17000 6 sec 355 sec 211 sec Add 25000 8 sec 547 sec 329 sec Add 36000 10 sec 875 sec 542 sec Even if three strings are NOT indexed and unique, on the addrecord of 8000 lines, Valentina 2 puts 114 seconds (cursor) or 54 seconds (API). The gain is modest. Valentina 2 is UTF16 but in Valentina 1 EVERY string is TRANSFORM of UTF8 in ASCII-FRENCH by a RB METHOD before the recording in DB. Le 8 janv. 05, ? 07:47, Olivier a ?crit : > Hi Ruslan and list, > > >> Right now no. I see that 2.0 now is 2 times slower on addRecord. >> Don't know about rest operations. >> But we yet in debug mode for indexes for example. >> >> I think that AFTER we spend a month or so on optimzation and tunning, >> We will make 2.0 faster of 1.x. > > Effectively, addRecord is MUCH slower. Far too much. > > Example : > AddRecord of records of 4 fields (3 strings, 1 boolean). The 3 strings > (lenght 38) are indexed. one unique. one indexedByWord. > > Valentina 1 : cache 20 Mo. > Valentina 2 : cache 50 Mo. > RB 5.5.4 > > Add x records Valentina 1 Valentina 2 > > Add 100 2 secondes 4 secondes > Add 600 2 sec 11 sec > Add 3000 3 sec 50 sec > Add 5000 3 sec 86 sec > Add 8000 4 sec 147 sec > Add 11000 5 sec 212 sec > Add 17000 6 sec 355 sec > Add 25000 8 sec 547 sec > Add 36000 10 sec 875 sec > > Even if three strings are NOT indexed and unique, > on the addrecord of 8000 lines, Valentina 2 puts 114 seconds. The gain > is modest. > > Valentina 2 is UTF16 but in Valentina 1 EVERY string is TRANSFORM of > UTF8 in ASCII-FRENCH by a RB METHOD. > > Are you SURE that you can optimize and accelerate significantly > Valentina 2.0 ? > That you will reach at least the same speed as Valentina 1? > Because the distance is really very important. > > thank you > > olivier > > _______________________________________________ > Valentina-beta mailing list > Valentina-beta@lists.macserve.net > http://lists.macserve.net/mailman/listinfo/valentina-beta > -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: text/enriched Size: 3379 bytes Desc: not available Url : http://lists.macserve.net/pipermail/valentina-beta/attachments/20050108/7c68a47f/attachment.bin From sunshine at public.kherson.ua Sat Jan 8 17:01:45 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Sat Jan 8 09:01:51 2005 Subject: valentina 2 In-Reply-To: <322A36BB-6141-11D9-9A27-000D93C35C66@yahoo.fr> Message-ID: On 1/8/05 8:47 AM, "Olivier" wrote: > Hi Ruslan and list, > > >> Right now no. I see that 2.0 now is 2 times slower on addRecord. >> Don't know about rest operations. >> But we yet in debug mode for indexes for example. >> >> I think that AFTER we spend a month or so on optimzation and tunning, >> We will make 2.0 faster of 1.x. > > Effectively, addRecord is MUCH slower. Far too much. > > Example : > AddRecord of records of 4 fields (3 strings, 1 boolean). The 3 strings > (lenght 38) are indexed. one unique. one indexedByWord. > > Valentina 1 : cache 20 Mo. > Valentina 2 : cache 50 Mo. > RB 5.5.4 > > Add x records Valentina 1 Valentina 2 > > Add 100 2 secondes 4 secondes > Add 600 2 sec 11 sec > Add 3000 3 sec 50 sec > Add 5000 3 sec 86 sec > Add 8000 4 sec 147 sec > Add 11000 5 sec 212 sec > Add 17000 6 sec 355 sec > Add 25000 8 sec 547 sec > Add 36000 10 sec 875 sec > > Even if three strings are NOT indexed and unique, Moment. Strings marked as unique? THEN Valentina will build index ALWAYS! > on the addrecord of 8000 lines, Valentina 2 puts 114 seconds. The gain > is modest. > > Valentina 2 is UTF16 but in Valentina 1 EVERY string is TRANSFORM of > UTF8 in ASCII-FRENCH by a RB METHOD. Right. > Are you SURE that you can optimize and accelerate significantly > Valentina 2.0 ? > That you will reach at least the same speed as Valentina 1? > Because the distance is really very important. Of course. As you see we have even not linear dependency. Starts as 2 times slower. 2/4 secs finish 80 times slower 10/875 It looks to me that somewhere is ON diagnose calls. So the bigger db the longer diagnose take place. We need find this place. -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From vidal_olivier at yahoo.fr Sat Jan 8 16:16:29 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Sat Jan 8 09:16:21 2005 Subject: valentina 2 In-Reply-To: References: Message-ID: <45D4AFF0-6188-11D9-9C63-000D93C35C66@yahoo.fr> >> Even if three strings are NOT indexed and unique, > > Moment. Strings marked as unique? > THEN Valentina will build index ALWAYS! > yes. That is why I made an test also by removing the properties " indexed " of all the fields and the property "unique" of 1 field which had it. But the gain of speed is not important. > > Of course. As you see we have even not linear dependency. > Starts as 2 times slower. 2/4 secs > finish 80 times slower 10/875 > > It looks to me that somewhere is ON diagnose calls. > So the bigger db the longer diagnose take place. > We need find this place. That reassures me. My application has to process many data, that would be catastrophic for me if this problem was not resolved. I sent you another e-mail where I tested the addRecord by API, it is effectively faster than with cursors but much slower than with Valentina 1. thank you olivier From vidal_olivier at yahoo.fr Sat Jan 8 16:28:26 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Sat Jan 8 09:28:21 2005 Subject: valentina 2 In-Reply-To: References: <322A36BB-6141-11D9-9A27-000D93C35C66@yahoo.fr> Message-ID: Le 8 janv. 05, ? 13:33, jda a ?crit : >> >> Effectively, addRecord is MUCH slower. Far too much. >> > > Hi Olivier, > > You might want to add your comments as a note to my feature request > #245 on Mantis: > > http://paradigma.ukrcom.kherson.ua/bt/login_page.php OK it is made. thank you Jon olivier > Jon > _______________________________________________ > Valentina-beta mailing list > Valentina-beta@lists.macserve.net > http://lists.macserve.net/mailman/listinfo/valentina-beta > From sunshine at public.kherson.ua Sat Jan 8 18:23:24 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Sat Jan 8 10:23:29 2005 Subject: valentina 2 In-Reply-To: <45D4AFF0-6188-11D9-9C63-000D93C35C66@yahoo.fr> Message-ID: On 1/8/05 5:16 PM, "Olivier" wrote: >>> Even if three strings are NOT indexed and unique, >> >> Moment. Strings marked as unique? >> THEN Valentina will build index ALWAYS! >> > yes. > That is why I made an test also by removing the properties " indexed " > of all the fields and the property "unique" of 1 field which had it. > But the gain of speed is not important. Oliver, another point. When you do this bench you must make sure that V4RB_Log File is disabled ! So is it ? -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From jda at his.com Sat Jan 8 11:47:40 2005 From: jda at his.com (jda) Date: Sat Jan 8 10:47:51 2005 Subject: Sorting on null text fields Message-ID: Hi Ruslan, I'm seeing some odd behavior, and I want to understand if I can fix it. I have records sorted by this method: authorsSortMethod = CreateTextField("authorsSortMethod", 128, EVFlag.fNone, "left(authors, 256)") When the authors field is empty (NULL), that record seems to be randomly returned. It's sometimes at the beginning, sometimes at the end, and sometimes in the middle of the cursor. Anything I can do about this? Jon From sunshine at public.kherson.ua Sat Jan 8 18:59:26 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Sat Jan 8 10:59:31 2005 Subject: Sorting on null text fields In-Reply-To: Message-ID: On 1/8/05 6:47 PM, "jda" wrote: Hi Jon, > I'm seeing some odd behavior, and I want to understand if I can fix it. > > I have records sorted by this method: > > authorsSortMethod = CreateTextField("authorsSortMethod", 128, > EVFlag.fNone, "left(authors, 256)") ^^^^^^^^ try fNullable > When the authors field is empty (NULL), that record seems to be Empty is not the same to NULL. > randomly returned. It's sometimes at the beginning, sometimes at the > end, and sometimes in the middle of the cursor. > > Anything I can do about this? If it is really NULL (such records must be found by WHERE fld IS NULL) Then they must go TOP in the sort order. -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From macsforever2000 at axint.net Sat Jan 8 10:02:44 2005 From: macsforever2000 at axint.net (Frank Schima) Date: Sat Jan 8 11:02:32 2005 Subject: open DB In-Reply-To: References: Message-ID: <1D536004-6197-11D9-A9E0-0003939246BC@axint.net> Hi guys, On Jan 8, 2005, at 5:27 AM, Ruslan Zasukhin wrote: > On 1/8/05 12:17 PM, "Olivier" wrote: > >> But then impossible to open my DB with my application ("DB.open(nf)" ) >> or viSQL or Vstudio. >> My DB is created with classes. > > Yes Frank have the same problem. > Although he can open db by his own app Actually, I have not tried to open it with my application yet. Best regards, Frank Schima From sunshine at public.kherson.ua Sat Jan 8 19:07:30 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Sat Jan 8 11:07:35 2005 Subject: [V4RB] 2.0 b36 uploaded Message-ID: kernel - 2.0 b36 ================ - 0000271: [Other] Throw away Loki::Factory. (Ivan Smahin) - 0000270: [Sets] Crash with AND - OR on empty table. (Ivan Smahin) V4RB - 2.0 b36 ============== - 0000262: [API] Update All Records causes a crash with large table (Igor Gomon) -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- -------------- next part -------------- An HTML attachment was scrubbed... URL: http://lists.macserve.net/pipermail/valentina-beta/attachments/20050108/858bff82/attachment.html From jda at his.com Sat Jan 8 12:08:24 2005 From: jda at his.com (jda) Date: Sat Jan 8 11:08:31 2005 Subject: Sorting on null text fields In-Reply-To: References: Message-ID: > > > >> authorsSortMethod = CreateTextField("authorsSortMethod", 128, >> EVFlag.fNone, "left(authors, 256)") > ^^^^^^^^ > try fNullable > > > > >> Anything I can do about this? > >If it is really NULL > (such records must be found by WHERE fld IS NULL) > >Then they must go TOP in the sort order. > You mean I have to change the method? I added EVFlag.fNullable to the current definition, and the records still sort randomly. I also realize that I wasn't using fNullable in my regular field definitions (I am nw) and was nulling them in my code. Valentina didn't complain, but perhaps just nothing was being done. Jon From sunshine at public.kherson.ua Sat Jan 8 19:17:53 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Sat Jan 8 11:17:56 2005 Subject: [V4MD] 2.0 b36 uploaded, for MAC. START BETA TESTING. Message-ID: Hi Director developers! We are happy introduce to you V4MD 2.0 b36 for MAC. * Please download it, * install following instructions. * download docs (they are draft yet) * run and play our examples. ** V4MD 2.0 have get SIGNIFICANT architectural changes. Now it have no 2 Xtras but 10 Xtras. ** We have implement the excellent OO design for V4MD. Yes, we have found that Macromedia Director allow this. ** Also we have implement usage of properties for Xtras. So now syntax is more simple for many things. E.g. Instead of pair methods db.SetName() + db.GetName() Now we have simply db.Name property. ** we have implement 3 ways to work with pictures: - 1.x way via cast member. - way via .Picture and variable - .Image - modern way of Director. - Now we do not depend on QT on any platform. - now we support any major bit depth of picture, 32, 24, 16, 8 - now we can store picture into JPG, TIFF, PICT, DIB - Now we can load into Picture field the file from HDD as is. ** SQL92 implemented. On the beta download page you will find link to our SQL grammar. ** Now V4MD will produce V4MD_Log file with a lots of debug info, which will simplify your development. ** And a lots of other new cool features. So We invite you for ACTIVE BETA TESTING! :-) Enjoy. P.S. If somebody cannot use MAC version and need WIN version, please let us know, we will prepare WIN archive also. -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From sunshine at public.kherson.ua Sat Jan 8 19:22:47 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Sat Jan 8 11:22:50 2005 Subject: Sorting on null text fields In-Reply-To: Message-ID: On 1/8/05 7:08 PM, "jda" wrote: >>> authorsSortMethod = CreateTextField("authorsSortMethod", 128, >>> EVFlag.fNone, "left(authors, 256)") >> ^^^^^^^^ >> try fNullable >> >> >>> >>> Anything I can do about this? >> >> If it is really NULL >> (such records must be found by WHERE fld IS NULL) >> >> Then they must go TOP in the sort order. >> > > You mean I have to change the method? > > I added EVFlag.fNullable to the current definition, and the records > still sort randomly. > > I also realize that I wasn't using fNullable in my regular field > definitions (I am nw) and was nulling them in my code. Valentina > didn't complain, but perhaps just nothing was being done. So looks as bug ? Then we need db + query or project -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From jda at his.com Sat Jan 8 12:43:30 2005 From: jda at his.com (jda) Date: Sat Jan 8 11:43:37 2005 Subject: Sorting on null text fields In-Reply-To: References: Message-ID: Are blogs automatically nullable? I tried setting the EVFlag.fNullable for a blob and got an error. The latest RB docs (updated today, p. 48) are ambiguous. Here is the definition: CreateBLOBField(inName as string, inSegmentSize as integer) as VBLOB But here is the example: fldAge = tblPerson.CreateBLOBField("notesStyle", 256, EVFlags.fNullable + EVFlags.fIndexed ) Jon From snw at paradise.net.nz Sun Jan 9 07:58:29 2005 From: snw at paradise.net.nz (Sean Wilson) Date: Sat Jan 8 12:59:47 2005 Subject: [V4MD] 2.0 b36 uploaded, for MAC. START BETA TESTING. In-Reply-To: References: Message-ID: <6.2.0.14.2.20050109075747.030a1610@pop3.paradise.net.nz> >P.S. If somebody cannot use MAC version and need WIN version, > please let us know, we will prepare WIN archive also. That would be me - could you upload a Windows version? Thanks, -Sean. From sunshine at public.kherson.ua Sat Jan 8 21:41:39 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Sat Jan 8 13:41:47 2005 Subject: [V4MD] 2.0 b36 uploaded, for MAC. START BETA TESTING. In-Reply-To: <6.2.0.14.2.20050109075747.030a1610@pop3.paradise.net.nz> Message-ID: On 1/8/05 8:58 PM, "Sean Wilson" wrote: > >> P.S. If somebody cannot use MAC version and need WIN version, >> please let us know, we will prepare WIN archive also. > > That would be me - could you upload a Windows version? Okay, Igor, please prepare archives for windows. -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From sunshine at public.kherson.ua Sat Jan 8 22:45:49 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Sat Jan 8 14:45:59 2005 Subject: Problem of viSQL open failure Message-ID: Frank, Oliver, I have found why viSQL fails. This is when it try to show structure of db. IT try show props of Links of db. So it come to some ObjectPtr, try cast it to Link2 and ops. Exception. Because I did not make ObjectPtr class as child of Link2. I cannot do this becaise ObjectPtr already is child of Vfield. The correct way is to use Interfaces, but they are broken in REALbasic, ARGH! Show stopper... I will think for now. -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From sunshine at public.kherson.ua Sat Jan 8 23:19:27 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Sat Jan 8 15:19:34 2005 Subject: FIXED: Problem of viSQL open failure In-Reply-To: Message-ID: On 1/8/05 10:45 PM, "Ruslan Zasukhin" wrote: Hi All, Well, I just disable few lines in the viSQL so now it works. I have re-uplaod b36 with the corrected viSQL. > Frank, Oliver, > > I have found why viSQL fails. > > This is when it try to show structure of db. > IT try show props of Links of db. > > So it come to some ObjectPtr, try cast it to Link2 and ops. Exception. > Because I did not make ObjectPtr class as child of Link2. > I cannot do this becaise ObjectPtr already is child of Vfield. > > The correct way is to use Interfaces, but they are broken in REALbasic, > ARGH! > > Show stopper... > > I will think for now. > -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From vbt at paradigma.ukrcom.kherson.ua Sat Jan 8 23:31:46 2005 From: vbt at paradigma.ukrcom.kherson.ua (vbt@paradigma.ukrcom.kherson.ua) Date: Sat Jan 8 15:35:19 2005 Subject: [V4RB 0000274]: added Enum EVDateFormat Message-ID: The following issue has been SUBMITTED. ====================================================================== ====================================================================== Reported By: Ruslan Zasukhin Assigned To: Ruslan Zasukhin ====================================================================== Project: V4RB Issue ID: 274 Category: API Reproducibility: always Severity: minor Priority: normal Status: assigned ====================================================================== Date Submitted: 01-08-2005 23:31 EET Last Modified: 01-08-2005 23:31 EET ====================================================================== Summary: added Enum EVDateFormat Description: kMDY = 0, kDMY = 1, kYMD = 2 ====================================================================== Issue History Date Modified Username Field Change ====================================================================== 01-08-05 23:31 Ruslan ZasukhinNew Issue ====================================================================== From sunshine at public.kherson.ua Sun Jan 9 01:09:18 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Sat Jan 8 17:09:24 2005 Subject: Jon, Message-ID: Jon, Please download b36. Download it after this letter if you have made this before. archive must be dated Jan 9 And check again index bug that you did report. Let me know -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From sunshine at public.kherson.ua Sun Jan 9 01:37:55 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Sat Jan 8 17:38:02 2005 Subject: Oliver, bench Message-ID: Hi Oliver, I have open Example/Example 1.x/ single TAble And run it with b36. This example adds into table with 16 fields 100,000 records. On my G5 DUAL 2.0 it have made this in 10 seconds. Please check this example on your side. -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From jda at his.com Sun Jan 9 07:55:19 2005 From: jda at his.com (jda) Date: Sun Jan 9 06:55:30 2005 Subject: Sorting on null text fields In-Reply-To: References: Message-ID: Hi Ruslan, I didn't see an answer to this, so I'm posting again. Thanks, Jon. >Are blogs automatically made nullable? I tried setting the >EVFlag.fNullable for a blob and got an error. The latest RB docs >(updated today, p. 48) are ambiguous. > >Here is the definition: > >CreateBLOBField(inName as string, inSegmentSize as integer) as VBLOB > >But here is the example: > >fldAge = tblPerson.CreateBLOBField("notesStyle", 256, >EVFlags.fNullable + EVFlags.fIndexed ) From sunshine at public.kherson.ua Sun Jan 9 19:20:04 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Sun Jan 9 11:20:19 2005 Subject: Sorting on null text fields In-Reply-To: Message-ID: On 1/9/05 2:55 PM, "jda" wrote: >> EVFlag.fNullable for a blob and got an error. What error ? >> The latest RB docs >> (updated today, p. 48) are ambiguous. >> >> Here is the definition: >> >> CreateBLOBField(inName as string, inSegmentSize as integer) as VBLOB >> >> But here is the example: >> >> fldAge = tblPerson.CreateBLOBField("notesStyle", 256, >> EVFlags.fNullable + EVFlags.fIndexed ) Example is wrong as I see. -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From jda at his.com Sun Jan 9 12:27:39 2005 From: jda at his.com (jda) Date: Sun Jan 9 11:27:51 2005 Subject: Sorting on null text fields In-Reply-To: References: Message-ID: >On 1/9/05 2:55 PM, "jda" wrote: > >>> EVFlag.fNullable for a blob and got an error. > >What error ? > Too many parameters. Jon From sunshine at public.kherson.ua Sun Jan 9 19:47:35 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Sun Jan 9 11:47:42 2005 Subject: Sorting on null text fields In-Reply-To: Message-ID: On 1/9/05 7:27 PM, "jda" wrote: >> On 1/9/05 2:55 PM, "jda" wrote: >> >>>> EVFlag.fNullable for a blob and got an error. >> >> What error ? >> > > Too many parameters. Aha, so compilation error. Yes, BLOB constructor do not have flags parameter, Because no sense index it. Yes, BLOB field is nullable always. So even no need set flag. NOTE that in the TEXT field again we have flags, because we can index this field, index by words it and so on. -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From sunshine at public.kherson.ua Sun Jan 9 20:30:26 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Sun Jan 9 12:30:34 2005 Subject: V4RB, Vlink.AsVObjectPtr, Vlink.AsVBinaryLink, Message-ID: Hi All, Taking into account that in V4RB 2.0 I still cannot use full power of REALbasic interfaces, because they are broken and/or not ready for normal usage, I have add the following cast functions Vlink.AsVObjectPtr() as VObjectPtr Vlink.AsVBinaryLink() as VBinaryLink VObjectPtr.AsVLink() as VLink They are effective, fast, so we can do type casting between these classes if needed. This is temporary solution until we get normal interfaces in the REALbasic. It can take up to 1-2 years. Usually this will need only few developers. Just I add this to have full picture of API. And I self have to use them in the viSQL. -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From jda at his.com Sun Jan 9 14:14:52 2005 From: jda at his.com (jda) Date: Sun Jan 9 13:15:03 2005 Subject: Sorting on null text fields In-Reply-To: References: Message-ID: >Aha, so compilation error. > >Yes, BLOB constructor do not have flags parameter, >Because no sense index it. > >Yes, BLOB field is nullable always. So even no need set flag. > > >NOTE that in the TEXT field again we have flags, because we can index this >field, index by words it and so on. > Yes, I saw. I added the fNullable flag to my text fields a few days ago. FWIW, before this when I opened my db in viSQL the fields were listed as NULL anyway. So maybe the nullable flag isn't necessary in any case (but I set it just to make sure). Jon From sunshine at public.kherson.ua Sun Jan 9 21:18:16 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Sun Jan 9 13:18:20 2005 Subject: Sorting on null text fields In-Reply-To: Message-ID: On 1/9/05 9:14 PM, "jda" wrote: >> Aha, so compilation error. >> >> Yes, BLOB constructor do not have flags parameter, >> Because no sense index it. >> >> Yes, BLOB field is nullable always. So even no need set flag. >> >> >> NOTE that in the TEXT field again we have flags, because we can index this >> field, index by words it and so on. >> > > Yes, I saw. I added the fNullable flag to my text fields a few days > ago. FWIW, before this when I opened my db in viSQL the fields were > listed as NULL anyway. So maybe the nullable flag isn't necessary in > any case (but I set it just to make sure). Yes, BLOB will be marked as Nullable anyway. So this is optional. Really. -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From vidal_olivier at yahoo.fr Mon Jan 10 05:52:51 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Sun Jan 9 22:52:48 2005 Subject: valentina 2 In-Reply-To: References: Message-ID: <7B705410-62C3-11D9-A06F-000D93C35C66@yahoo.fr> > > Oliver, another point. > > When you do this bench you must make sure that > > V4RB_Log File is disabled ! > > So is it ? > Arrghh....no. I do new bench now. olivier From vidal_olivier at yahoo.fr Mon Jan 10 05:56:07 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Sun Jan 9 22:55:59 2005 Subject: Oliver, bench In-Reply-To: References: Message-ID: > > I have open Example/Example 1.x/ single TAble > > And run it with b36. > > This example adds into table with 16 fields 100,000 records. > On my G5 DUAL 2.0 it have made this in 10 seconds. > > Please check this example on your side. > on my eMac 1.25Ghz : 9/10 seconds also ! From sunshine at public.kherson.ua Mon Jan 10 08:16:16 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Mon Jan 10 00:16:23 2005 Subject: Oliver, bench In-Reply-To: Message-ID: On 1/10/05 6:56 AM, "Olivier" wrote: >> >> I have open Example/Example 1.x/ single TAble >> >> And run it with b36. >> >> This example adds into table with 16 fields 100,000 records. >> On my G5 DUAL 2.0 it have made this in 10 seconds. >> >> Please check this example on your side. >> > > on my eMac 1.25Ghz : 9/10 seconds also ! Ok, can you send me YOUR project where you see slow Add -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From vidal_olivier at yahoo.fr Mon Jan 10 08:19:45 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Mon Jan 10 01:19:43 2005 Subject: valentina 2 In-Reply-To: <7B705410-62C3-11D9-A06F-000D93C35C66@yahoo.fr> References: <7B705410-62C3-11D9-A06F-000D93C35C66@yahoo.fr> Message-ID: <012B3E5D-62D8-11D9-BDBB-000D93C35C66@yahoo.fr> New tests : Stopwatches indicated below are valid ONLY as comparison between Valentina 1 and 2. They does not need to take them theoretically. They do not calculate the RAW speed of Valentina 1 or 2. Because there are several factors which SLOW DOWN Adds (display of a progressbar, several methods of processing of strings before "addrecord", strings got back by a listBox, etc.). BUT they are the SAME factors for Valentina 1 and Valentina 2. Valentina 1 cache : 20 Mo Valentina 2 cache : 50 Mo DebugLevel disabled. 3 string fields. 1 boolean field. 2 strings are indexed. 1 string is UNIQUE. Add x lines With cursor With API Valentina 1 100 lines 2 s 2 s 2 s 600 2 2 2 3000 9 8 3 5000 17 16 3 8000 35 34 4 17000 109 6 25000 180 8 36000 329 322 10 There the cursor is pratically also fast as the API. ------------------- NO unique or indexed field : Add x lines With cursor With API Valentina 1 3000 lines 3 s 3 s 8000 6 5 17000 13 10 4 36000 32 22 7 The difference of speed between Valentina 1 and 2 is sharply less important here. Besides, the speed of Valentina seems proportional among adds. There is thus a problem in the processing of the indexs / unique by Valentina 2. Here, API is a little faster than the cursor. olivier Le 10 janv. 05, ? 05:52, Olivier a ?crit : >> >> Oliver, another point. >> >> When you do this bench you must make sure that >> >> V4RB_Log File is disabled ! >> >> So is it ? >> > Arrghh....no. > > I do new bench now. > > olivier > > _______________________________________________ > Valentina-beta mailing list > Valentina-beta@lists.macserve.net > http://lists.macserve.net/mailman/listinfo/valentina-beta > -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: text/enriched Size: 1843 bytes Desc: not available Url : http://lists.macserve.net/pipermail/valentina-beta/attachments/20050110/cd9ff58a/attachment.bin From sunshine at public.kherson.ua Mon Jan 10 09:29:10 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Mon Jan 10 01:29:17 2005 Subject: valentina 2 In-Reply-To: <012B3E5D-62D8-11D9-BDBB-000D93C35C66@yahoo.fr> Message-ID: On 1/10/05 9:19 AM, "Olivier" wrote: > New tests : > > Stopwatches indicated below are valid ONLY as comparison between Valentina 1 > and 2. > They does not need to take them theoretically. They do not calculate the RAW > speed of Valentina 1 or 2. > Because there are several factors which SLOW DOWN Adds (display of a > progressbar, several methods of processing of strings before "addrecord", > strings got back by a listBox, etc.). > > BUT they are the SAME factors for Valentina 1 and Valentina 2. > > Valentina 1 cache : 20 Mo > Valentina 2 cache : 50 Mo > DebugLevel disabled. > > 3 string fields. 1 boolean field. 2 strings are indexed. 1 string is UNIQUE. > > Add x lines With cursor With API Valentina 1 > > 100 lines 2 s 2 s 2 s > 600 2 2 2 > 3000 9 8 3 > 5000 17 16 3 > 8000 35 34 4 > 17000 109 6 > 25000 180 8 > 36000 329 322 10 > > There the cursor is pratically also fast as the API. Oliver, remind please first column? Second ? Third ? -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From vidal_olivier at yahoo.fr Mon Jan 10 08:36:00 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Mon Jan 10 01:35:52 2005 Subject: valentina 2 In-Reply-To: References: Message-ID: <46797286-62DA-11D9-BDBB-000D93C35C66@yahoo.fr> >> >> Add x lines With cursor With API Valentina 1 >> >> 100 lines 2 s 2 s 2 s >> 600 2 2 2 >> 3000 9 8 3 >> 5000 17 16 3 >> 8000 35 34 4 >> 17000 109 6 >> 25000 180 8 >> 36000 329 322 10 >> >> There the cursor is pratically also fast as the API. > > Oliver, remind please > > first column? Second ? Third ? > first : cursor second : API third : valentina 1 From vidal_olivier at yahoo.fr Mon Jan 10 08:48:58 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Mon Jan 10 01:48:52 2005 Subject: valentina 2 In-Reply-To: <012B3E5D-62D8-11D9-BDBB-000D93C35C66@yahoo.fr> References: <7B705410-62C3-11D9-A06F-000D93C35C66@yahoo.fr> <012B3E5D-62D8-11D9-BDBB-000D93C35C66@yahoo.fr> Message-ID: <16265C7D-62DC-11D9-BDBB-000D93C35C66@yahoo.fr> I redid a test by keeping indexes but by removing "unique" of the field string which had it. (I replaced it by " indexed "). NOW it's PROPORTIONAL : 8000 lines : 11 seconds 17000 lines : 22 s 25000 lines : 35 s The processing of the "unique" is thus not proportional on Valentina 2. precisions : DB created with classes. The "flush" is included in the stopwatch. PS: I cannot connect a lot in internet today Le 10 janv. 05, ? 08:19, Olivier a ?crit : > NO unique or indexed field : > > Add x lines With cursor With API Valentina 1 > > 3000 lines 3 s 3 s > 8000 6 5 > 17000 13 10 4 > 36000 32 22 7 > > The difference of speed between Valentina 1 and 2 is sharply less > important here. > Besides, the speed of Valentina seems proportional among adds. > There is thus a problem in the processing of the indexs / unique by > Valentina 2. From vidal_olivier at yahoo.fr Mon Jan 10 08:57:18 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Mon Jan 10 01:57:13 2005 Subject: FIXED: Problem of viSQL open failure In-Reply-To: References: Message-ID: <3FE3FE4C-62DD-11D9-BDBB-000D93C35C66@yahoo.fr> Le 8 janv. 05, ? 22:19, Ruslan Zasukhin a ?crit : > On 1/8/05 10:45 PM, "Ruslan Zasukhin" > wrote: > > Hi All, > > Well, I just disable few lines in the viSQL so now it works. > > I have re-uplaod b36 with the corrected viSQL. > Now I can open my DB with VISQL AND MY APP ! good work ! olivier > _______________ > Valentina-beta mailing list > Valentina-beta@lists.macserve.net > http://lists.macserve.net/mailman/listinfo/valentina-beta > From bibiko at eva.mpg.de Mon Jan 10 09:18:23 2005 From: bibiko at eva.mpg.de (Hans-Joerg Bibiko) Date: Mon Jan 10 02:18:50 2005 Subject: [V4MD] 2.0 b36 uploaded, for MAC. START BETA TESTING. In-Reply-To: References: Message-ID: <41E23A4F.3090800@eva.mpg.de> HI Ruslan, please, can you send me the URL for downloading the V4MD 2.0 b36 for MAC. Thanks Hans From sunshine at public.kherson.ua Mon Jan 10 10:32:46 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Mon Jan 10 02:32:58 2005 Subject: [V4MD] 2.0 b36 uploaded, for MAC. START BETA TESTING. In-Reply-To: <41E23A4F.3090800@eva.mpg.de> Message-ID: On 1/10/05 10:18 AM, "Hans-Joerg Bibiko" wrote: > HI Ruslan, > > > please, can you send me the URL for downloading the V4MD 2.0 b36 for MAC. Hi Hans, http://paradigmasoft.com/download/v2_beta/download.html -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From bibiko at eva.mpg.de Mon Jan 10 10:42:40 2005 From: bibiko at eva.mpg.de (Hans-Joerg Bibiko) Date: Mon Jan 10 03:42:52 2005 Subject: [V4MD] 2.0 b36 uploaded, for MAC. START BETA TESTING. In-Reply-To: References: Message-ID: <41E24E10.6040405@eva.mpg.de> HI Ruslan, well, this is the URL I knew but there is no way for me to reach it. www.paradigmasoft.com is also unavaible. Is you server down? Hans From sunshine at public.kherson.ua Mon Jan 10 12:20:38 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Mon Jan 10 04:20:41 2005 Subject: [V4MD] 2.0 b36 uploaded, for MAC. START BETA TESTING. In-Reply-To: <41E24E10.6040405@eva.mpg.de> Message-ID: On 1/10/05 11:42 AM, "Hans-Joerg Bibiko" wrote: > HI Ruslan, > > well, this is the URL I knew but there is no way for me to reach it. > > www.paradigmasoft.com is also unavaible. Is you server down? Seems to be down at least last 30 minutes -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From sunshine at public.kherson.ua Mon Jan 10 12:45:32 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Mon Jan 10 04:45:38 2005 Subject: FIXED: Sorting on null text fields In-Reply-To: Message-ID: On 1/9/05 2:55 PM, "jda" wrote: > >> Are blogs automatically made nullable? I tried setting the >> EVFlag.fNullable for a blob and got an error. The latest RB docs >> (updated today, p. 48) are ambiguous. >> >> Here is the definition: >> >> CreateBLOBField(inName as string, inSegmentSize as integer) as VBLOB >> >> But here is the example: >> >> fldAge = tblPerson.CreateBLOBField("notesStyle", 256, >> EVFlags.fNullable + EVFlags.fIndexed ) -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From ernestogiannotta at tiscalinet.it Mon Jan 10 11:46:58 2005 From: ernestogiannotta at tiscalinet.it (erne) Date: Mon Jan 10 04:47:29 2005 Subject: V4RB1.11 export empty bug Message-ID: Hello Ruslan, I've found that if try to export an empty cursor to a text file Vale will crash. Not a big deal since no sense in exporting an empty cursor, yet annoying since we must add check in code before export to avoid crash. may I send you a demo project? Cool Runnings, Erne. |er| musical box |ne| a media store From sunshine at public.kherson.ua Mon Jan 10 12:59:25 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Mon Jan 10 04:59:29 2005 Subject: V4RB1.11 export empty bug In-Reply-To: Message-ID: On 1/10/05 12:46 PM, "erne" wrote: > Hello Ruslan, > > I've found that if try to export an empty cursor to a text file Vale > will crash. > Not a big deal since no sense in exporting an empty cursor, > yet annoying since we must add check in code before export to avoid > crash. > > may I send you a demo project? I think this bug already is fixed -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From vidal_olivier at yahoo.fr Mon Jan 10 12:39:13 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Mon Jan 10 05:39:11 2005 Subject: first cursor Message-ID: <401F5323-62FC-11D9-BDBB-000D93C35C66@yahoo.fr> Hi list, A strange thing with my app happens to me. I have a DB with 83 000 small records. When I click a menu of my app, a window appears with an Einhugur' Datagrid (sort of fast listBox) fills 83 000 records. Thus, when the window opens, a cursor is created to obtain 83 000 records of the DB. With Valentina 1, the first time when I open this window, the datagrid puts approximately 1 second to fill. If I close the window and what I open it again, the datagrid fills almost immediately. With Valentina 2, the first time when I open this window, the datagrid puts ......34 seconds to fill ! If I close the window and what I open it again, the datagrid puts approximately 1 second to fill. As the cursor is destroyed every time I close the window, I thought that the problem was linked to the cache. But if I puts the cache on 50 Mb, 1Mb or even 0.01 Mb, that changes nothing! (Same on Valentina 1) Nevertheless the DB makes 15 Mb with Valentina 1 and 37 Mb with Valentina 2 ( UTF16). I also tried my DB with ViSQL for Valentina 1 and ViSQL for Valentina 2. If I launch a query to select all the records of my table (83 000) : With ViSQL 1: - The first time: 0.13 seconds - If I begin again: 0.06 seconds on average With ViSQL 2: - The first time: 0.28 seconds - If I begin again: 0.22 seconds on average Olivier From vidal_olivier at yahoo.fr Mon Jan 10 12:46:26 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Mon Jan 10 05:46:19 2005 Subject: valentina 2 Message-ID: <42B5584A-62FD-11D9-BDBB-000D93C35C66@yahoo.fr> Ruslan, I speak only about problems, it is the purpose of this list! But I am impressed all the same by the possibilities of Valentina 2. I think that you made a very good work and that you will manage to improve this beta. The storage Unicode is very practical. ( But I worry about the size of files. My complete DB will make approximately 500 to 600 Mb with Valentina 1. With Valentina 2 I am afraid that it takes too much place (UTF16). Finally I shall see well) The API methods are very elegant and very promising. But if my application has to work later on a small local area network, I am obliged to use cursors ? thank you olivier From sunshine at public.kherson.ua Mon Jan 10 13:47:47 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Mon Jan 10 05:47:58 2005 Subject: first cursor In-Reply-To: <401F5323-62FC-11D9-BDBB-000D93C35C66@yahoo.fr> Message-ID: On 1/10/05 1:39 PM, "Olivier" wrote: > Hi list, > > A strange thing with my app happens to me. > I have a DB with 83 000 small records. > > When I click a menu of my app, a window appears with an Einhugur' > Datagrid (sort of fast listBox) fills 83 000 records. > Thus, when the window opens, a cursor is created to obtain 83 000 > records of the DB. ??? You sure ? Datagrid is perfect that it allow you NOT load 80,000 records! Please study it. It allow load only 20-30 records that must be displayed. > With Valentina 1, the first time when I open this window, the datagrid > puts approximately 1 second to fill. > If I close the window and what I open it again, the datagrid fills > almost immediately. > > With Valentina 2, the first time when I open this window, the datagrid > puts ......34 seconds to fill ! > If I close the window and what I open it again, the datagrid puts > approximately 1 second to fill. > > As the cursor is destroyed every time I close the window, I thought > that the problem was linked to the cache. > But if I puts the cache on 50 Mb, 1Mb or even 0.01 Mb, that changes > nothing! (Same on Valentina 1) > Nevertheless the DB makes 15 Mb with Valentina 1 and 37 Mb with > Valentina 2 ( UTF16). > > I also tried my DB with ViSQL for Valentina 1 and ViSQL for Valentina 2. > If I launch a query to select all the records of my table (83 000) : > > With ViSQL 1: > > - The first time: 0.13 seconds > - If I begin again: 0.06 seconds on average > > With ViSQL 2: > > - The first time: 0.28 seconds > - If I begin again: 0.22 seconds on average > > Olivier > > _______________________________________________ > Valentina-beta mailing list > Valentina-beta@lists.macserve.net > http://lists.macserve.net/mailman/listinfo/valentina-beta -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From sunshine at public.kherson.ua Mon Jan 10 13:50:16 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Mon Jan 10 05:50:26 2005 Subject: valentina 2 In-Reply-To: <42B5584A-62FD-11D9-BDBB-000D93C35C66@yahoo.fr> Message-ID: On 1/10/05 1:46 PM, "Olivier" wrote: > Ruslan, > > I speak only about problems, it is the purpose of this list! > But I am impressed all the same by the possibilities of Valentina 2. > I think that you made a very good work and that you will manage to > improve this beta. Thank you. > The storage Unicode is very practical. > ( But I worry about the size of files. My complete DB will make > approximately 500 to 600 Mb with Valentina 1. With Valentina 2 I am > afraid that it takes too much place (UTF16). Finally I shall see well) In worse case (if ALL your fields are strings) db will be bigger 2 times with UTF16. If you have many numeric fields, then they will have the same size of course. > The API methods are very elegant and very promising. > But if my application has to work later on a small local area network, > I am obliged to use cursors ? Also no. We have put even API way on network protocol. BUT!!!! You should use API on network only if you will have FAST network. If you going work via inet then SQL is the best way. -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From sunshine at public.kherson.ua Mon Jan 10 13:56:20 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Mon Jan 10 05:56:25 2005 Subject: first cursor In-Reply-To: <401F5323-62FC-11D9-BDBB-000D93C35C66@yahoo.fr> Message-ID: On 1/10/05 1:39 PM, "Olivier" wrote: > Hi list, > > A strange thing with my app happens to me. > I have a DB with 83 000 small records. > > When I click a menu of my app, a window appears with an Einhugur' > Datagrid (sort of fast listBox) fills 83 000 records. > Thus, when the window opens, a cursor is created to obtain 83 000 > records of the DB. > > With Valentina 1, the first time when I open this window, the datagrid > puts approximately 1 second to fill. > If I close the window and what I open it again, the datagrid fills > almost immediately. > > With Valentina 2, the first time when I open this window, the datagrid > puts ......34 seconds to fill ! > If I close the window and what I open it again, the datagrid puts > approximately 1 second to fill. Long. This again can be result of not cached INDEXES. > As the cursor is destroyed every time I close the window, I thought > that the problem was linked to the cache. > But if I puts the cache on 50 Mb, 1Mb or even 0.01 Mb, that changes > nothing! (Same on Valentina 1) > Nevertheless the DB makes 15 Mb with Valentina 1 and 37 Mb with > Valentina 2 ( UTF16). 15MB difference this is just 0.5 second to read from HDD > I also tried my DB with ViSQL for Valentina 1 and ViSQL for Valentina 2. > If I launch a query to select all the records of my table (83 000) : > > With ViSQL 1: > > - The first time: 0.13 seconds > - If I begin again: 0.06 seconds on average > > With ViSQL 2: > > - The first time: 0.28 seconds > - If I begin again: 0.22 seconds on average -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From vidal_olivier at yahoo.fr Mon Jan 10 13:06:43 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Mon Jan 10 06:06:37 2005 Subject: first cursor In-Reply-To: References: Message-ID: <17C12A28-6300-11D9-BDBB-000D93C35C66@yahoo.fr> >> > > ??? You sure ? > > Datagrid is perfect that it allow you NOT load 80,000 records! > Please study it. It allow load only 20-30 records that must be > displayed. > > Yes, sorry, I badly expressed myself. I meant that a cursor is created in the opening of the window. This cursor includes all the records of the table (thus 83000). But, effectively, the datagrid takes only in the cursor the records which it needs for the display in the screen. (According to the position of the slider which moves the user. When we move the slider, no problem, the datagrid and Valentina quickly react. It is only in the first display of the datagrid that it is very slow.) The problem is that in the first opening of the window, the display of the datagrid is very very long. Far too much. While there was no this problem with Valentina 1. olivier From vidal_olivier at yahoo.fr Mon Jan 10 13:09:48 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Mon Jan 10 06:09:44 2005 Subject: first cursor In-Reply-To: References: Message-ID: <85FF6896-6300-11D9-BDBB-000D93C35C66@yahoo.fr> >> >> With Valentina 2, the first time when I open this window, the datagrid >> puts ......34 seconds to fill ! >> If I close the window and what I open it again, the datagrid puts >> approximately 1 second to fill. > > Long. This again can be result of not cached INDEXES. Even for the first display of the datagrid? 34 seconds, it is really very long. >> As the cursor is destroyed every time I close the window, I thought >> that the problem was linked to the cache. >> But if I puts the cache on 50 Mb, 1Mb or even 0.01 Mb, that changes >> nothing! (Same on Valentina 1) >> Nevertheless the DB makes 15 Mb with Valentina 1 and 37 Mb with >> Valentina 2 ( UTF16). > > > 15MB difference this is just 0.5 second to read from HDD > ok. From vidal_olivier at yahoo.fr Mon Jan 10 16:28:58 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Mon Jan 10 09:29:05 2005 Subject: vCursor.import_text In-Reply-To: <1C012856-5FAA-11D9-B97F-0003939246BC@axint.net> References: <1C012856-5FAA-11D9-B97F-0003939246BC@axint.net> Message-ID: <58E3544C-631C-11D9-BDBB-000D93C35C66@yahoo.fr> Le 6 janv. 05, ? 07:13, Frank Schima a ?crit : > > My bug report is about how picky V 2 importing is compared to V 1. In > V 2, the number of fields in the text file must match *exactly* with > the number of fields in the SELECT of the cursor. If they do not match > in *any* single line of text, it will crash. > > V 1 was much more forgiving about filling in Null or blank values if > there were less text fields than SELECT fields in the cursor - or > ignoring the extra text fields if there were too many. I believe that Igor indicated on the mantis that it bugs was corrected. But with it b036, I always have crashes if THE ONLY ONE of the fields of A SINGLE LINE is empty. olivier From sunshine at public.kherson.ua Mon Jan 10 17:37:07 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Mon Jan 10 09:37:18 2005 Subject: vCursor.import_text In-Reply-To: <58E3544C-631C-11D9-BDBB-000D93C35C66@yahoo.fr> Message-ID: On 1/10/05 5:28 PM, "Olivier" wrote: > > Le 6 janv. 05, ? 07:13, Frank Schima a ?crit : >> >> My bug report is about how picky V 2 importing is compared to V 1. In >> V 2, the number of fields in the text file must match *exactly* with >> the number of fields in the SELECT of the cursor. If they do not match >> in *any* single line of text, it will crash. >> >> V 1 was much more forgiving about filling in Null or blank values if >> there were less text fields than SELECT fields in the cursor - or >> ignoring the extra text fields if there were too many. > > I believe that Igor indicated on the mantis that it bugs was corrected. > But with it b036, I always have crashes if THE ONLY ONE of the fields > of A SINGLE LINE is empty. Igor have correct bug reported by Frank. So if still some exists please send us DATA FILE you import, And how you do this -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From vidal_olivier at yahoo.fr Mon Jan 10 18:23:53 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Mon Jan 10 11:23:50 2005 Subject: vCursor.import_text In-Reply-To: References: Message-ID: <6665CF1C-632C-11D9-B3D8-000D93C35C66@yahoo.fr> Sorry. Effectively, that works. I think that I had had to try to import a UTF16 file text. (there : crash) By this method, we can import only ASCII texts? No accented characters for example? Because file texts in Mac Roman are correctly imported except the accented characters. thank you olivier Le 10 janv. 05, ? 16:37, Ruslan Zasukhin a ?crit : > On 1/10/05 5:28 PM, "Olivier" wrote: > >> >> Le 6 janv. 05, ? 07:13, Frank Schima a ?crit : >>> >>> My bug report is about how picky V 2 importing is compared to V 1. In >>> V 2, the number of fields in the text file must match *exactly* with >>> the number of fields in the SELECT of the cursor. If they do not >>> match >>> in *any* single line of text, it will crash. >>> >>> V 1 was much more forgiving about filling in Null or blank values if >>> there were less text fields than SELECT fields in the cursor - or >>> ignoring the extra text fields if there were too many. >> >> I believe that Igor indicated on the mantis that it bugs was >> corrected. >> But with it b036, I always have crashes if THE ONLY ONE of the fields >> of A SINGLE LINE is empty. > > Igor have correct bug reported by Frank. > > So if still some exists please send us DATA FILE you import, > And how you do this > > -- > Best regards, > Ruslan Zasukhin [ I feel the need...the need for speed ] > ------------------------------------------------------------- > e-mail: ruslan@paradigmasoft.com > web: http://www.paradigmasoft.com > > To subscribe to the Valentina mail list go to: > http://lists.macserve.net/mailman/listinfo/valentina > ------------------------------------------------------------- > > > _______________________________________________ > Valentina-beta mailing list > Valentina-beta@lists.macserve.net > http://lists.macserve.net/mailman/listinfo/valentina-beta > From sunshine at public.kherson.ua Mon Jan 10 19:32:56 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Mon Jan 10 11:33:00 2005 Subject: vCursor.import_text In-Reply-To: <6665CF1C-632C-11D9-B3D8-000D93C35C66@yahoo.fr> Message-ID: On 1/10/05 7:23 PM, "Olivier" wrote: > Sorry. Effectively, that works. > I think that I had had to try to import a UTF16 file text. (there : > crash) > By this method, we can import only ASCII texts? No accented characters > for example? No, it must accept files in any encoding, and with any line-ending. Right Igor ? We need check that we have open all parameters into V4RB. > Because file texts in Mac Roman are correctly imported except the > accented characters. -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From sunshine at public.kherson.ua Mon Jan 10 19:39:15 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Mon Jan 10 11:39:22 2005 Subject: [V4MD] 2.0 b36 WIN uploaded Message-ID: Hi Director lovers, You can download 2.0 b36 for Windows. Let us know how it works. Please help us test Valentina 2.0 as good as possible. -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From vidal_olivier at yahoo.fr Mon Jan 10 19:57:57 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Mon Jan 10 12:58:04 2005 Subject: crash on empty cursor Message-ID: <8AF1960C-6339-11D9-B3D8-000D93C35C66@yahoo.fr> Hi list, I have a cursor : SELECT recId, string1, string2 FROM Table1 ORDER by string1 If this cursor is empty, then, later, the "cursor.addrecord()" method crashes the app. Naturally, it is better to use " where RecId=0 " but sometimes we have not the choice. olivier From sunshine at public.kherson.ua Mon Jan 10 21:05:21 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Mon Jan 10 13:05:26 2005 Subject: crash on empty cursor In-Reply-To: <8AF1960C-6339-11D9-B3D8-000D93C35C66@yahoo.fr> Message-ID: On 1/10/05 8:57 PM, "Olivier" wrote: > Hi list, > > I have a cursor : > SELECT recId, string1, string2 FROM Table1 ORDER by string1 > > If this cursor is empty, then, later, the "cursor.addrecord()" method > crashes the app. > > Naturally, it is better to use " where RecId=0 " but sometimes we have > not the choice. Hmm, The correct way NOW produce empty cursor for adding of records is SELECT recId, string1, string2 FROM Table1 WHERE false ORDER by string1 -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From sunshine at public.kherson.ua Mon Jan 10 21:07:12 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Mon Jan 10 13:07:19 2005 Subject: V4RB1.11 export empty bug In-Reply-To: Message-ID: On 1/10/05 12:46 PM, "erne" wrote: > Hello Ruslan, > > I've found that if try to export an empty cursor to a text file Vale > will crash. > Not a big deal since no sense in exporting an empty cursor, > yet annoying since we must add check in code before export to avoid > crash. > > may I send you a demo project? Erne, Please try TODAYS build. Btw, today just in case I have rebuild and re-upload All 3 archives including VComponents -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From vidal_olivier at yahoo.fr Mon Jan 10 20:17:19 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Mon Jan 10 13:17:12 2005 Subject: crash on empty cursor In-Reply-To: References: Message-ID: <3F60E41B-633C-11D9-B3D8-000D93C35C66@yahoo.fr> Le 10 janv. 05, ? 20:05, Ruslan Zasukhin a ?crit : > On 1/10/05 8:57 PM, "Olivier" wrote: > >> Hi list, >> >> I have a cursor : >> SELECT recId, string1, string2 FROM Table1 ORDER by string1 >> >> If this cursor is empty, then, later, the "cursor.addrecord()" method >> crashes the app. In Valentina 1, it works. >> >> Naturally, it is better to use " where RecId=0 " but sometimes we have >> not the choice. > > Hmm, > > The correct way NOW produce empty cursor for adding of records is > > SELECT recId, string1, string2 > FROM Table1 > WHERE false > ORDER by string1 > > It is necessary to replace every " where recid=0 " by " where false "? (It would be necessary to correct doc) In the presented case, I do not know early if the cursor will be empty. It is more practical to be able to add records that the cursor is empty OR not. In Valentina 1, this query worked: >> SELECT recId, string1, string2 FROM Table1 ORDER by string1 rs=cursor.add() From sunshine at public.kherson.ua Mon Jan 10 21:44:25 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Mon Jan 10 13:44:32 2005 Subject: crash on empty cursor In-Reply-To: <3F60E41B-633C-11D9-B3D8-000D93C35C66@yahoo.fr> Message-ID: On 1/10/05 9:17 PM, "Olivier" wrote: >> Hmm, >> >> The correct way NOW produce empty cursor for adding of records is >> >> SELECT recId, string1, string2 >> FROM Table1 >> WHERE false >> ORDER by string1 >> >> > > It is necessary to replace every " where recid=0 " by " where false "? > (It would be necessary to correct doc) RecID = 0 still will work. Just WHERE false is the most effective > In the presented case, I do not know early if the cursor will be empty. > It is more practical to be able to add records that the cursor is empty > OR not. -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From sunshine at public.kherson.ua Mon Jan 10 21:44:54 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Mon Jan 10 13:45:00 2005 Subject: crash on empty cursor In-Reply-To: <3F60E41B-633C-11D9-B3D8-000D93C35C66@yahoo.fr> Message-ID: On 1/10/05 9:17 PM, "Olivier" wrote: > In the presented case, I do not know early if the cursor will be empty. > It is more practical to be able to add records that the cursor is empty > OR not. So again, Where is problem ? -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From snw at paradise.net.nz Tue Jan 11 09:25:51 2005 From: snw at paradise.net.nz (Sean Wilson) Date: Mon Jan 10 14:26:02 2005 Subject: [V4MD] 2.0 b36 WIN uploaded In-Reply-To: References: Message-ID: <6.2.0.14.2.20050111090714.03286dd0@pop3.paradise.net.nz> >You can download 2.0 b36 for Windows. > >Let us know how it works. It doesn't work yet. I downloaded the xtra, components and engine. I had been testing the earlier beta and so already had it (the earlier version) working correctly with the components folder in the correct place and listed in my system's PATH variable. When I replace the previous components (and engine files) folder with the new one, the xtra doesn't appear available to Director. I tried adding msvcp71.dll (7.10.3077.0) and msvcr71.dll (7.10.3052.4) from the earlier working components folder but the xtra still doesn't appear. I have rebooted. Any advice/suggestions? Thanks, -Sean. From sunshine at public.kherson.ua Tue Jan 11 00:49:56 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Mon Jan 10 16:50:00 2005 Subject: V4MD, VEngine_Win_VC.exe archive updated Message-ID: Hi All, I have update VEngine_Win_VC.exe and VComponents_win_VC We hope that now V4MD WIN will work. Please try again -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From snw at paradise.net.nz Tue Jan 11 12:08:11 2005 From: snw at paradise.net.nz (Sean Wilson) Date: Mon Jan 10 17:08:47 2005 Subject: V4MD, VEngine_Win_VC.exe archive updated In-Reply-To: References: Message-ID: <6.2.0.14.2.20050111120751.02768d10@pop3.paradise.net.nz> > >I have update VEngine_Win_VC.exe and VComponents_win_VC > >We hope that now V4MD WIN will work. Yes, it works now. Thanks, -Sean. From sunshine at public.kherson.ua Tue Jan 11 01:13:02 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Mon Jan 10 17:13:10 2005 Subject: V4MD, VEngine_Win_VC.exe archive updated In-Reply-To: <6.2.0.14.2.20050111120751.02768d10@pop3.paradise.net.nz> Message-ID: On 1/11/05 1:08 AM, "Sean Wilson" wrote: >> I have update VEngine_Win_VC.exe and VComponents_win_VC >> >> We hope that now V4MD WIN will work. > > Yes, it works now. ECXELLENT !!! :-)) Martin, please also try V4MD WIN b36 -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From vidal_olivier at yahoo.fr Tue Jan 11 05:46:53 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Mon Jan 10 22:47:03 2005 Subject: crash on empty cursor In-Reply-To: References: Message-ID: Hi Ruslan and list, >> In the presented case, I do not know early if the cursor will be >> empty. >> It is more practical to be able to add records that the cursor is >> empty >> OR not. > > So again, > > Where is problem ? The problem is that the cursor.addrecord method crashes if, BEFORE this method, the result of the cursor was empty (while we waited for records), what was not the case in Valentina 1. In my application I have the following cursor : >> SELECT recId, string1, string2 FROM Table1 ORDER by string1 This cursor returns generally records BUT in certain cases it is possible that it is empty. Much more late, I have a method : call cursor.Addrecord If the cursor was empty, the application crashes. olivier From vidal_olivier at yahoo.fr Tue Jan 11 08:25:58 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Tue Jan 11 01:26:11 2005 Subject: first cursor very very slow when it is sorted In-Reply-To: References: Message-ID: <0A0A756B-63A2-11D9-B070-000D93C35C66@yahoo.fr> Le 10 janv. 05, ? 12:56, Ruslan Zasukhin a ?crit : > Long. This again can be result of not cached INDEXES. after several tests, this SQL request take 1 second with 110 000 little records (4 strings, 1 boolean) : cur=new vcursor(DB,"SELECT recid,string1, string2, string3, boolean1 FROM Table ",2,3,2) this SQL request (on same 110 000 records) take .... 1 minute, 1 MINUTE !!! : cur=new vcursor(DB,"SELECT recid,string1, string2, string3, boolean1 FROM Table ORDER BY string1",2,3,2) string1 is indexed ! cache : 50 Mb This query put one minute THE FIRST TIME. If I relaunch the SAME query then, it will put ONLY one second. The sorting by "ORDER BY" is thus very very long. olivier -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: text/enriched Size: 770 bytes Desc: not available Url : http://lists.macserve.net/pipermail/valentina-beta/attachments/20050111/a28a7533/attachment.bin From sunshine at public.kherson.ua Tue Jan 11 10:17:47 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Tue Jan 11 02:17:55 2005 Subject: first cursor very very slow when it is sorted In-Reply-To: <0A0A756B-63A2-11D9-B070-000D93C35C66@yahoo.fr> Message-ID: On 1/11/05 9:25 AM, "Olivier" wrote: > Long. This again can be result of not cached INDEXES. > > after several tests, this SQL request take 1 second with 110 000 little > records (4 strings, 1 boolean) : > > cur=new vcursor(DB,"SELECT recid,string1, string2, string3, boolean1 FROM > Table ",2,3,2) > > > > this SQL request (on same 110 000 records) take .... 1 minute, 1 MINUTE !!! : > > cur=new vcursor(DB,"SELECT recid,string1, string2, string3, boolean1 FROM > Table ORDER BY string1",2,3,2) > > string1 is indexed ! > cache : 50 Mb > > This query put one minute THE FIRST TIME. > If I relaunch the SAME query then, it will put ONLY one second. > > The sorting by "ORDER BY" is thus very very long. Oliver, first time Valentina build special index for sorting. I again say, INDEXES are NOT cached now. This is a reason. Do not test for now big number of records. I think on this week we will make indexed to be cached. -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From ernestogiannotta at tiscalinet.it Tue Jan 11 09:21:20 2005 From: ernestogiannotta at tiscalinet.it (erne) Date: Tue Jan 11 02:21:51 2005 Subject: V4RB1.11 export empty bug In-Reply-To: Message-ID: Hi Ruslan, on 10-01-2005 20:07, Ruslan Zasukhin at sunshine@public.kherson.ua wrote: >> I've found that if try to export an empty cursor to a text file Vale >> will crash. >> Not a big deal since no sense in exporting an empty cursor, >> yet annoying since we must add check in code before export to avoid >> crash. >> >> may I send you a demo project? > > Erne, > > Please try TODAYS build. > > Btw, today just in case I have rebuild and re-upload > All 3 archives including VComponents > please note I'm talking about V1.11 Cool Runnings, Erne. -- | e r | Ernesto Giannotta | n e | Musical Box - a media store From sunshine at public.kherson.ua Tue Jan 11 10:34:21 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Tue Jan 11 02:34:26 2005 Subject: V4RB1.11 export empty bug In-Reply-To: Message-ID: On 1/11/05 10:21 AM, "erne" wrote: > >>> I've found that if try to export an empty cursor to a text file Vale >>> will crash. >>> Not a big deal since no sense in exporting an empty cursor, >>> yet annoying since we must add check in code before export to avoid >>> crash. >>> >>> may I send you a demo project? >> >> Erne, >> >> Please try TODAYS build. >> >> Btw, today just in case I have rebuild and re-upload >> All 3 archives including VComponents >> > > please note I'm talking about V1.11 Erne, beta list this is for 2.0 discussion. -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From bibiko at eva.mpg.de Tue Jan 11 09:54:11 2005 From: bibiko at eva.mpg.de (Hans-Joerg Bibiko) Date: Tue Jan 11 02:54:41 2005 Subject: [V4MD] 2.0 b36 MAC - hint+question In-Reply-To: References: Message-ID: <41E39433.9010000@eva.mpg.de> Hi Ruslan, 1) could it be that you confused the folder names for installing V4MD_2 on a Mac? On http://paradigmasoft.com/download/v2_beta/v4md_beta.html you wrote: Install VEngine files into VComponents folder within and the VComponents file into Library/CFMSupport/. I believe just the opposite: VEngine files into Library/CFMSupport/ and VComponents into Library/CFMSupport/VComponents. Then it works for me. ? 2) Is there a routine to convert a V4MD_1.x DB into V4MD_2 DB or I have to do it via dump? 3) I rebuilt my DB via dump. OK Init,Open,Close,Flush,Defragment,Compact Valentina OK gCur = gDB.sqlSelect( "select ....") OK cnt = gCur.recordCount OK res = getRecordsAsPropList(gCur) OK **BUT** If my field type is 'TEXT' then I get something wrong: e.g. I stored a string of a number like 13 within this TEXT field. (I know this shouldn't be done, but ...) with integer(gCur.getRecordAsString) I will get 31 ? I stored a huge string contains a proplist. I was returned 'only' hex values of the string like '41 72 69 61 6c' instead of 'Arial' Is this correctly? Did I make a mistake? Hans From vidal_olivier at yahoo.fr Tue Jan 11 10:03:38 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Tue Jan 11 03:03:37 2005 Subject: first cursor very very slow when it is sorted In-Reply-To: References: Message-ID: Le 11 janv. 05, ? 09:17, Ruslan Zasukhin a ?crit : > I think on this week we will make indexed to be cached. perfect ! From vidal_olivier at yahoo.fr Tue Jan 11 10:07:11 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Tue Jan 11 03:07:04 2005 Subject: =?iso-8859-1?q?m=E9thods_API_vs_SQL?= Message-ID: <2D88D31E-63B0-11D9-B070-000D93C35C66@yahoo.fr> Hi list, In my tests, the API methods are faster than the SQL methods, but not as long as that. When the software will be optimized, the distance between both is going to increase? Or both will go more quickly but by keeping the same distance? thank you olivier From sunshine at public.kherson.ua Tue Jan 11 11:41:54 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Tue Jan 11 03:42:01 2005 Subject: [V4MD] 2.0 b36 MAC - hint+question In-Reply-To: <41E39433.9010000@eva.mpg.de> Message-ID: On 1/11/05 10:54 AM, "Hans-Joerg Bibiko" wrote: > Hi Ruslan, > > 1) > could it be that you confused the folder names for installing V4MD_2 on a Mac? > > On http://paradigmasoft.com/download/v2_beta/v4md_beta.html you wrote: > > Install VEngine files into VComponents folder within and the VComponents file > into > Library/CFMSupport/. I believe just the opposite: > > VEngine files into Library/CFMSupport/ and > VComponents into Library/CFMSupport/VComponents. > > Then it works for me. > > ? No no no. step1: Drag and drop VComponents folder into "/Library/CFMSupport" step2: you need COPY all files from archive Vengine into /Library/CFMSupport/VComponents So on HDD you have only ONE folder /Library/CFMSupport/VComponents All files from both our archives are inside of Vcomponents folder. You must COPY files of Vengine into Vcomponents !!! Not a folder itself. Vcomponents VEngine -- this is wrong! Must be Vcomponents kernel.shlb resources icu libs -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From sunshine at public.kherson.ua Tue Jan 11 11:55:54 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Tue Jan 11 03:56:23 2005 Subject: [V4MD] 2.0 b36 MAC - hint+question In-Reply-To: <41E39433.9010000@eva.mpg.de> Message-ID: On 1/11/05 10:54 AM, "Hans-Joerg Bibiko" wrote: > 2) > Is there a routine to convert a V4MD_1.x DB into V4MD_2 DB or I have to do it > via dump? We going to introduce it soon. XML also is a way, but we want faster -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From sunshine at public.kherson.ua Tue Jan 11 12:00:56 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Tue Jan 11 04:01:01 2005 Subject: [V4MD] 2.0 b36 MAC - hint+question In-Reply-To: <41E39433.9010000@eva.mpg.de> Message-ID: On 1/11/05 10:54 AM, "Hans-Joerg Bibiko" wrote: > 3) > I rebuilt my DB via dump. > OK cool > Init,Open,Close,Flush,Defragment,Compact Valentina > OK > > gCur = gDB.sqlSelect( "select ....") > OK > > cnt = gCur.recordCount > OK > > res = getRecordsAsPropList(gCur) > OK **BUT** > If my field type is 'TEXT' then I get something wrong: Igor, you need check this > e.g. > I stored a string of a number like 13 within this TEXT field. (I know this > shouldn't be done, but ...) > with integer(gCur.getRecordAsString) I will get 31 ? > > I stored a huge string contains a proplist. > I was returned 'only' hex values of the string like '41 72 69 61 6c' instead > of 'Arial' > > Is this correctly? Did I make a mistake? I think our bug -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From sunshine at public.kherson.ua Tue Jan 11 12:06:26 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Tue Jan 11 04:06:32 2005 Subject: m =?iso-8859-1?q?=E9?= thods API vs SQL In-Reply-To: <2D88D31E-63B0-11D9-B070-000D93C35C66@yahoo.fr> Message-ID: On 1/11/05 11:07 AM, "Olivier" wrote: Hi Oliver, > In my tests, the API methods are faster than the SQL methods, but not > as long as that. > When the software will be optimized, the distance between both is going > to increase? > Or both will go more quickly but by keeping the same distance? RULE: SQL way will be always somehow slower because exists overhead of SQL parser. SQL after parsing use the same API methods. -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From bibiko at eva.mpg.de Tue Jan 11 11:38:58 2005 From: bibiko at eva.mpg.de (Hans-Joerg Bibiko) Date: Tue Jan 11 04:39:15 2005 Subject: [V4MD] 2.0b MAC - LIKE problem on a VarChar field In-Reply-To: References: Message-ID: <41E3ACC2.2040609@eva.mpg.de> Hi Ruslan, I copied the folder like you mentioned. OK Problem: gDB.SQLSelect("select * name,code from language where name = 'German'") OK gDB.SQLSelect("select * name,code from language where name LIKE 'German'") OK but it finds only one record but I have 17 names beginning with 'German' gDB.SQLSelect("select * name,code from language where name like 'german' no_case") SQLError: token 'german' couldn't parsed Hans -- **************************************************************** Hans-Joerg Bibiko Max Planck Institute for Evolutionary Anthropology Department of Linguistics Deutscher Platz 6 phone: +49 (0) 341 35 50 341 D-04103 Leipzig fax: +49 (0) 341 35 50 333 Germany e-mail: bibiko@eva.mpg.de **************************************************************** From giv at tlc.kherson.ua Tue Jan 11 12:58:56 2005 From: giv at tlc.kherson.ua (Igor Gomon) Date: Tue Jan 11 04:58:56 2005 Subject: [V4MD] 2.0 b36 MAC - hint+question References: Message-ID: <008d01c4f7cc$8c0abec0$3b04a8c0@giv> > > res = getRecordsAsPropList(gCur) > > OK **BUT** > > If my field type is 'TEXT' then I get something wrong: > > Igor, you need check this I wrote the next code to check this: global Valentina -- on startMovie Valentina = new (Xtra "Valentina") if CheckValError() then exit Valentina.init("", 8 * 1024 * 1024, "", "") if CheckValError() then exit db = new (Xtra "VDatabase") if CheckValError() then exit db.create(the moviePath & "Db1.vdb") if CheckValError() then exit t1 = db.createTable("Table1") if CheckValError() then exit fLong = t1.createLongField("fLong") if CheckValError() then exit fText = t1.createTextField("fText", 1024) if CheckValError() then exit -- Add one record. fLong.value = 1 fText.value = "AAA" t1.addRecord() if CheckValError() then exit cur = db.sqlSelect("SELECT * FROM Table1") if CheckValError() then exit rec = cur.getRecordAsPropList() if CheckValError() then exit put rec db.close() db.throwOut() Valentina.shutDown() end The resulting output is: -- [#fLong: 1, #fText: "AAA"] So, on my side all is working fine. Can you give me an example of code that gives you wrong result? -- Best regards, Igor Gomon ------------------------------------------------------------- e-mail: giv@tlc.kherson.ua web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://listserv.macserve.net/mailman/listinfo/valentina From sunshine at public.kherson.ua Tue Jan 11 13:29:02 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Tue Jan 11 05:29:09 2005 Subject: [V4MD] 2.0b MAC - LIKE problem on a VarChar field In-Reply-To: <41E3ACC2.2040609@eva.mpg.de> Message-ID: On 1/11/05 12:38 PM, "Hans-Joerg Bibiko" wrote: > Hi Ruslan, > > I copied the folder like you mentioned. OK > > Problem: > > gDB.SQLSelect("select * name,code from language where name = 'German'") > OK > > gDB.SQLSelect("select * name,code from language where name LIKE 'German'") > OK but it finds only one record but I have 17 names beginning with 'German' > > gDB.SQLSelect("select * name,code from language where name like 'german' > no_case") > SQLError: token 'german' couldn't parsed 2.0 do not have key word no_case....at least now -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From bibiko at eva.mpg.de Tue Jan 11 13:21:59 2005 From: bibiko at eva.mpg.de (Hans-Joerg Bibiko) Date: Tue Jan 11 06:22:14 2005 Subject: [V4MD] 2.0 b36 MAC - TEXT field problem In-Reply-To: <008d01c4f7cc$8c0abec0$3b04a8c0@giv> References: <008d01c4f7cc$8c0abec0$3b04a8c0@giv> Message-ID: <41E3C4E7.8050009@eva.mpg.de> HI Igor, your code is working fine ! but I guess I found the problem. The problem is I rebuilt my DB via XML LoadDump. In my XML file which is coded with ISO-8859-1 is written: .... 2 VarChar 1 0 0 0 0 20 ASCII Text 0 0 1 0 0 ASCII 128 ... 7 1 wsize 31 34 2 wface 41 72 69 61 6C 3 li_lang 5B 5B 23 74 69 74 6C 65 3A 20 22 6D 6F 76 65 20 61 6C 6C 20 6C 61 6E 67 75 61 67 65 73 20 74 6F 20 74 68 65 20 53 65 6C 65 63 74 69 6F 6E 20 57 ... As you can see above the VCursor routine get the right data but there are saved in a wrong way. I will rebuild my DB without using loadDump later if I have a bit more time. Thank you and I found a lot of good stuff within the tutorials ;) You all did and are doing a great job! Hans From sunshine at public.kherson.ua Wed Jan 12 11:48:14 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Wed Jan 12 03:48:22 2005 Subject: RegisterModule or class, AFTER PluginEntry() In-Reply-To: <12602574.1105522362178.JavaMail.james.milne@mac.com> Message-ID: On 1/12/05 11:32 AM, "James Milne" wrote: >>> Hi All, >>> >>> Is it possible to register another class/module/else >>> Not in the PluginEntry() function, but LATER ? >>> In some other function ? >>> >>> Any technical limitations present here? >>> Or we must be able do this ? >> >> So anybody can comment ? > > It's not possible. > > REALbasic expects you to have registered everything your plugin provides in > the PluginEntry function. > > The REALbasic plugin loader loads all the plugins it can find and calls each > plugin's PluginEntry() function. It assumes that after it has called all the > PluginEntry functions that all the resources provided by the plugin have been > registered. It then sorts out dependencies by resolving superclasses for > plugin classes, etc, so that you can have inter-plugin dependencies. > > I'd say that you can safely assume none of the REALRegister* Plugin API > functions are usable outside of the PluginEntry() function. I see, thank you James. Yes I see 100% crashes. Bad. Then I have problems....may be somebody have an idea. Task is next: * when REALbasic starts and loads plugin we have in the /Library/CFMSupport/Vcomponents folder with our dlls and other stuf. in particular this folder contains several XML files, which during PluginEntry() we parse and build error constants for V4RB. * Now I want allow to RB developers build Application, and move all files (DLLs and rest) into Application folder, to have distribution as SINGLE FOLDER. For this developer will need self assign directory where these files are. But developer can do this only AFTER PluginEntry(). Hmm, so plugin self during PluginEntry() must try to find where are required files....well I think this is the only solution. -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From vidal_olivier at yahoo.fr Thu Jan 13 07:24:55 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Thu Jan 13 00:25:01 2005 Subject: exception on open of my DB Message-ID: Hi Ruslan and list, this morning, I have an exception (catchable) N? 365826, message : "file 'Ox600050007' does not exist' ". This exception occurs at open of my DB : app.open : WindowBase.base.open(nf) strangely, after, my app works, appently, without problems. the exception occurs then at each launch of my app. I created new DB. There are not problems anymore. olivier From sunshine at public.kherson.ua Thu Jan 13 09:27:16 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Thu Jan 13 01:27:24 2005 Subject: exception on open of my DB In-Reply-To: Message-ID: On 1/13/05 8:24 AM, "Olivier" wrote: > Hi Ruslan and list, > > this morning, I have an exception (catchable) N? 365826, message : > "file 'Ox600050007' does not exist' ". > This exception occurs at open of my DB : > > app.open : > > WindowBase.base.open(nf) > > strangely, after, my app works, appently, without problems. > > the exception occurs then at each launch of my app. May be you have stop app in debugger ? > I created new DB. There are not problems anymore. Yes, only new one -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From vidal_olivier at yahoo.fr Thu Jan 13 09:22:28 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Thu Jan 13 02:22:30 2005 Subject: exception on open of my DB In-Reply-To: References: Message-ID: <4301C766-653C-11D9-9FA3-000D93C35C66@yahoo.fr> Le 13 janv. 05, ? 08:27, Ruslan Zasukhin a ?crit : > May be you have stop app in debugger ? I do not think. But I am not sure. The next time when it occurs, I shall prevent you. olivier From vidal_olivier at yahoo.fr Thu Jan 13 09:37:43 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Thu Jan 13 02:37:39 2005 Subject: SQL cursors and API Message-ID: <64A3DFA0-653E-11D9-9FA3-000D93C35C66@yahoo.fr> Hi, I am upgrading my project of Valentina 1 to Valentina 2. In certain methods, I replace the SQL cursors by API methods. But for other windows, I keep SQL cursors. Does it raise problems to make live SQL cursors and API methods? I suppose that if one day I want that my app works in local network, all the application has to work with cursors OR with API methods ? olivier From sunshine at public.kherson.ua Thu Jan 13 12:24:18 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Thu Jan 13 04:24:25 2005 Subject: SQL cursors and API In-Reply-To: <64A3DFA0-653E-11D9-9FA3-000D93C35C66@yahoo.fr> Message-ID: On 1/13/05 10:37 AM, "Olivier" wrote: > Hi, > > I am upgrading my project of Valentina 1 to Valentina 2. > In certain methods, I replace the SQL cursors by API methods. > But for other windows, I keep SQL cursors. > > Does it raise problems to make live SQL cursors and API methods? In 2.0 this should work... Although may be will be side effects as after Cursor.NextRecord Your Table current record is changed. Deal is Table is only one. You can have only one current record in the table. Cursors can be many around table. Each cursor have own buffer of field vales. But they use Table to read that values. > I suppose that if one day I want that my app works in local network, > all the application has to work with cursors OR with API methods ? API methods also work on netwrok. Just I did say, this will have sense only in the local fast netwrok. For REMOTE dbs on inet it is ALWAYS should be used SQL, Because of speed. -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From sunshine at public.kherson.ua Thu Jan 13 14:59:53 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Thu Jan 13 07:00:12 2005 Subject: [V4RB 0000059]: Set path to VComponents In-Reply-To: Message-ID: On 1/13/05 2:44 PM, "jda" wrote: >> In fact Valentina now self will do all job. Property is not needed! >> > > Hi Ruslan, that's great. But, hm, I'm not sure what your comment > means. Valentina searches the hard drive for VComponents? What > happens if there is more than one folder? For MacOS we have next rule 1) search in APP folder. 2) if not found in App folder then search Home/Library/CFMSupport/Vcomponents point 2 is not implemented yet. 3) if not found then search Library/CFMSupport/Vcomponents So the most correct step I think will be for developer to 1) compile own app as CARBON (not Macho) 2) simply copy all files from VComponents to App folder. You will get something as: MyAppFolder MyApplication icu lis resources kenrel libs ----------- Still is open question: should we future provide support of Library/CFMSupport/MyAppComponents I.e. Developer may wish still to have VComponents folder as separate one. Hmm, I think I can easy do this. Just use "nameofapp SPACE Vcomponents" "viSQL Components" -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From macsforever2000 at axint.net Thu Jan 13 08:13:52 2005 From: macsforever2000 at axint.net (Frank Schima) Date: Thu Jan 13 09:13:34 2005 Subject: [V4RB 0000059]: Set path to VComponents In-Reply-To: References: Message-ID: Hi guys, On Jan 13, 2005, at 5:59 AM, Ruslan Zasukhin wrote: > On 1/13/05 2:44 PM, "jda" wrote: > >>> In fact Valentina now self will do all job. Property is not needed! >>> >> >> Hi Ruslan, that's great. But, hm, I'm not sure what your comment >> means. Valentina searches the hard drive for VComponents? What >> happens if there is more than one folder? > > For MacOS we have next rule > > 1) search in APP folder. > > 2) if not found in App folder then search > Home/Library/CFMSupport/Vcomponents > > point 2 is not implemented yet. > > 3) if not found then search > Library/CFMSupport/Vcomponents > > > > So the most correct step I think will be for developer to > 1) compile own app as CARBON (not Macho) > 2) simply copy all files from VComponents to App folder. > > You will get something as: > > MyAppFolder > MyApplication > icu lis > resources > kenrel libs Can we have the option of compiling a MachO application and placing the VComponents inside the .app? So they would reside in MyApp.app/Contents/MacOS/VComponents/ That way they are completely hidden from view and we can be a first class Mac OS X application. > ----------- > Still is open question: > should we future provide support of > > Library/CFMSupport/MyAppComponents > > I.e. Developer may wish still to have VComponents folder as separate > one. > Hmm, I think I can easy do this. Just use "nameofapp SPACE Vcomponents" > > "viSQL Components" Nice, but I'd rather have the option mentioned above. From sunshine at public.kherson.ua Thu Jan 13 17:18:19 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Thu Jan 13 09:18:29 2005 Subject: [V4RB 0000059]: Set path to VComponents In-Reply-To: Message-ID: On 1/13/05 5:13 PM, "Frank Schima" wrote: >> For MacOS we have next rule >> >> 1) search in APP folder. >> >> 2) if not found in App folder then search >> Home/Library/CFMSupport/Vcomponents >> >> point 2 is not implemented yet. >> >> 3) if not found then search >> Library/CFMSupport/Vcomponents >> >> >> >> So the most correct step I think will be for developer to >> 1) compile own app as CARBON (not Macho) >> 2) simply copy all files from VComponents to App folder. >> >> You will get something as: >> >> MyAppFolder >> MyApplication >> icu lis >> resources >> kenrel libs > > Can we have the option of compiling a MachO application and placing the > VComponents inside the .app? So they would reside in > > MyApp.app/Contents/MacOS/VComponents/ > > That way they are completely hidden from view and we can be a first > class Mac OS X application. I have try this Frank, but it crashes on not clear reason. Crashes it seems in the code of REALbasic. It needs to play more with this. Just even now I have kill 3 full days to make it working. In Valentina studio this works, but Jochen use our C++ macho libs. V4RB is related to our Carbon libs. I do not see why this can't work. So may be in near future we will find solution. -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From vidal_olivier at yahoo.fr Thu Jan 13 18:19:12 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Thu Jan 13 11:19:19 2005 Subject: virtuals fields / Bo methods Message-ID: <3E6E0118-6587-11D9-9FA3-000D93C35C66@yahoo.fr> Hi, My virtuals fields does not work any more correctly. I have the following structure (created by classes) : DB1 "Table1" "String1" "String2" String2.methodText="String1" String2.isindexed=true String1.indexbyword=true This structure worked in Valentina 1. If I ask for a cursor sorted by String2, it worked. now no. in ViSQL, We cannot see virtual fields? olivier From sunshine at public.kherson.ua Thu Jan 13 20:47:14 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Thu Jan 13 12:47:21 2005 Subject: virtuals fields / Bo methods In-Reply-To: <3E6E0118-6587-11D9-9FA3-000D93C35C66@yahoo.fr> Message-ID: On 1/13/05 7:19 PM, "Olivier" wrote: > Hi, > > My virtuals fields does not work any more correctly. > > I have the following structure (created by classes) : > > DB1 > > "Table1" > > "String1" > "String2" > String2.methodText="String1" > String2.isindexed=true > String1.indexbyword=true > > This structure worked in Valentina 1. This may not work now. Now note that constructors of fields have parameter Method. So try set method text via this parameter, but not via String2.methodText Also now exists 3d parameter flags, so you can Vstring( "strring2", EVFlag.kIndexed + EVFlag.kIndexByWords, ="String1" ) > If I ask for a cursor sorted by String2, it worked. now no. > > in ViSQL, We cannot see virtual fields? You can! They have fldName = formula -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From vidal_olivier at yahoo.fr Thu Jan 13 20:14:44 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Thu Jan 13 13:14:45 2005 Subject: virtuals fields / Bo methods In-Reply-To: References: Message-ID: <62413824-6597-11D9-9FA3-000D93C35C66@yahoo.fr> >> String2.methodText="String1" > This may not work now. > it's in new doc. > Now note that constructors of fields have parameter Method. > So try set method text via this parameter, but not via > String2.methodText > how ? string2=new vstring("string2",38,EVflag.fmethod("string1") ) does not work. > Also now exists 3d parameter flags, so you can > 3d parameters ?! It is what? > Vstring( "strring2", EVFlag.kIndexed + EVFlag.kIndexByWords, > ="String1" ) > in RB, I have no EVflag.kindexed or EVflag.kindexbyword. From vidal_olivier at yahoo.fr Thu Jan 13 20:25:03 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Thu Jan 13 13:24:59 2005 Subject: virtuals fields / Bo methods In-Reply-To: <62413824-6597-11D9-9FA3-000D93C35C66@yahoo.fr> References: <62413824-6597-11D9-9FA3-000D93C35C66@yahoo.fr> Message-ID: > >> Now note that constructors of fields have parameter Method. >> So try set method text via this parameter, but not via >> String2.methodText >> > how ? > string2=new vstring("string2",38,EVflag.fmethod("string1") ) > does not work. > > excuse me. I m tired this evening ! ok thats works.. thank you olivier From sunshine at public.kherson.ua Thu Jan 13 21:27:59 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Thu Jan 13 13:28:07 2005 Subject: virtuals fields / Bo methods In-Reply-To: <62413824-6597-11D9-9FA3-000D93C35C66@yahoo.fr> Message-ID: On 1/13/05 9:14 PM, "Olivier" wrote: >>> String2.methodText="String1" >> This may not work now. >> > it's in new doc. I already have correct this. Property will be used mainly for CHANGE of text. >> Also now exists 3d parameter flags, so you can >> > 3d parameters ?! It is what? I mean third parameter >> Vstring( "strring2", EVFlag.kIndexed + EVFlag.kIndexByWords, >> ="String1" ) >> > in RB, I have no EVflag.kindexed or EVflag.kindexbyword. Ops Vstring( "strring2", 50 EVFlag.fIndexed + EVFlag.fIndexByWords, "String1" ) -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From sunshine at public.kherson.ua Thu Jan 13 21:30:37 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Thu Jan 13 13:30:41 2005 Subject: virtuals fields / Bo methods In-Reply-To: Message-ID: On 1/13/05 9:25 PM, "Olivier" wrote: >>> Now note that constructors of fields have parameter Method. >>> So try set method text via this parameter, but not via >>> String2.methodText >>> >> how ? >> string2=new vstring("string2",38,EVflag.fmethod("string1") ) >> does not work. >> >> > excuse me. I m tired this evening ! > > ok thats works.. Wait wait wait! this is not correct! string2=new vstring("string2",38,EVflag.fmethod("string1") ) This should not even compile! -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From vidal_olivier at yahoo.fr Thu Jan 13 20:45:21 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Thu Jan 13 13:45:22 2005 Subject: virtuals fields / Bo methods In-Reply-To: References: Message-ID: Le 13 janv. 05, ? 20:30, Ruslan Zasukhin a ?crit : >> excuse me. I m tired this evening ! >> >> ok thats works.. > > Wait wait wait! > > this is not correct! > > string2=new vstring("string2",38,EVflag.fmethod("string1") ) > > This should not even compile! yes, I had tried : string2=new vstring("string2",38,17,"string1") And that worked. in valentina 1 doc : indexed=1 method=16 but indexbyword ? olivier From sunshine at public.kherson.ua Thu Jan 13 22:10:14 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Thu Jan 13 14:10:22 2005 Subject: virtuals fields / Bo methods In-Reply-To: Message-ID: On 1/13/05 9:45 PM, "Olivier" wrote: >> Wait wait wait! >> >> this is not correct! >> >> string2=new vstring("string2",38,EVflag.fmethod("string1") ) >> >> This should not even compile! > > yes, I had tried : > > string2=new vstring("string2",38,17,"string1") > > And that worked. > > > in valentina 1 doc : > > indexed=1 > method=16 > > but indexbyword ? Oliver, once again. Must be Vstring( "strring2", 38, EVFlag.fIndexed + EVFlag.fIndexByWords, "String1" ) Use enum contants. Do not use numbers. In 2.0 values of constants was changed. So if you use constants you do not care about their values. -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From jordan at digi-net.com Thu Jan 13 16:20:42 2005 From: jordan at digi-net.com (Ken Jordan) Date: Thu Jan 13 15:18:24 2005 Subject: V2 compatibility Message-ID: <41E6E62A.2020204@digi-net.com> Is V2 V4RB compatible with Windows? If so, what is the installation procedure? From vidal_olivier at yahoo.fr Fri Jan 14 08:29:28 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Fri Jan 14 01:29:34 2005 Subject: virtuals fields / Bo methods In-Reply-To: References: Message-ID: <05F3FD78-65FE-11D9-9687-000D93C35C66@yahoo.fr> > Hi Ruslan, > Oliver, once again. Must be > > Vstring( "strring2", 38, > EVFlag.fIndexed + EVFlag.fIndexByWords, "String1" ) > > > Use enum contants. Do not use numbers. > In 2.0 values of constants was changed. > > So if you use constants you do not care about their values. > ok, Thus now, we can have several indexs (indexed, indexbyword) for a field without needing virtual field? -> make doc. For the efficiency and of the size of indexs and DB, it is the same thing as the supplementary virtual field? thank you olivier From sunshine at public.kherson.ua Fri Jan 14 09:42:18 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Fri Jan 14 01:42:31 2005 Subject: V2 compatibility In-Reply-To: <41E6E62A.2020204@digi-net.com> Message-ID: On 1/13/05 11:20 PM, "Ken Jordan" wrote: > Is V2 V4RB compatible with Windows? If so, what is the installation > procedure? We here work with V4RB windows, but we did not upload it yet. You work with REALbasic on PC ? -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From sunshine at public.kherson.ua Fri Jan 14 10:03:13 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Fri Jan 14 02:03:21 2005 Subject: virtuals fields / Bo methods In-Reply-To: <05F3FD78-65FE-11D9-9687-000D93C35C66@yahoo.fr> Message-ID: On 1/14/05 9:29 AM, "Olivier" wrote: > Hi Ruslan, > >> Oliver, once again. Must be >> >> Vstring( "strring2", 38, >> EVFlag.fIndexed + EVFlag.fIndexByWords, "String1" ) >> >> >> Use enum contants. Do not use numbers. >> In 2.0 values of constants was changed. >> >> So if you use constants you do not care about their values. >> > > ok, > > Thus now, we can have several indexs (indexed, indexbyword) for a field > without needing virtual field? > -> make doc. No no. Oliver, everything here like in 1.x ! Field can have INDEX. INDEX can be marked as UNIQUE, or IndexByWords. Virtual fields solve other tasks. > For the efficiency and of the size of indexs and DB, it is the same > thing as the supplementary virtual field? I think you have not correctly understand me. This YOUR OLD code "String2" String2.methodText="String1" String2.isindexed=true String1.indexbyword=true Is the same to Vstring( "strring2", 38, EVFlag.fIndexed + EVFlag.fIndexByWords, "String1" ) -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From vidal_olivier at yahoo.fr Fri Jan 14 09:22:55 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Fri Jan 14 02:22:56 2005 Subject: virtuals fields / Bo methods In-Reply-To: References: Message-ID: <7DB2AB74-6605-11D9-9687-000D93C35C66@yahoo.fr> Le 14 janv. 05, ? 09:03, Ruslan Zasukhin a ?crit : > This YOUR OLD code > > "String2" > String2.methodText="String1" > String2.isindexed=true > String1.indexbyword=true > > Is the same to > > Vstring( "strring2", 38, > EVFlag.fIndexed + EVFlag.fIndexByWords, "String1" ) then string1=new vstring("string1",38,EVflag.fIndexByWords) string2=new vstring("string2",38,EVflag.fIndexed,"string1") = string1=new vstring("string1",38) string2=new vstring("string2",38,EVflag.fIndexed + EVflag.fIndexByWords, "string1") right ? From sunshine at public.kherson.ua Fri Jan 14 10:27:27 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Fri Jan 14 02:27:37 2005 Subject: virtuals fields / Bo methods In-Reply-To: <7DB2AB74-6605-11D9-9687-000D93C35C66@yahoo.fr> Message-ID: On 1/14/05 10:22 AM, "Olivier" wrote: >> This YOUR OLD code >> >> "String2" >> String2.methodText="String1" >> String2.isindexed=true >> String1.indexbyword=true >> >> Is the same to >> >> Vstring( "strring2", 38, >> EVFlag.fIndexed + EVFlag.fIndexByWords, "String1" ) > > then Aha, yes I see. You assign flags to different fields. > string1=new vstring("string1",38,EVflag.fIndexByWords) > string2=new vstring("string2",38,EVflag.fIndexed,"string1") > > = > > string1=new vstring("string1",38) > string2=new vstring("string2",38,EVflag.fIndexed + > EVflag.fIndexByWords, "string1") This will be correct string1=new vstring("string1",38, EVflag.fIndexed + EVflag.fIndexByWords) string2=new vstring("string2",38,EVflag.fIndexed,"string1") String1 - is indexed AND index is by words String2 - is indexed and it is method -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From vidal_olivier at yahoo.fr Fri Jan 14 13:15:54 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Fri Jan 14 06:16:03 2005 Subject: virtuals fields / ViSQL In-Reply-To: References: Message-ID: <09B4D542-6626-11D9-9687-000D93C35C66@yahoo.fr> > >> string1=new vstring("string1",38,EVflag.fIndexByWords) >> string2=new vstring("string2",38,EVflag.fIndexed,"string1") > > > This will be correct > > string1=new vstring("string1",38, EVflag.fIndexed + > EVflag.fIndexByWords) > string2=new vstring("string2",38,EVflag.fIndexed,"string1") > > > String1 - is indexed AND index is by words > String2 - is indexed and it is method > ok, that's works. But I does not understand why it needs there to put "string1" indexed. We have a string field. For this string field, we want two indexs: indexbyword + indexed. In Valentina 1: We put, for the string field, "indexbyword". Then we created a second string field with for method : the first string field + flag indexed. Logic. Why in Valentina 2, we have to put the flag indexed in both fields? In ViSQL, I think that there is a problem with the methods. I have a string field which is a method, "string2". In ViSQL, if I launch " SELECT * FROM table1 Where string2=' xxx ' : OK, no problems, it finds. But if I launch " SELECT * FROM table1 ORDER BY string2 ", ViSQL indicates "Field string2 not exist". While this query works in my application. olivier From sunshine at public.kherson.ua Fri Jan 14 15:08:23 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Fri Jan 14 07:08:32 2005 Subject: virtuals fields / ViSQL In-Reply-To: <09B4D542-6626-11D9-9687-000D93C35C66@yahoo.fr> Message-ID: On 1/14/05 2:15 PM, "Olivier" wrote: >>> string1=new vstring("string1",38,EVflag.fIndexByWords) >>> string2=new vstring("string2",38,EVflag.fIndexed,"string1") >> >> >> This will be correct >> >> string1=new vstring("string1",38, EVflag.fIndexed + >> EVflag.fIndexByWords) >> string2=new vstring("string2",38,EVflag.fIndexed,"string1") >> >> >> String1 - is indexed AND index is by words >> String2 - is indexed and it is method >> > > ok, that's works. > But I does not understand why it needs there to put "string1" indexed. Because Valentina 2.0 little differ. If you set only flag fIndexByWords this not means yet create index. Valentina 2.0 now can do searches on NON-indexed fields. So to really build index, you need explicitly say: fIndexed. -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From sunshine at public.kherson.ua Fri Jan 14 15:11:44 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Fri Jan 14 07:11:52 2005 Subject: virtuals fields / ViSQL In-Reply-To: <09B4D542-6626-11D9-9687-000D93C35C66@yahoo.fr> Message-ID: On 1/14/05 2:15 PM, "Olivier" wrote: >> This will be correct >> >> string1=new vstring("string1",38, EVflag.fIndexed + >> EVflag.fIndexByWords) >> string2=new vstring("string2",38,EVflag.fIndexed,"string1") >> >> >> String1 - is indexed AND index is by words >> String2 - is indexed and it is method >> > > ok, that's works. > But I does not understand why it needs there to put "string1" indexed. > > We have a string field. For this string field, we want two indexs: > indexbyword + indexed. > > In Valentina 1: > > We put, for the string field, "indexbyword". > Then we created a second string field with for method : the first > string field + flag indexed. > Logic. :-) in fact in 1.x for method you can skip indexed! > Why in Valentina 2, we have to put the flag indexed in both fields? Because you WANT TO HAVE that both fields have INDEX. Right? And you want to have for f1 index by words., For f2 regular index. > In ViSQL, I think that there is a problem with the methods. > I have a string field which is a method, "string2". > In ViSQL, if I launch " SELECT * FROM table1 Where string2=' xxx ' : > OK, no problems, it finds. > > But if I launch " SELECT * FROM table1 ORDER BY string2 ", ViSQL > indicates "Field string2 not exist". > While this query works in my application. Strange. Look into window Structure of ViSQL. Do you see both fields ? -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From jda at his.com Fri Jan 14 08:21:25 2005 From: jda at his.com (jda) Date: Fri Jan 14 07:21:38 2005 Subject: virtuals fields / ViSQL In-Reply-To: References: Message-ID: >Because Valentina 2.0 little differ. > >If you set only flag fIndexByWords this not means yet create index. > >Valentina 2.0 now can do searches on NON-indexed fields. >So to really build index, you need explicitly say: fIndexed. > Ruslan, can you explain this a little more, please? This applies to SQL, too? So indexes are not needed for full text searching (left(), whole word, etc.)? So in V2 indexes are only needed for speed? Jon From rjb at robelko.com Fri Jan 14 14:31:23 2005 From: rjb at robelko.com (Robert Brenstein) Date: Fri Jan 14 07:35:20 2005 Subject: virtuals fields / ViSQL In-Reply-To: References: Message-ID: > > But I does not understand why it needs there to put "string1" indexed. >> >> We have a string field. For this string field, we want two indexs: >> indexbyword + indexed. >> >> In Valentina 1: >> >> We put, for the string field, "indexbyword". >> Then we created a second string field with for method : the first >> string field + flag indexed. >> Logic. > >:-) in fact in 1.x for method you can skip indexed! > > >> Why in Valentina 2, we have to put the flag indexed in both fields? > >Because you WANT TO HAVE that both fields have INDEX. Right? > >And you want to have for f1 index by words., >For f2 regular index. > In V1, indexed is set automatically when index-by-words is set, if my memory does not fail me. May be in V2, setting indexed-by-words should automatically imply indexed as well, so indexed and indexed-by-words are independent settings. I think that having the latter apply only if the former is set (that is it being a flag modifier flag) might be confusing. Robert From vidal_olivier at yahoo.fr Fri Jan 14 14:38:49 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Fri Jan 14 07:38:52 2005 Subject: virtuals fields / ViSQL In-Reply-To: References: Message-ID: <9F2354C2-6631-11D9-9687-000D93C35C66@yahoo.fr> > Look into window Structure of ViSQL. > Do you see both fields ? yes, "string1 string[38] IW" "string2=string1 string[38] Method I" ahh, sorry.. in my SELECT, I forgot to put the second " * "... now, that's works.. thank you olivier From vidal_olivier at yahoo.fr Fri Jan 14 14:41:03 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Fri Jan 14 07:40:57 2005 Subject: virtuals fields / ViSQL In-Reply-To: References: Message-ID: Le 14 janv. 05, ? 14:11, Ruslan Zasukhin a ?crit : > Because you WANT TO HAVE that both fields have INDEX. Right? > > And you want to have for f1 index by words., > For f2 regular index. ok thank you olivier From sunshine at public.kherson.ua Fri Jan 14 15:55:09 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Fri Jan 14 07:55:17 2005 Subject: virtuals fields / ViSQL In-Reply-To: Message-ID: On 1/14/05 3:21 PM, "jda" wrote: >> Because Valentina 2.0 little differ. >> >> If you set only flag fIndexByWords this not means yet create index. >> >> Valentina 2.0 now can do searches on NON-indexed fields. >> So to really build index, you need explicitly say: fIndexed. >> > > Ruslan, can you explain this a little more, please? This applies to > SQL, too? Of course > So indexes are not needed for full text searching (left(), > whole word, etc.)? Guys where you read this?! Or I explain in bad way ... :-) > So in V2 indexes are only needed for speed? Again. * FIELD can have index. * To get index you need say fIndexed. * Valentina 1.x in case of search on NOT indexed field automatically have set it to be indexed. Valentina 2.0 will not do this and will do NOT-indexed search. * Valentina 1.x also automatically build index IF you do sorting on Not indexed field. Valentina 2.0 ALSO will do this yet. * fUnique and fIndexByWords in 2.0 do not means that INDEX is built. this is just modifiers. fld.IndexByWords = true // you set modifier, this not means INDEX was built. fld.Find() // NON indexed search fld.Indexed = true // BUILDS INDEX fld.Find() // indexed search -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From jda at his.com Fri Jan 14 09:05:07 2005 From: jda at his.com (jda) Date: Fri Jan 14 08:05:15 2005 Subject: virtuals fields / ViSQL In-Reply-To: References: Message-ID: > >Guys where you read this?! >Or I explain in bad way ... :-) I'm sure you are explaining it very well. I'm just understanding it badly. :-) > > fld.IndexByWords = true > > // you set modifier, this not means INDEX was built. > > fld.Find() > > // NON indexed search > > fld.Indexed = true // BUILDS INDEX > > fld.Find() > > // indexed search > So, very simply, if I create a field like this pages = CreateTextField("pages", 128, EVFlag.fNullable + EVFlag.fIndexed + EVFlag.fIndexByWords) will be be indexed when I populate it? Or do I also have to have this line in code to actually have an index made: pages.indexed = true Jon From sunshine at public.kherson.ua Fri Jan 14 16:13:12 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Fri Jan 14 08:13:20 2005 Subject: virtuals fields / ViSQL In-Reply-To: Message-ID: On 1/14/05 4:05 PM, "jda" wrote: >> fld.IndexByWords = true >> >> // you set modifier, this not means INDEX was built. >> >> fld.Find() >> >> // NON indexed search >> >> fld.Indexed = true // BUILDS INDEX >> >> fld.Find() >> >> // indexed search >> > > So, very simply, if I create a field like this > > pages = CreateTextField("pages", 128, EVFlag.fNullable + > EVFlag.fIndexed + EVFlag.fIndexByWords) > > will be be indexed when I populate it? yes > Or do I also have to have this line in code to actually have an index made: > > pages.indexed = true No, this is old style. Now we have FLAGS in the CreateField() functions. -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From sunshine at public.kherson.ua Fri Jan 14 16:14:44 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Fri Jan 14 08:14:48 2005 Subject: virtuals fields / ViSQL In-Reply-To: Message-ID: On 1/14/05 3:31 PM, "Robert Brenstein" wrote: >> >>> Why in Valentina 2, we have to put the flag indexed in both fields? >> >> Because you WANT TO HAVE that both fields have INDEX. Right? >> >> And you want to have for f1 index by words., >> For f2 regular index. >> > > In V1, indexed is set automatically when index-by-words is set, if my > memory does not fail me. Right. In 2.0 this is CHANGED. > May be in V2, setting indexed-by-words > should automatically imply indexed as well, so indexed and > indexed-by-words are independent settings. > I think that having the > latter apply only if the former is set (that is it being a flag > modifier flag) might be confusing. -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From rjb at robelko.com Fri Jan 14 15:39:03 2005 From: rjb at robelko.com (Robert Brenstein) Date: Fri Jan 14 08:54:39 2005 Subject: virtuals fields / ViSQL In-Reply-To: References: Message-ID: >On 1/14/05 3:31 PM, "Robert Brenstein" wrote: > >>> >>>> Why in Valentina 2, we have to put the flag indexed in both fields? >>> >>> Because you WANT TO HAVE that both fields have INDEX. Right? >>> >>> And you want to have for f1 index by words., >>> For f2 regular index. >>> >> > > In V1, indexed is set automatically when index-by-words is set, if my >> memory does not fail me. > >Right. In 2.0 this is CHANGED. > >> May be in V2, setting indexed-by-words >> should automatically imply indexed as well, so indexed and >> indexed-by-words are independent settings. > >> I think that having the >> latter apply only if the former is set (that is it being a flag > > modifier flag) might be confusing. But having index-by-words set but being not in effect IS confusing/misleading. This is why I am suggesting that it becomes an independent flag: one flag (index) controls normal indexing and another flag (index-by-words) control word indexing. When both are set, the latter (higher functionality) wins. I think this would make things clearer for developers. Robert From jordan at digi-net.com Fri Jan 14 10:00:10 2005 From: jordan at digi-net.com (Ken Jordan) Date: Fri Jan 14 08:57:38 2005 Subject: V2 compatibility In-Reply-To: References: Message-ID: <41E7DE7A.3010904@digi-net.com> Ruslan Zasukhin wrote: >On 1/13/05 11:20 PM, "Ken Jordan" wrote: > > > >>Is V2 V4RB compatible with Windows? If so, what is the installation >>procedure? >> >> > >We here work with V4RB windows, but we did not upload it yet. > >You work with REALbasic on PC ? > > I am using REALbasic on PC and Mac. I'm doing most of the coding and testing on the PC right now because the only Mac that I have is a PowerBook (laptop) and I find the desktop much easier and more comfortable to use. Is there any chance of getting the Windows version? :-) Please :-) -------------- next part -------------- An HTML attachment was scrubbed... URL: http://lists.macserve.net/pipermail/valentina-beta/attachments/20050114/e443a515/attachment.html From vidal_olivier at yahoo.fr Fri Jan 14 16:08:51 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Fri Jan 14 09:09:01 2005 Subject: virtuals fields / ViSQL In-Reply-To: References: Message-ID: <335ECF2A-663E-11D9-9687-000D93C35C66@yahoo.fr> I agree. olivier Le 14 janv. 05, ? 15:39, Robert Brenstein a ?crit : > But having index-by-words set but being not in effect IS > confusing/misleading. > > This is why I am suggesting that it becomes an independent flag: one > flag (index) controls normal indexing and another flag > (index-by-words) control word indexing. When both are set, the latter > (higher functionality) wins. > > I think this would make things clearer for developers. From sunshine at public.kherson.ua Fri Jan 14 17:34:35 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Fri Jan 14 09:34:41 2005 Subject: V2 compatibility In-Reply-To: <41E7DE7A.3010904@digi-net.com> Message-ID: On 1/14/05 5:00 PM, "Ken Jordan" wrote: > I am using REALbasic on PC and Mac. I'm doing most of the coding and testing > on the PC right now because the only Mac that I have is a PowerBook (laptop) > and I find the desktop much easier and more comfortable to use. > > Is there any chance of getting the Windows version? :-) > > Please :-) Ok, next beta we will make for Windows also -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From sunshine at public.kherson.ua Fri Jan 14 17:40:06 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Fri Jan 14 09:40:09 2005 Subject: virtuals fields / ViSQL In-Reply-To: <335ECF2A-663E-11D9-9687-000D93C35C66@yahoo.fr> Message-ID: On 1/14/05 5:08 PM, "Olivier" wrote: > I agree. > > olivier > > Le 14 janv. 05, ? 15:39, Robert Brenstein a ?crit : > >> But having index-by-words set but being not in effect IS >> confusing/misleading. >> >> This is why I am suggesting that it becomes an independent flag: one >> flag (index) controls normal indexing and another flag >> (index-by-words) control word indexing. When both are set, the latter >> (higher functionality) wins. >> >> I think this would make things clearer for developers. Think about this behavior. I have field (Indexed + ByWords) Now I want set index OFF for some time, then return it back. So I do fld.Indexed = false. ... fld.Indexed = true. Now in 2.0, the first line DO NOT remove flag IndexByWords (1.x did). So the second line is able __correctly__ return back THE SAME index kind. Even do not have idea how todo this in 1.x. * You need self get and save flag IndexByWords ? * and flag Unique? * and for each field if you do this for several fields. ---------------- I want underline, that in 2.0 we have REMOVE several such "automatic things" We think this is more correct way. Because if developer DID NOT ask build index for a field, Then search must be NOT indexed. -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From vidal_olivier at yahoo.fr Fri Jan 14 17:24:55 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Fri Jan 14 10:24:57 2005 Subject: db created with classes or API ? Message-ID: Hi, My project was created with classes. I study the possibility of modifying it to create and manage the DB with the API. That will bring me more flexibility? olivier From vidal_olivier at yahoo.fr Fri Jan 14 18:09:24 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Fri Jan 14 11:09:33 2005 Subject: precisions on set, bitset and arrayset Message-ID: <0A39D87A-664F-11D9-9687-000D93C35C66@yahoo.fr> Hi, What is that an arrayset ? What is the difference between a set and a bitset ? A set is thus a record selection. A sort of temporary table ? Does It take a lot of memory ? thank you olivier From jda at his.com Fri Jan 14 12:13:56 2005 From: jda at his.com (jda) Date: Fri Jan 14 11:14:07 2005 Subject: virtuals fields / ViSQL In-Reply-To: References: Message-ID: >>Valentina 2.0 now can do searches on NON-indexed fields. >>So to really build index, you need explicitly say: fIndexed. >> I've been looking at the SQL docs but don't see an obvious answer to this question: what is the SQL way to search an *unindexed* text field for a 1. word begining with "aaa" and a 2. word "aaa" without using REGEX? Thanks, Jon From sunshine at public.kherson.ua Fri Jan 14 20:12:16 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Fri Jan 14 12:12:20 2005 Subject: virtuals fields / ViSQL In-Reply-To: Message-ID: On 1/14/05 7:13 PM, "jda" wrote: >>> Valentina 2.0 now can do searches on NON-indexed fields. >>> So to really build index, you need explicitly say: fIndexed. >>> > > I've been looking at the SQL docs but don't see an obvious answer to > this question: what is the SQL way to search an *unindexed* text > field for a > > 1. word begining with "aaa" But no way Jon! You need to have index BY WORDS to be search WORD START WITH. Otherwise you will search LINE START WITH > and a > > 2. word "aaa" > > without using REGEX? -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From sunshine at public.kherson.ua Fri Jan 14 20:37:43 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Fri Jan 14 12:37:49 2005 Subject: db created with classes or API ? In-Reply-To: Message-ID: On 1/14/05 6:24 PM, "Olivier" wrote: > Hi, > > My project was created with classes. > I study the possibility of modifying it to create and manage the DB > with the API. > That will bring me more flexibility? But classes is an API way. Classes is the same as API, just you mirror structure of db on classes. Classes are good IF you develop app which have quite fixed structure of tables. Classes give more comfortable and little faster way of accesst to Tables/Fields. Because you ALREADY have pointers to them. In the pure API way you need each time do db.Table("Person") But API way is more flexible, and it is the only way for such apps as viSQL, which do not know structure of db which they will open. -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From sunshine at public.kherson.ua Fri Jan 14 20:44:07 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Fri Jan 14 12:44:08 2005 Subject: precisions on set, bitset and arrayset In-Reply-To: <0A39D87A-664F-11D9-9687-000D93C35C66@yahoo.fr> Message-ID: On 1/14/05 7:09 PM, "Olivier" wrote: > Hi, > > What is that an arrayset ? > What is the difference between a set and a bitset ? ArraySet -- array of integers. BitSet -- set of bits. You use this to get/keep SELECTION OF records. ArraySet is good for small selections. BitSet is godo for big selections. ** Let you have million records table. And Find() returns 100 records. BitSet always million bits. = 128 KB AraySet 100 * 4 = 400 bytes ** Now let Find() return 300,000 records. BitSet always million bits. = 128 KB AraySet 300,000 * 4 = 1.2 MB For all records BitSet always million bits. = 128 KB AraySet 1M * 4 = 4 MB Also note, that SORT() always return ArraySet, Because bitset cannot keep sort order. > A set is thus a record selection. A sort of temporary table ? :-) Not exactly. You have ONE table. You can have many selections for it. You can have many threads that will use that selections. Selection this is RAM only array that keep RecID of records. BitSet also indirectly keep RecIDs. > Does It take a lot of memory ? This is minimal we can do. Cursor for example use BitSets and ArraySets also. Just you did not know about thus before. :-) -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From vidal_olivier at yahoo.fr Fri Jan 14 20:00:16 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Fri Jan 14 13:01:07 2005 Subject: db created with classes or API ? In-Reply-To: References: Message-ID: <87260152-665E-11D9-9687-000D93C35C66@yahoo.fr> The structure of the DB should remain fixed. But during updates, I have to can add easily tables or fields. If I can make it EASILY (for customer especially) with the classes, perfect ! With the classes, I can create " binary links " as with the API, right? Thank you very much for all this precision. olivier > But classes is an API way. > > Classes is the same as API, just you mirror structure of db on classes. > > Classes are good IF you develop app which have quite fixed structure of > tables. > > Classes give more comfortable and little faster way of accesst to > Tables/Fields. Because you ALREADY have pointers to them. > In the pure API way you need each time do > > db.Table("Person") > > > But API way is more flexible, and it is the only way for such apps as > viSQL, > which do not know structure of db which they will open. From vidal_olivier at yahoo.fr Fri Jan 14 20:06:46 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Fri Jan 14 13:06:42 2005 Subject: precisions on set, bitset and arrayset In-Reply-To: References: Message-ID: <6F5F0502-665F-11D9-9687-000D93C35C66@yahoo.fr> OK, it is very clear! I decided to replace in my project all the SQL cursors by the API, and obectPtrs by binary links. The API of Valentina is really attractive! The 'sets' look like very powerful... Thank you very much Ruslan good evening ! olivier > ArraySet -- array of integers. > BitSet -- set of bits. > > You use this to get/keep SELECTION OF records. > > > ArraySet is good for small selections. > BitSet is godo for big selections. > > ** Let you have million records table. > And Find() returns 100 records. > > BitSet always million bits. = 128 KB > AraySet 100 * 4 = 400 bytes > > ** Now let Find() return 300,000 records. > > BitSet always million bits. = 128 KB > AraySet 300,000 * 4 = 1.2 MB > > For all records > > BitSet always million bits. = 128 KB > AraySet 1M * 4 = 4 MB > > > Also note, that SORT() always return ArraySet, > Because bitset cannot keep sort order. > >> A set is thus a record selection. A sort of temporary table ? > > :-) > > Not exactly. > > You have ONE table. You can have many selections for it. > You can have many threads that will use that selections. > > Selection this is RAM only array that keep RecID of records. > BitSet also indirectly keep RecIDs. > > >> Does It take a lot of memory ? > > This is minimal we can do. > > Cursor for example use BitSets and ArraySets also. > Just you did not know about thus before. :-) > > > -- > Best regards, > Ruslan Zasukhin [ I feel the need...the need for speed ] > ------------------------------------------------------------- > e-mail: ruslan@paradigmasoft.com > web: http://www.paradigmasoft.com > > To subscribe to the Valentina mail list go to: > http://lists.macserve.net/mailman/listinfo/valentina > ------------------------------------------------------------- > > > _______________________________________________ > Valentina-beta mailing list > Valentina-beta@lists.macserve.net > http://lists.macserve.net/mailman/listinfo/valentina-beta > From sunshine at public.kherson.ua Fri Jan 14 21:16:37 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Fri Jan 14 13:16:41 2005 Subject: precisions on set, bitset and arrayset In-Reply-To: <6F5F0502-665F-11D9-9687-000D93C35C66@yahoo.fr> Message-ID: On 1/14/05 9:06 PM, "Olivier" wrote: > OK, it is very clear! > > I decided to replace in my project all the SQL cursors by the API, and > obectPtrs by binary links. :-) so we expect stress - testing for binary links! This is very good!!! > The API of Valentina is really attractive! The 'sets' look like very > powerful... Yes a lots of power is hidden here. If you will fill something is missing to you, let us know. We have a lots more things in the c++ kernel, Just we in huge rush to open, test, example and document what we already have open to RB developers. > Thank you very much Ruslan Enjoy. -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From sunshine at public.kherson.ua Fri Jan 14 21:17:41 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Fri Jan 14 13:17:42 2005 Subject: db created with classes or API ? In-Reply-To: <87260152-665E-11D9-9687-000D93C35C66@yahoo.fr> Message-ID: On 1/14/05 9:00 PM, "Olivier" wrote: > The structure of the DB should remain fixed. > But during updates, I have to can add easily tables or fields. > If I can make it EASILY (for customer especially) with the classes, > perfect ! Yes, your users do not see your classes. > With the classes, I can create " binary links " as with the API, right? Yes. Classes_way folder have example of BinaryLink class. > Thank you very much for all this precision. -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From jda at his.com Fri Jan 14 14:52:37 2005 From: jda at his.com (jda) Date: Fri Jan 14 13:52:47 2005 Subject: virtuals fields / ViSQL In-Reply-To: References: Message-ID: >On 1/14/05 7:13 PM, "jda" wrote: > >>>> Valentina 2.0 now can do searches on NON-indexed fields. >>>> So to really build index, you need explicitly say: fIndexed. >>>> >> >> I've been looking at the SQL docs but don't see an obvious answer to >> this question: what is the SQL way to search an *unindexed* text >> field for a >> >> 1. word begining with "aaa" > >But no way Jon! > >You need to have index BY WORDS to be search WORD START WITH. > >Otherwise you will search LINE START WITH > OK, that's what I thought (but was confused a bit by your previous statement -- I told you I didn't understand too well). Thanks for clearing that up. Jon From vidal_olivier at yahoo.fr Sat Jan 15 19:04:06 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Sat Jan 15 12:04:06 2005 Subject: update in set Message-ID: Hi, I have one set. I want to UPDATE, to add or to delete a record. How can it be done? I do not see methods (as in cursors) in the sets. It is necessary to use directly the API methods (table1.field ("string1") .setstring (" xxxx "), table1.updaterecord () ) ? But in that case the set save the modifications at once? Which is the difference between setIterator.value and arraySet.ItemAt ? thank you olivier From sunshine at public.kherson.ua Sat Jan 15 22:08:51 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Sat Jan 15 14:09:00 2005 Subject: update in set In-Reply-To: Message-ID: On 1/15/05 8:04 PM, "Olivier" wrote: > Hi, > > I have one set. > I want to UPDATE, to add or to delete a record. > How can it be done? > I do not see methods (as in cursors) in the sets. right > It is necessary to use directly the API methods (table1.field > ("string1") .setstring (" xxxx "), table1.updaterecord () ) ? Right! > But in that case the set save the modifications at once? Oliver, set this is just AN ARRAY OF Integer numbers! This is very very simple abstraction. You need get an Iteger value from Set, this is RecID. So dim recid as Integer for I = 1 to count recId = set.ItemIt( I ) Table.RecID = recID // Now you can do Update or Delete. Or read values of this record end if > Which is the difference between setIterator.value and arraySet.ItemAt ? No difference. * Just ArraySet.ItemAt() allow you work with ArraySet without iterator. * But for BitSet there is no itemAt(), and to get item of bitset you MUST to work with Iterator. * Iterator is an abstraction that allow you work with both, BitSet and ArraySet and the same way. -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From bibiko at eva.mpg.de Mon Jan 17 01:47:13 2005 From: bibiko at eva.mpg.de (bibiko@eva.mpg.de) Date: Sun Jan 16 18:47:24 2005 Subject: Search for accented characters In-Reply-To: <41E7DE7A.3010904@digi-net.com> References: <41E7DE7A.3010904@digi-net.com> Message-ID: <1105922833.41eb0b114a8ae@email.eva.mpg.de> Dear developers, are there any thoughts about an easy way for searching accented characters? I don't know if there is an implementation yet. Problem: One record has stored the string 'Wich?'. If I'm looking for 'Wichi' then I don't get the match for 'Wich?'. You have to write something like SELECT field FROM tbl WHERE field LIKE 'Wich[i|?]' OK. But it would be a very nice feature to implement a keyword like 'no_accent' for searching automatically this kind of occurances :) The problem is if a user has its own keyboard and s/he is looking for a word which contains foreign characters with diacritics and you are not able to type this character, well ... I know we are talking about Unicode and there a lot of such stuff but nevertheless it would be a further argument to buy Valentina ;) The following is only a thought: Such character classes could generated via the internal Unicode names. If you are thinking further you can implement other class of Cyrillic, Greek, Japanese, Tamil etc. Then you can search for the Russian word for 'language' via ASCII 'jazyk'. Of course that would mean that there is a standard for transliteration of such writing systems with Roman letters but ... I write this because this is my special area of interest and a part of my daily work. If there is an interest maybe I could help to generate such classes. Thanks a lot Hans From vidal_olivier at yahoo.fr Mon Jan 17 08:22:22 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Mon Jan 17 01:22:34 2005 Subject: type casting in API Message-ID: <879BDF25-6858-11D9-8C55-000D93C35C66@yahoo.fr> Hi Ruslan and list, I noticed in the API examples that fields are not type casting as in cursors. Then, it is not better to put ? : table1.stringfield("string1").getsring On the place of table1.field("string1").getsring. thank you olivier From sunshine at public.kherson.ua Mon Jan 17 09:29:06 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Mon Jan 17 01:29:11 2005 Subject: type casting in API In-Reply-To: <879BDF25-6858-11D9-8C55-000D93C35C66@yahoo.fr> Message-ID: On 1/17/05 9:22 AM, "Olivier" wrote: > Hi Ruslan and list, > > I noticed in the API examples that fields are not type casting as in > cursors. I will fix this right now! > Then, it is not better to put ? : > > table1.stringfield("string1").getsring > > On the place of > > table1.field("string1").getsring. -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From sunshine at public.kherson.ua Mon Jan 17 09:33:40 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Mon Jan 17 01:33:45 2005 Subject: Search for accented characters In-Reply-To: <1105922833.41eb0b114a8ae@email.eva.mpg.de> Message-ID: On 1/17/05 2:47 AM, "bibiko@eva.mpg.de" wrote: Hi Hans, > Dear developers, > > are there any thoughts about an easy way for searching accented characters? > > I don't know if there is an implementation yet. > > Problem: > > One record has stored the string 'Wich?'. > > If I'm looking for 'Wichi' then I don't get the match for 'Wich?'. > You have to write something like > SELECT field FROM tbl WHERE field LIKE 'Wich[i|?]' > > OK. > > But it would be a very nice feature to implement a keyword like 'no_accent' > for searching automatically this kind of occurances :) We talk about Valentina 2.0 ? Then solution is the next: * IF you want that your whole db consider accent chars as regular chars, Then you can create db and set its collation Attribute STRENGTH = kSecondary. db.ColAttribute( EVColAttribute.kStrength ) = EVColAttrValue.kSecondary We have example on this in 2.0. You can send this also for separate table and a field. You can have fields/methods with different settings. So if you set this option, then search WHERE fld = 'Wich?' Will find also 'Wich?'. > The problem is if a user has its own keyboard and s/he is looking for a word > which contains foreign characters with diacritics and you are not able to type > this character, well ... > > I know we are talking about Unicode and there a lot of such stuff but > nevertheless it would be a further argument to buy Valentina ;) > > The following is only a thought: > > Such character classes could generated via the internal Unicode names. > If you are thinking further you can implement other class of Cyrillic, Greek, > Japanese, Tamil etc. Then you can search for the Russian word for 'language' > via > ASCII 'jazyk'. Of course that would mean that there is a standard for > transliteration of such writing systems with Roman letters but ... > > I write this because this is my special area of interest and a part of my > daily > work. If there is an interest maybe I could help to generate such classes. -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From sunshine at public.kherson.ua Mon Jan 17 09:51:58 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Mon Jan 17 01:52:13 2005 Subject: type casting in API In-Reply-To: <879BDF25-6858-11D9-8C55-000D93C35C66@yahoo.fr> Message-ID: On 1/17/05 9:22 AM, "Olivier" wrote: > Hi Ruslan and list, > > I noticed in the API examples that fields are not type casting as in > cursors. Oliver, Type casting methods ALREADY present in the Vtable class > Then, it is not better to put ? : > > table1.stringfield("string1").getsring > > On the place of > > table1.field("string1").getsring. You can use any way you prefer or which is faster for your task! -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From rjb at robelko.com Mon Jan 17 12:52:33 2005 From: rjb at robelko.com (Robert Brenstein) Date: Mon Jan 17 06:05:20 2005 Subject: virtuals fields / ViSQL In-Reply-To: References: Message-ID: >On 1/14/05 5:08 PM, "Olivier" wrote: > >> I agree. >> >> olivier >> >> Le 14 janv. 05, ? 15:39, Robert Brenstein a ?crit : >> >>> But having index-by-words set but being not in effect IS >>> confusing/misleading. >>> >>> This is why I am suggesting that it becomes an independent flag: one >>> flag (index) controls normal indexing and another flag >>> (index-by-words) control word indexing. When both are set, the latter >>> (higher functionality) wins. >>> >>> I think this would make things clearer for developers. > >Think about this behavior. > >I have field (Indexed + ByWords) > >Now I want set index OFF for some time, then return it back. > >So I do > > fld.Indexed = false. > > ... > > fld.Indexed = true. > >Now in 2.0, the first line DO NOT remove flag IndexByWords (1.x did). >So the second line is able __correctly__ return back THE SAME index kind. Then in 2, these lines would tackle both flags as needed. At low level,I mean in kernel, you should probably have a separate flag that tells kernel whether to index or not, so kernel does not have to touch what users set. >Even do not have idea how todo this in 1.x. >* You need self get and save flag IndexByWords ? >* and flag Unique? >* and for each field if you do this for several fields. V1.x should stay as it is in my opinion. There is a discontinuity from going from 1 to 2 anyway, so I see no problem. You should really use this opportunity to clean up and improve all APIs. Robert From sunshine at public.kherson.ua Mon Jan 17 15:03:08 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Mon Jan 17 07:03:16 2005 Subject: virtuals fields / ViSQL In-Reply-To: Message-ID: On 1/17/05 1:52 PM, "Robert Brenstein" wrote: >> Even do not have idea how todo this in 1.x. >> * You need self get and save flag IndexByWords ? >> * and flag Unique? >> * and for each field if you do this for several fields. > > V1.x should stay as it is in my opinion. There is > a discontinuity from going from 1 to 2 anyway, so > I see no problem. You should really use this > opportunity to clean up and improve all APIs. Wait, Robert! Of course we NOT going change behavior 1.x We change behavior for 2.0 -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From vidal_olivier at yahoo.fr Mon Jan 17 14:17:11 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Mon Jan 17 07:17:25 2005 Subject: local pointer and auto-completion Message-ID: <18FD663A-688A-11D9-8C55-000D93C35C66@yahoo.fr> Hi Ruslan, In certain examples, you indicate that it is preferable (because faster) to create a variable or a local property for the used table or the DB. But in that case (VarTable1.field ("NameString") .value), we have no more the auto-completion of RB. While if we use directly "Wind1.table1. NameString.value", we have the auto-completion of RB. It is, sometimes, thus simpler. To have a local pointer of the table is it really faster? thank you olivier From sunshine at public.kherson.ua Mon Jan 17 15:22:47 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Mon Jan 17 07:23:02 2005 Subject: local pointer and auto-completion In-Reply-To: <18FD663A-688A-11D9-8C55-000D93C35C66@yahoo.fr> Message-ID: On 1/17/05 3:17 PM, "Olivier" wrote: > Hi Ruslan, > > In certain examples, you indicate that it is preferable (because > faster) to create a variable or a local property for the used table or > the DB. > > But in that case (VarTable1.field ("NameString") .value), we have no > more the auto-completion of RB. > While if we use directly "Wind1.table1. NameString.value", we have the > auto-completion of RB. It is, sometimes, thus simpler. > To have a local pointer of the table is it really faster? It is faster, but use it only for big loops. -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From rjb at robelko.com Mon Jan 17 14:06:44 2005 From: rjb at robelko.com (Robert Brenstein) Date: Mon Jan 17 07:25:13 2005 Subject: virtuals fields / ViSQL In-Reply-To: References: Message-ID: >On 1/17/05 1:52 PM, "Robert Brenstein" wrote: > >>> Even do not have idea how todo this in 1.x. >>> * You need self get and save flag IndexByWords ? >>> * and flag Unique? >>> * and for each field if you do this for several fields. >> >> V1.x should stay as it is in my opinion. There is >> a discontinuity from going from 1 to 2 anyway, so >> I see no problem. You should really use this >> opportunity to clean up and improve all APIs. > >Wait, Robert! > > >Of course we NOT going change behavior 1.x > >We change behavior for 2.0 Then I did understand what you wrote above, Ruslan. From sunshine at public.kherson.ua Mon Jan 17 16:02:36 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Mon Jan 17 08:02:46 2005 Subject: virtuals fields / ViSQL In-Reply-To: Message-ID: On 1/17/05 3:06 PM, "Robert Brenstein" wrote: >>>> Even do not have idea how todo this in 1.x. >>>> * You need self get and save flag IndexByWords ? >>>> * and flag Unique? >>>> * and for each field if you do this for several fields. >>> >>> V1.x should stay as it is in my opinion. There is >>> a discontinuity from going from 1 to 2 anyway, so >>> I see no problem. You should really use this >>> opportunity to clean up and improve all APIs. >> >> Wait, Robert! >> >> >> Of course we NOT going change behavior 1.x >> >> We change behavior for 2.0 > > Then I did understand what you wrote above, Ruslan. I mean that 2.0 will work in DIFFERENT way to 1.x. -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From rjb at robelko.com Mon Jan 17 15:30:27 2005 From: rjb at robelko.com (Robert Brenstein) Date: Mon Jan 17 08:44:19 2005 Subject: virtuals fields / ViSQL In-Reply-To: References: Message-ID: >On 1/17/05 3:06 PM, "Robert Brenstein" wrote: > >>>>> Even do not have idea how todo this in 1.x. >>>>> * You need self get and save flag IndexByWords ? >>>>> * and flag Unique? >>>>> * and for each field if you do this for several fields. >>>> >>>> V1.x should stay as it is in my opinion. There is >>>> a discontinuity from going from 1 to 2 anyway, so >>>> I see no problem. You should really use this >>>> opportunity to clean up and improve all APIs. >>> >>> Wait, Robert! >>> >>> >>> Of course we NOT going change behavior 1.x >>> >>> We change behavior for 2.0 >> >> Then I did understand what you wrote above, Ruslan. > >I mean that 2.0 will work in DIFFERENT way to 1.x. > Okay, so we agree. From jordan at digi-net.com Mon Jan 17 10:17:24 2005 From: jordan at digi-net.com (Ken Jordan) Date: Mon Jan 17 09:14:59 2005 Subject: V2 beta install Message-ID: <41EBD704.2020003@digi-net.com> I'm having a little trouble getting V4RB V2 to work - wondering what I might be doing wrong. Trying to set up on Powerbook with OS X, RB 5.5.4. Copied V4RB to plug-ins folder of RB. Didn't find a "/Library/CFMSupport" folder, so created one. Copied VComponenets folder into this folder. Copied Kernel_Carbon_Final.shlb and vclient_carbon_cw.shlb from VEngine folder to "/Library/CFMSupport/VComponents" folder. RB would no longer start. Saw that there was an "Application Support" folder, thought I'd try that. Moved VComponents folder w/VEngine files to "Application Support". Now RB would start, but Valentina classes are not found. Thought maybe the space before "Support" was important - created a "/Library/CFM Support" folder, moved VComponents folder there. Again, RB starts, but can't find Valentina classes. What am I missing here?? Ken From sunshine at public.kherson.ua Mon Jan 17 17:19:09 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Mon Jan 17 09:19:15 2005 Subject: V2 beta install In-Reply-To: <41EBD704.2020003@digi-net.com> Message-ID: On 1/17/05 5:17 PM, "Ken Jordan" wrote: Hi Ken, > I'm having a little trouble getting V4RB V2 to work - wondering what I > might be doing wrong. > > Trying to set up on Powerbook with OS X, RB 5.5.4. > > Copied V4RB to plug-ins folder of RB. ok > Didn't find a "/Library/CFMSupport" folder, so created one. Copied > VComponenets folder into this folder. ok > Copied Kernel_Carbon_Final.shlb and vclient_carbon_cw.shlb from VEngine > folder to "/Library/CFMSupport/VComponents" folder. > > RB would no longer start. But Vengine folder also have "resources" folder. You also should copy it into VComponents -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From jordan at digi-net.com Mon Jan 17 10:27:41 2005 From: jordan at digi-net.com (Ken Jordan) Date: Mon Jan 17 09:24:58 2005 Subject: V2 beta install In-Reply-To: References: Message-ID: <41EBD96D.1060505@digi-net.com> Ruslan Zasukhin wrote: >On 1/17/05 5:17 PM, "Ken Jordan" wrote: > >Hi Ken, > > > >>I'm having a little trouble getting V4RB V2 to work - wondering what I >>might be doing wrong. >> >>Trying to set up on Powerbook with OS X, RB 5.5.4. >> >>Copied V4RB to plug-ins folder of RB. >> >> > >ok > > > >>Didn't find a "/Library/CFMSupport" folder, so created one. Copied >>VComponenets folder into this folder. >> >> > >ok > > > >>Copied Kernel_Carbon_Final.shlb and vclient_carbon_cw.shlb from VEngine >>folder to "/Library/CFMSupport/VComponents" folder. >> >>RB would no longer start. >> >> > >But Vengine folder also have "resources" folder. >You also should copy it into VComponents > > > There is already a "resources" folder in VComponents that appears to have the same contents as the one in VEngine. Replace it? From sunshine at public.kherson.ua Mon Jan 17 17:26:58 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Mon Jan 17 09:27:05 2005 Subject: V2 beta install In-Reply-To: <41EBD96D.1060505@digi-net.com> Message-ID: On 1/17/05 5:27 PM, "Ken Jordan" wrote: >> But Vengine folder also have "resources" folder. >> You also should copy it into VComponents >> >> >> > There is already a "resources" folder in VComponents that appears to > have the same contents as the one in VEngine. Replace it? No, that is enough -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From sunshine at public.kherson.ua Mon Jan 17 17:29:17 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Mon Jan 17 09:29:20 2005 Subject: V2 beta install In-Reply-To: <41EBD704.2020003@digi-net.com> Message-ID: On 1/17/05 5:17 PM, "Ken Jordan" wrote: > I'm having a little trouble getting V4RB V2 to work - wondering what I > might be doing wrong. > > Trying to set up on Powerbook with OS X, RB 5.5.4. > > Copied V4RB to plug-ins folder of RB. > Didn't find a "/Library/CFMSupport" folder, so created one. Copied > VComponenets folder into this folder. > Copied Kernel_Carbon_Final.shlb and vclient_carbon_cw.shlb from VEngine > folder to "/Library/CFMSupport/VComponents" folder. > > RB would no longer start. > > Saw that there was an "Application Support" folder, thought I'd try > that. Moved VComponents folder w/VEngine files to "Application Support". > Now RB would start, but Valentina classes are not found. No, do not use "Application Support" > Thought maybe the space before "Support" was important - created a > "/Library/CFM Support" folder, moved VComponents folder there. Again, RB > starts, but can't find Valentina classes. > > What am I missing here?? 1) do not put into other folders. Everything must be in the /Library/CFMSupport/VComponents 2) make sure that you have made /Library/CFMSupport In the ROOT, but not in the HOME -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From jordan at digi-net.com Mon Jan 17 10:39:26 2005 From: jordan at digi-net.com (Ken Jordan) Date: Mon Jan 17 09:36:46 2005 Subject: V2 beta install In-Reply-To: References: Message-ID: <41EBDC2E.3040409@digi-net.com> Ruslan Zasukhin wrote: >On 1/17/05 5:17 PM, "Ken Jordan" wrote: > > > >>I'm having a little trouble getting V4RB V2 to work - wondering what I >>might be doing wrong. >> >>Trying to set up on Powerbook with OS X, RB 5.5.4. >> >>Copied V4RB to plug-ins folder of RB. >>Didn't find a "/Library/CFMSupport" folder, so created one. Copied >>VComponenets folder into this folder. >>Copied Kernel_Carbon_Final.shlb and vclient_carbon_cw.shlb from VEngine >>folder to "/Library/CFMSupport/VComponents" folder. >> >>RB would no longer start. >> >>Saw that there was an "Application Support" folder, thought I'd try >>that. Moved VComponents folder w/VEngine files to "Application Support". >>Now RB would start, but Valentina classes are not found. >> >> > >No, do not use "Application Support" > > > >>Thought maybe the space before "Support" was important - created a >>"/Library/CFM Support" folder, moved VComponents folder there. Again, RB >>starts, but can't find Valentina classes. >> >>What am I missing here?? >> >> > >1) do not put into other folders. >Everything must be in the /Library/CFMSupport/VComponents > >2) make sure that you have made /Library/CFMSupport >In the ROOT, but not in the HOME > > > That's it! Thanks again! From jda at his.com Tue Jan 18 08:03:53 2005 From: jda at his.com (jda) Date: Tue Jan 18 07:04:13 2005 Subject: Dispensable libraries Message-ID: Hi Ruslan, I'm using V4RB for a standalone application in OS X only. By trial and error I have found that I can launch my app without these two libraries installed: VDK_PPC.shlb and vclient_carbon_cw.shlb. My question is, is it safe to assume I will not need these and can ship my app without them? Thanks, Jon From sunshine at public.kherson.ua Tue Jan 18 15:15:33 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Tue Jan 18 07:15:39 2005 Subject: Dispensable libraries In-Reply-To: Message-ID: On 1/18/05 3:03 PM, "jda" wrote: Hi Jon, > I'm using V4RB for a standalone application in OS X only. By trial > and error I have found that I can launch my app without these two > libraries installed: VDK_PPC.shlb and vclient_carbon_cw.shlb. > > My question is, is it safe to assume I will not need these and can > ship my app without them? VDK_PPC this is 1.x engine. So YOU will need it if you want convert Client -- yes is not required for LOCAL db -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From jda at his.com Tue Jan 18 08:30:54 2005 From: jda at his.com (jda) Date: Tue Jan 18 07:31:05 2005 Subject: Dispensable libraries In-Reply-To: References: Message-ID: > > >VDK_PPC this is 1.x engine. So YOU will need it if you want convert > >Client -- yes is not required for LOCAL db > Thanks, I understand. Hm, if we can't use a folder to store all these files, might it be possible to combine some of the libraries into one? There are 4 different icu libraries, for example. Jon From sunshine at public.kherson.ua Tue Jan 18 15:52:26 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Tue Jan 18 07:52:37 2005 Subject: Dispensable libraries In-Reply-To: Message-ID: On 1/18/05 3:30 PM, "jda" wrote: >> VDK_PPC this is 1.x engine. So YOU will need it if you want convert >> >> Client -- yes is not required for LOCAL db >> > > Thanks, I understand. > > Hm, if we can't use a folder to store all these files, might it be > possible to combine some of the libraries into one? There are 4 > different icu libraries, for example. No, ICU we can't change. This is IBM -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From jda at his.com Tue Jan 18 09:14:47 2005 From: jda at his.com (jda) Date: Tue Jan 18 08:15:04 2005 Subject: Dispensable libraries In-Reply-To: References: Message-ID: > > Hm, if we can't use a folder to store all these files, might it be >> possible to combine some of the libraries into one? There are 4 >> different icu libraries, for example. > >No, ICU we can't change. This is IBM > > You mean that you don't have the source? This is the way they distribut it? Ruslan, lots of other apps in OS X use the ICU library, and Valentina is the only one that has to include all of the libraries and .dat file. Are you sure you can't simply link to the libraries included with OS X (I'm pretty sure that RB does, for example)? This alone adds most of the overhead of a V2 distribution. Jon From sunshine at public.kherson.ua Tue Jan 18 16:51:34 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Tue Jan 18 08:51:38 2005 Subject: Dispensable libraries In-Reply-To: Message-ID: On 1/18/05 4:14 PM, "jda" wrote: >>> Hm, if we can't use a folder to store all these files, might it be >>> possible to combine some of the libraries into one? There are 4 >>> different icu libraries, for example. >> >> No, ICU we can't change. This is IBM >> >> > > You mean that you don't have the source? This is the way they distribut it? I have, but projects are huge, so it is very bad idea try to change them. Each new release again and again... > Ruslan, lots of other apps in OS X use the ICU library, and Valentina > is the only one that has to include all of the libraries and .dat > file. Why you think so?! I very doubt that somebody use ICU libs from Apple. I have discuss this with the main engineer of Apple that integrate ICU to OS X. He have told that 1) Apple do not open now ICU for developers access. they work with IBM to improve ICU to allow this. 2) Apple's libs are in MACHO format, so I cannot use them in V4RB Carbon, and V4MD Carobn. In any case. 3) even to use them it needs use NOT C++ API of ICU but C only. not very pleasant. > Are you sure you can't simply link to the libraries included > with OS X (I'm pretty sure that RB does, for example)? This alone > adds most of the overhead of a V2 distribution. I know. It is possible reduce overhead only with .DAT file. They say it is possible to make it not 8MB but 3.5MB If remove rare languages, and other not needed resources. You can do this self using ICU command line utils. But I have not yet play with them. -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From valentina-list at vermontsoftworks.com Tue Jan 18 17:05:58 2005 From: valentina-list at vermontsoftworks.com (Erik Mueller-Harder) Date: Tue Jan 18 16:06:06 2005 Subject: [V4RB] "ascending" parameter ignored in VSet sorts Message-ID: <21C2025E-699D-11D9-AFE7-000D936D2E58@vermontsoftworks.com> FYI: 0000303: "ascending" parameter ignored in VSet sorts Sorting a VSet's records for a VArraySet ignores the "ascending" parameter. Records are sorted "ascending" regardless of its value. Example: ??????????iSet = iTable.Sort(vs, iTable.chosenDT, False) sorts on field "chosenDT" with records ascending. Also, both the documentation and the REALBasic "tips" window for the Sort function incorrectly spell "ascending" without its "s," as "acending". Thanks, -- Erik From sunshine at public.kherson.ua Wed Jan 19 00:13:59 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Tue Jan 18 16:14:06 2005 Subject: FIXED: [V4RB] "ascending" parameter ignored in VSet sorts In-Reply-To: <21C2025E-699D-11D9-AFE7-000D936D2E58@vermontsoftworks.com> Message-ID: On 1/19/05 12:05 AM, "Erik Mueller-Harder" wrote: > FYI: > > 0000303: "ascending" parameter ignored in VSet sorts > > Sorting a VSet's records for a VArraySet ignores the "ascending" > parameter. Records are sorted "ascending" regardless of its value. > > Example: > ??????????iSet = iTable.Sort(vs, iTable.chosenDT, False) > > sorts on field "chosenDT" with records ascending. > > Also, both the documentation and the REALBasic "tips" window for the > Sort function incorrectly spell "ascending" without its "s," as > "acending". > > Thanks, > > -- Erik > > _______________________________________________ > Valentina-beta mailing list > Valentina-beta@lists.macserve.net > http://lists.macserve.net/mailman/listinfo/valentina-beta -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From valentina-list at vermontsoftworks.com Tue Jan 18 18:39:17 2005 From: valentina-list at vermontsoftworks.com (Erik Mueller-Harder) Date: Tue Jan 18 17:39:23 2005 Subject: FIXED: [V4RB] "ascending" parameter ignored in VSet sorts In-Reply-To: References: Message-ID: <2AF3290D-69AA-11D9-AFE7-000D936D2E58@vermontsoftworks.com> 8 minutes! It took you 8 minutes to solve this one! Paint me black and white and call me a zebra! Well done -- and thank you! -- Erik From sunshine at public.kherson.ua Wed Jan 19 02:28:48 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Tue Jan 18 18:28:55 2005 Subject: [V4RB] b37 Message-ID: kernel - 2.0 b37 ================ - 0000316: [Files] I_Location::get_Parent() (Ivan Smahin) - 0000293: [Database] int GetDatabaseVersion( Location_Ptr file ). (Ivan Smahin) - 0000286: [SQL Parser] CONVERT RDB LINK (Sergey Duminskiy) - 0000285: [SQL Parser] CREATE TABLE must support METHOD('txt') (Sergey Duminskiy) - 0000284: [Installation] DLLs can be in the Application folder now. (Ruslan Zasukhin) V4RB - 2.0 b37 ============== - 0000315: [Examples] NEW: Common/Convert_1_2 (Ruslan Zasukhin) - 0000314: [Examples] NEW: Common/GetDatabaseVersion (Ruslan Zasukhin) - 0000313: [Examples] NEW: SQL-way/MethodCreate (Ruslan Zasukhin) - 0000303: [API-way] "ascending" parameter ignored in VSet sorts (Ruslan Zasukhin) - 0000283: [NEW] Valentina.Convert_1_2( OldDb as Location, NewDb as Location, CopyRecords as boolean) (Ivan Smahin) - 0000059: [Feature Request] Set path to VComponents (Ruslan Zasukhin) - 0000279: [API] New enum is added EVDateFormat { kMDY, kDMY, kYMD } (Ruslan Zasukhin) - 0000278: [API] New enum is added EVTableKind { kTblPermanent, kTblTemporary } (Ruslan Zasukhin) - 0000277: [Tables] Vatabase.CreateTable() now have parameters inTableKind, inStorageType (Ruslan Zasukhin) - 0000274: [API] added Enum EVDateFormat (Ruslan Zasukhin) -------------- next part -------------- An HTML attachment was scrubbed... URL: http://lists.macserve.net/pipermail/valentina-beta/attachments/20050119/7f195b76/attachment.html From jda at his.com Tue Jan 18 20:02:33 2005 From: jda at his.com (jda) Date: Tue Jan 18 19:02:45 2005 Subject: [V4RB] b37 In-Reply-To: References: Message-ID: Hi Ruslan, First, it converts my db without crashing (good). But the db size went from 196 kb (1.x) to 4 MB (bad). CopyRecords was set to true. Jon From jda at his.com Tue Jan 18 20:10:07 2005 From: jda at his.com (jda) Date: Tue Jan 18 19:10:13 2005 Subject: [V4RB] b37 In-Reply-To: References: Message-ID: >Hi Ruslan, > >First, it converts my db without crashing (good). But the db size >went from 196 kb (1.x) to 4 MB (bad). CopyRecords was set to true. > Hm, when I try to open the 4 MB converted file I get this error: File "  %o," does not exist. Error number = 365826 This is the same database I sent you, so you should be able to reproduce this. Jon P.S. if CopyRecords = false the new db is still 4 MB. From jda at his.com Tue Jan 18 20:53:34 2005 From: jda at his.com (jda) Date: Tue Jan 18 19:53:40 2005 Subject: Convert In-Reply-To: References: Message-ID: Hi Ruslan, One other thing with convert -- my db's have a resource fork, and that's not copied to the new db. I can create it myself, of course, and I will if I have to. But I think V2 should do that automatically. Jon From sunshine at public.kherson.ua Wed Jan 19 09:42:11 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Wed Jan 19 01:42:22 2005 Subject: Convert In-Reply-To: Message-ID: On 1/19/05 3:53 AM, "jda" wrote: Hi Jon, > One other thing with convert -- my db's have a resource fork, and > that's not copied to the new db. I can create it myself, of course, > and I will if I have to. But I think V2 should do that automatically. No, Valentina db do not have resource forks, so it should not. If you have made resources then you should copy them self. -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From sunshine at public.kherson.ua Wed Jan 19 09:42:39 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Wed Jan 19 01:42:47 2005 Subject: [V4RB] b37 In-Reply-To: Message-ID: On 1/19/05 3:02 AM, "jda" wrote: > Hi Ruslan, > > First, it converts my db without crashing (good). But the db size > went from 196 kb (1.x) to 4 MB (bad). CopyRecords was set to true. Aha, db.segmentsize was not copied -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From sunshine at public.kherson.ua Wed Jan 19 10:49:04 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Wed Jan 19 02:49:31 2005 Subject: V4RB, New documentation uploaded Message-ID: Hi All, You can download updated V4RB_Refernce.pdf -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From vidal_olivier at yahoo.fr Wed Jan 19 15:52:15 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Wed Jan 19 08:52:28 2005 Subject: indexes cached + UTF8 In-Reply-To: References: Message-ID: Le 11 janv. 05, ? 09:17, Ruslan Zasukhin a ?crit : > Oliver, first time Valentina build special index for sorting. > I again say, INDEXES are NOT cached now. This is a reason. > > Do not test for now big number of records. > I think on this week we will make indexed to be cached. Hi list, Ruslan, Do you think that that will be made this week? (So that I can work with the true (big) data) You said that the data in UTF16 took twice more place than in UTF8. If my application is sold only in my country (France), I have no interest to store my data in UTF16, right? (Moreover, it must be slower I suppose) If I understood well, the only interest to use UTF16 is to be able to use certain oriental languages? Soon, we can choose in Valentina to store the data in UTF8 rather than in UTF16? When do you think of being ready for it? thank you olivier From jda at his.com Wed Jan 19 09:59:02 2005 From: jda at his.com (jda) Date: Wed Jan 19 08:59:15 2005 Subject: indexes cached + UTF8 In-Reply-To: References: Message-ID: > > >Do you think that that will be made this week? (So that I can work >with the true (big) data) > >You said that the data in UTF16 took twice more place than in UTF8. >If my application is sold only in my country (France), I have no >interest to store my data in UTF16, right? >(Moreover, it must be slower I suppose) > >If I understood well, the only interest to use UTF16 is to be able >to use certain oriental languages? > >Soon, we can choose in Valentina to store the data in UTF8 rather >than in UTF16? >When do you think of being ready for it? > >thank you > >olivier > Olivier, you read my mind. Once Convert is working without problems, the only two issues I need to get my app out the door are: (1) Unicode/long file names and (2) using cache for databases. UTF8 storage is important for me, too, though, and would be nice in the 2.0 release. But if not, then hopefully in 2.0.1 a week or two later? Jon From jordan at digi-net.com Wed Jan 19 11:07:18 2005 From: jordan at digi-net.com (Ken Jordan) Date: Wed Jan 19 10:04:45 2005 Subject: [V4RB] b37 In-Reply-To: References: Message-ID: <41EE85B6.3050503@digi-net.com> Hey - where's the Windows version :-) Ruslan Zasukhin wrote: > > kernel - 2.0 b37 > ================ > - 0000316: *[Files]* I_Location::get_Parent() (Ivan Smahin) > - 0000293: *[Database]* int GetDatabaseVersion( Location_Ptr file ). > (Ivan Smahin) > - 0000286: *[SQL Parser]* CONVERT RDB LINK (Sergey Duminskiy) > - 0000285: *[SQL Parser]* CREATE TABLE must support METHOD('txt') > (Sergey Duminskiy) > - 0000284: *[Installation]* DLLs can be in the Application folder now. > (Ruslan Zasukhin) > > > > V4RB - 2.0 b37 > ============== > - 0000315: *[Examples]* NEW: Common/Convert_1_2 (Ruslan Zasukhin) > - 0000314: *[Examples]* NEW: Common/GetDatabaseVersion (Ruslan Zasukhin) > - 0000313: *[Examples]* NEW: SQL-way/MethodCreate (Ruslan Zasukhin) > - 0000303: *[API-way]* "ascending" parameter ignored in VSet sorts > (Ruslan Zasukhin) > - 0000283: *[NEW]* Valentina.Convert_1_2( OldDb as Location, NewDb as > Location, CopyRecords as boolean) (Ivan Smahin) > - 0000059: *[Feature Request]* Set path to VComponents (Ruslan Zasukhin) > - 0000279: *[API]* New enum is added EVDateFormat { kMDY, kDMY, kYMD } > (Ruslan Zasukhin) > - 0000278: *[API]* New enum is added EVTableKind { kTblPermanent, > kTblTemporary } (Ruslan Zasukhin) > - 0000277: *[Tables]* Vatabase.CreateTable() now have parameters > inTableKind, inStorageType (Ruslan Zasukhin) > - 0000274: *[API]* added Enum EVDateFormat (Ruslan Zasukhin) > >------------------------------------------------------------------------ > >_______________________________________________ >Valentina-beta mailing list >Valentina-beta@lists.macserve.net >http://lists.macserve.net/mailman/listinfo/valentina-beta > > -------------- next part -------------- An HTML attachment was scrubbed... URL: http://lists.macserve.net/pipermail/valentina-beta/attachments/20050119/cee29c62/attachment.html From sunshine at public.kherson.ua Wed Jan 19 18:06:29 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Wed Jan 19 10:06:36 2005 Subject: [V4RB] b37 In-Reply-To: <41EE85B6.3050503@digi-net.com> Message-ID: On 1/19/05 6:07 PM, "Ken Jordan" wrote: > Hey - where's the Windows version :-) Yes right. 1-3 hours please yet -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From sunshine at public.kherson.ua Wed Jan 19 18:58:58 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Wed Jan 19 10:59:10 2005 Subject: Just small news on Vserver and V4MD and Shockwave, and V4RB Message-ID: Hi All, Well, we have now working * Valentina Server 2.0 for Windows Valentina Server 2.0 for MAC * V4MD 2.0 It is marked as Shockwave safe. V4MD.w32 -- SW Package for Windows (it is 300KB size) tested on Windows it works! Igor still polish although. ------------------------------------ Local + Client/Server technologies Let me give again info how this will work in 2.0. * so we have only ONE plugin V4RB or V4MD. This plugin have WEAK link with 2 DLLS. - kernel.dll - client.dll. If you want only local, you can remove client.dll. if you want only client, you can remove kernel.dll And now attention! if you want to have app that in the same time work with local and remote database you need both this dlls. if I not mistake, any other database vendor for RB or Director do not allow such feature: in the same time work with local and remote database. If you know such database please point us. -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From sunshine at public.kherson.ua Wed Jan 19 20:33:32 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Wed Jan 19 12:33:37 2005 Subject: indexes cached + UTF8 In-Reply-To: Message-ID: On 1/19/05 4:52 PM, "Olivier" wrote: > Do you think that that will be made this week? (So that I can work with > the true (big) data) Indexes yes, UTF8 no. -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From sunshine at public.kherson.ua Wed Jan 19 20:52:26 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Wed Jan 19 12:52:33 2005 Subject: indexes cached + UTF8 In-Reply-To: Message-ID: On 1/19/05 4:59 PM, "jda" wrote: > > Olivier, you read my mind. Once Convert is working without problems, > the only two issues I need to get my app out the door are: (1) > Unicode/long file names and (2) using cache for databases. UTF8 > storage is important for me, too, though, and would be nice in the > 2.0 release. But if not, then hopefully in 2.0.1 a week or two later? Good plan :-) -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From sunshine at public.kherson.ua Wed Jan 19 21:33:53 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Wed Jan 19 13:34:01 2005 Subject: [V4RB] b37 for Win-Hosting uploaded Message-ID: -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From jda at his.com Wed Jan 19 14:56:31 2005 From: jda at his.com (jda) Date: Wed Jan 19 13:56:44 2005 Subject: Convert Message-ID: Hi Ruslan, I don't understand the logic in your Convert example: ------------- DbVersion = valentina.GetDatabaseVersion( f ) s = Hex( DbVersion ) // If database is not of 1.x version then return. if DbVersion = 0 or DbVersion >= &H200 then return end if ------------- &H200 has a value of 512. When I try to open my v1.x db I get a DBVersion value of 400 and a hex value of 190. So it is seen as NOT 1.x. When I open my v2 db I get a DBVersion value of 2 and a hex value of 2. So it is also seen as NOT 1.x. Shouldn't V2 db's return a (hex) value of &h200? Jon From sunshine at public.kherson.ua Wed Jan 19 23:17:16 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Wed Jan 19 15:17:22 2005 Subject: Convert In-Reply-To: Message-ID: On 1/19/05 9:56 PM, "jda" wrote: > Hi Ruslan, > > I don't understand the logic in your Convert example: > > ------------- > DbVersion = valentina.GetDatabaseVersion( f ) > > s = Hex( DbVersion ) > > // If database is not of 1.x version then return. > if DbVersion = 0 or DbVersion >= &H200 then > return > end if > ------------- > > &H200 has a value of 512. > > When I try to open my v1.x db I get a DBVersion value of 400 and a > hex value of 190. So it is seen as NOT 1.x. 1.9.0 H190 = 400 2.0.0 H200 = 512 2.0 > 1.9 right? so 512 > 400 All right here. Just HEX arithmetic > When I open my v2 db I get a DBVersion value of 2 and a hex value of > 2. So it is also seen as NOT 1.x. > > Shouldn't V2 db's return a (hex) value of &h200? -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From jda at his.com Wed Jan 19 16:24:12 2005 From: jda at his.com (jda) Date: Wed Jan 19 15:24:35 2005 Subject: Convert In-Reply-To: References: Message-ID: > > > >> When I try to open my v1.x db I get a DBVersion value of 400 and a >> hex value of 190. So it is seen as NOT 1.x. > > 1.9.0 H190 = 400 > 2.0.0 H200 = 512 > > > 2.0 > 1.9 right? >so 512 > 400 > >All right here. > >Just HEX arithmetic > > >> When I open my v2 db I get a DBVersion value of 2 and a hex value of >> 2. So it is also seen as NOT 1.x. >> > > Shouldn't V2 db's return a (hex) value of &h200? > Yes. But a V2 db returns a value of &h002 (decimal 2), NOT &h200 (decimal 512). Jon From sunshine at public.kherson.ua Wed Jan 19 23:28:34 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Wed Jan 19 15:28:48 2005 Subject: Convert In-Reply-To: Message-ID: On 1/19/05 11:24 PM, "jda" wrote: >>> When I open my v2 db I get a DBVersion value of 2 and a hex value of >>> 2. So it is also seen as NOT 1.x. >>> >>> Shouldn't V2 db's return a (hex) value of &h200? >> > > Yes. But a V2 db returns a value of &h002 (decimal 2), NOT &h200 (decimal > 512). I will check this -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From sunshine at public.kherson.ua Wed Jan 19 23:57:17 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Wed Jan 19 15:57:35 2005 Subject: Convert In-Reply-To: Message-ID: On 1/19/05 11:28 PM, "Ruslan Zasukhin" wrote: > On 1/19/05 11:24 PM, "jda" wrote: > >>>> When I open my v2 db I get a DBVersion value of 2 and a hex value of >>>> 2. So it is also seen as NOT 1.x. >>>> >>>> Shouldn't V2 db's return a (hex) value of &h200? >>> >> >> Yes. But a V2 db returns a value of &h002 (decimal 2), NOT &h200 (decimal >> 512). > > I will check this I see bug. Fixing.. So must be H200 of course -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From ralf at end-if.de Wed Jan 19 23:15:40 2005 From: ralf at end-if.de (Ralf Sander) Date: Wed Jan 19 16:16:04 2005 Subject: OID field Message-ID: Hi Ruslan, will the OID field fnd it's way into 2.0? Cheers, Ralf > Hi All, > > Yes we plan to add auto-increment field > For better support of Relational model. > > RecID is more or less feature of Object-Relational model. > > As I have told in 2.0 we also introduce OID field, > This is feature of OO model. > > > -- > Best regards, > Ruslan Zasukhin [ I feel the need...the need for speed ] From sunshine at public.kherson.ua Thu Jan 20 00:29:35 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Wed Jan 19 16:29:51 2005 Subject: OID field In-Reply-To: Message-ID: On 1/20/05 12:15 AM, "Ralf Sander" wrote: > Hi Ruslan, > > will the OID field fnd it's way into 2.0? Yes. You can see it right now in the SQL query. Try SELECT ** FROM T Or SELECT RecId, OID, f1, f4 FROM T Hmm, it seems OID works as RecID... I will check right now... -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From vidal_olivier at yahoo.fr Thu Jan 20 11:08:46 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Thu Jan 20 04:09:20 2005 Subject: error 537860 "unique violation" Message-ID: <4599F619-6ACB-11D9-8F04-000D93C35C66@yahoo.fr> Hi Ruslan and list, I have in my DB : table.constructor string1=new vstring("string1",5,EVflag.fIndexed) string2=new vvarChar("string2",100,EVflag.fIndexed+EVflag.findexByWords) string3=new vvarchar("string3",105,Evflag.fIndexed+EVflag.funique,"CONCAT(string1,st ring2)") thus, - string1 = indexed - string2 = indexbyword - sting3 = (string1+string2).unique record1 : string1 = "75001" string2="PARIS" string3="75001PARIS" if I try to add the same record, I have the error 537860 "unique vilation" and not the error 144643. if I catch this error (537860), error is catched BUT the ROW RECORDING ALL THE SAME. thus, I have : record1 : string1 = "75001" string2="PARIS" string3="75001PARIS" record2 : string1 = "75001" string2="PARIS" string3="75001PARIS" Olivier From sunshine at public.kherson.ua Thu Jan 20 12:32:57 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Thu Jan 20 04:33:08 2005 Subject: error 537860 "unique violation" In-Reply-To: <4599F619-6ACB-11D9-8F04-000D93C35C66@yahoo.fr> Message-ID: On 1/20/05 12:08 PM, "Olivier" wrote: Oliver, Please put this bug into Mantis And can you send some project that reproduce this problem. I think tip here that method marked as unique. > Hi Ruslan and list, > > I have in my DB : > > table.constructor > > string1=new vstring("string1",5,EVflag.fIndexed) > string2=new > vvarChar("string2",100,EVflag.fIndexed+EVflag.findexByWords) > string3=new > vvarchar("string3",105,Evflag.fIndexed+EVflag.funique,"CONCAT(string1,st > ring2)") > > thus, > - string1 = indexed > - string2 = indexbyword > - sting3 = (string1+string2).unique > > record1 : > > string1 = "75001" > string2="PARIS" > string3="75001PARIS" > > if I try to add the same record, I have the error 537860 "unique > vilation" and not the error 144643. > if I catch this error (537860), error is catched BUT the ROW RECORDING > ALL THE SAME. > > thus, I have : > > record1 : > > string1 = "75001" > string2="PARIS" > string3="75001PARIS" > > record2 : > > string1 = "75001" > string2="PARIS" > string3="75001PARIS" > > > Olivier > > _______________________________________________ > Valentina-beta mailing list > Valentina-beta@lists.macserve.net > http://lists.macserve.net/mailman/listinfo/valentina-beta -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From vidal_olivier at yahoo.fr Thu Jan 20 11:54:33 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Thu Jan 20 04:54:34 2005 Subject: error 537860 "unique violation" In-Reply-To: References: Message-ID: Le 20 janv. 05, ? 11:32, Ruslan Zasukhin a ?crit : > Please put this bug into Mantis > > And can you send some project that reproduce this problem. ok, I try today. olivier From jordan at digi-net.com Thu Jan 20 09:45:40 2005 From: jordan at digi-net.com (Ken Jordan) Date: Thu Jan 20 08:42:51 2005 Subject: [V4RB] b37 for Win-Hosting uploaded In-Reply-To: References: Message-ID: <41EFC414.1080103@digi-net.com> Seems to be missing a dll - MSL_All-90_x86_D.dll Ken From sunshine at public.kherson.ua Thu Jan 20 18:11:07 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Thu Jan 20 10:11:13 2005 Subject: [V4RB] b37 for Win-Hosting uploaded In-Reply-To: <41EFC414.1080103@digi-net.com> Message-ID: On 1/20/05 4:45 PM, "Ken Jordan" wrote: > Seems to be missing a dll - MSL_All-90_x86_D.dll In the Vcomponents? Right. Will fix now -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From sunshine at public.kherson.ua Thu Jan 20 18:21:54 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Thu Jan 20 10:21:58 2005 Subject: FIXED: [V4RB] b37 for Win-Hosting uploaded In-Reply-To: <41EFC414.1080103@digi-net.com> Message-ID: On 1/20/05 4:45 PM, "Ken Jordan" wrote: > Seems to be missing a dll - MSL_All-90_x86_D.dll You need re-download vengine_win_cw archive -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From jordan at digi-net.com Thu Jan 20 12:52:57 2005 From: jordan at digi-net.com (Ken Jordan) Date: Thu Jan 20 11:50:06 2005 Subject: FIXED: [V4RB] b37 for Win-Hosting uploaded In-Reply-To: References: Message-ID: <41EFEFF9.30601@digi-net.com> With yesterday's vclient_win32_cw.dll RealBasic startup stopped with the message that MSL_All-90_x86_D.dll could not be found. With today's vclient_win32_cw.dll RealBasic startup just aborts - no message of any kind. Location of all dlls is in the path. ????? Ken Ruslan Zasukhin wrote: >On 1/20/05 4:45 PM, "Ken Jordan" wrote: > > > >>Seems to be missing a dll - MSL_All-90_x86_D.dll >> >> > >You need re-download vengine_win_cw archive > > > > -------------- next part -------------- An HTML attachment was scrubbed... URL: http://lists.macserve.net/pipermail/valentina-beta/attachments/20050120/3ea07ddc/attachment.html From sunshine at public.kherson.ua Thu Jan 20 19:57:06 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Thu Jan 20 11:57:13 2005 Subject: FIXED: [V4RB] b37 for Win-Hosting uploaded In-Reply-To: <41EFEFF9.30601@digi-net.com> Message-ID: On 1/20/05 7:52 PM, "Ken Jordan" wrote: > With yesterday's vclient_win32_cw.dll RealBasic startup stopped with the > message that MSL_All-90_x86_D.dll could not be found. That was mistake, _D this is debug library. > With today's vclient_win32_cw.dll RealBasic startup just aborts - no message > of any kind. > > Location of all dlls is in the path. > > ????? Strange. I need check again. -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From sunshine at public.kherson.ua Thu Jan 20 20:38:19 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Thu Jan 20 12:38:24 2005 Subject: FIXED: [V4RB] b37 for Win-Hosting uploaded In-Reply-To: Message-ID: On 1/20/05 7:57 PM, "Ruslan Zasukhin" wrote: > On 1/20/05 7:52 PM, "Ken Jordan" wrote: > >> With yesterday's vclient_win32_cw.dll RealBasic startup stopped with the >> message that MSL_All-90_x86_D.dll could not be found. > > That was mistake, _D this is debug library. > >> With today's vclient_win32_cw.dll RealBasic startup just aborts - no message >> of any kind. >> >> Location of all dlls is in the path. >> >> ????? > > Strange. > > I need check again. Sorry, Ken, We have start redesign of sources, so until we finish, I can't make new build and test it. -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From jordan at digi-net.com Thu Jan 20 14:44:05 2005 From: jordan at digi-net.com (Ken Jordan) Date: Thu Jan 20 13:41:10 2005 Subject: FIXED: [V4RB] b37 for Win-Hosting uploaded In-Reply-To: References: Message-ID: <41F00A05.8070505@digi-net.com> OK - thanks for the info. Will keep working on Mac and watch for the update. Ken Ruslan Zasukhin wrote: >On 1/20/05 7:57 PM, "Ruslan Zasukhin" wrote: > > > >>On 1/20/05 7:52 PM, "Ken Jordan" wrote: >> >> >> >>>With yesterday's vclient_win32_cw.dll RealBasic startup stopped with the >>>message that MSL_All-90_x86_D.dll could not be found. >>> >>> >>That was mistake, _D this is debug library. >> >> >> >>>With today's vclient_win32_cw.dll RealBasic startup just aborts - no message >>>of any kind. >>> >>>Location of all dlls is in the path. >>> >>>????? >>> >>> >>Strange. >> >>I need check again. >> >> > >Sorry, Ken, > >We have start redesign of sources, so until we finish, >I can't make new build and test it. > > > -------------- next part -------------- An HTML attachment was scrubbed... URL: http://lists.macserve.net/pipermail/valentina-beta/attachments/20050120/002c7ce9/attachment.html From jordan at digi-net.com Thu Jan 20 14:51:33 2005 From: jordan at digi-net.com (Ken Jordan) Date: Thu Jan 20 13:48:36 2005 Subject: FIXED: [V4RB] b37 for Win-Hosting uploaded In-Reply-To: References: Message-ID: <41F00BC5.10607@digi-net.com> OK - thanks for the info. Will keep working on Mac and watch for the update. Ken Ruslan Zasukhin wrote: >On 1/20/05 7:57 PM, "Ruslan Zasukhin" wrote: > > > >>On 1/20/05 7:52 PM, "Ken Jordan" wrote: >> >> >> >>>With yesterday's vclient_win32_cw.dll RealBasic startup stopped with the >>>message that MSL_All-90_x86_D.dll could not be found. >>> >>> >>That was mistake, _D this is debug library. >> >> >> >>>With today's vclient_win32_cw.dll RealBasic startup just aborts - no message >>>of any kind. >>> >>>Location of all dlls is in the path. >>> >>>????? >>> >>> >>Strange. >> >>I need check again. >> >> > >Sorry, Ken, > >We have start redesign of sources, so until we finish, >I can't make new build and test it. > > > -------------- next part -------------- An HTML attachment was scrubbed... URL: http://lists.macserve.net/pipermail/valentina-beta/attachments/20050120/77df676a/attachment.html From sunshine at public.kherson.ua Fri Jan 21 15:21:08 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Fri Jan 21 07:21:23 2005 Subject: "unique" flag doesn't work In-Reply-To: <0B31C9A5-6BAF-11D9-96AD-000D93C35C66@yahoo.fr> Message-ID: On 1/21/05 3:19 PM, "Olivier" wrote: > > Ruslan, > > In spite of "unique" flag, two identical records are recorded. > > example : > > > > > > 1- run project > 2- click on "Add record" : ok > 2- click again on "Add record" : exception "unique violation on index.." > 3- nevertheless "unique flag" and exception, the two records are > enregistred. Oliver, Ivan already have fix this bug! -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From ralf at end-if.de Fri Jan 21 15:39:37 2005 From: ralf at end-if.de (End If Software - Ralf Sander) Date: Fri Jan 21 08:39:57 2005 Subject: OID field In-Reply-To: Message-ID: Hi Ruslan, is the OID field only accessible via sql? No Api like the RecID field? Cheers, Ralf on 19.01.2005 23:29 Uhr, Ruslan Zasukhin wrote: > On 1/20/05 12:15 AM, "Ralf Sander" wrote: > >> Hi Ruslan, >> >> will the OID field fnd it's way into 2.0? > > Yes. > > You can see it right now in the SQL query. > > Try > > SELECT ** FROM T > > Or > > SELECT RecId, OID, f1, f4 FROM T > > > Hmm, it seems OID works as RecID... > I will check right now... > ----- Die MediaMaid-Familie ? professionelles Media-Asset-Management GrabbyWindow ? Drag-and-Drop neu erfunden ----- End If Software Eckendorfer Stra?e 109, 33609 Bielefeld http://www.end-if.de ----- From vidal_olivier at yahoo.fr Fri Jan 21 15:50:49 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Fri Jan 21 08:50:50 2005 Subject: methods Message-ID: Hi Ruslan and list, In Valentina 2 also, the methods of fields work only on the fields of their table ? We cannot make a method which would work on fields of the other tables ? thank you olivier From sunshine at public.kherson.ua Fri Jan 21 18:26:45 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Fri Jan 21 10:26:52 2005 Subject: OID field In-Reply-To: Message-ID: On 1/21/05 4:39 PM, "End If Software - Ralf Sander" wrote: > Hi Ruslan, > > is the OID field only accessible via sql? No Api like the RecID field? Table.OID ? Problem is that REALbiasic and Director both do not support Llong values. You can access this field as Table.Field("OID").GetString -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From sunshine at public.kherson.ua Fri Jan 21 18:29:24 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Fri Jan 21 10:29:29 2005 Subject: methods In-Reply-To: Message-ID: On 1/21/05 4:50 PM, "Olivier" wrote: Hi Oliver, Very good question. > In Valentina 2 also, the methods of fields work only on the fields of > their table ? > We cannot make a method which would work on fields of the other tables ? 1) In general yes. 2) Only exception -- new methods that work on Link abstraction. PLEASE CHECK ValentinaSQL.pdf it have this section of functions. read it then ask any questions. 3) what is your task? give us example of what you need ? -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From vidal_olivier at yahoo.fr Fri Jan 21 18:53:39 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Fri Jan 21 11:53:42 2005 Subject: methods In-Reply-To: References: Message-ID: <617FB0BF-6BD5-11D9-96AD-000D93C35C66@yahoo.fr> > > > 3) what is your task? > give us example of what you need ? I have a table of ZIP-Citys : zip=new vstring("zip",5,EVflag.fIndexed) city=new vvarChar("city",100,EVflag.fIndexed+EVflag.findexByWords) CityIdx=new vvarChar("cityIdx",100,EVflag.fIndexed,"city") and a table of streets : street=new vvarChar("street",100,EVflag.fIndexed+EVflag.fIndexByWords) streetIdx=new vvarChar("streetIdx",25,EVflag.findexed,"street") Every street belongs to a couple zip / city : link_Street_ZipCity=new vbinaryLink("link_Street_ZipCity", streetTb, ZipCityTb, EVlinkType.kOne, EVlinkType.kMany) In a couple zip-city, it cannot have two identical streets there. I cannot use of "unique" flag on the column " street " because it can have several streets with the same name but in different cities there. I thus added in my table "streets", a column " Zip+city+street " which contains RecId of the couple zip / city + the name of the street. This column has the "unique" flag. OK ZipCityStreet=new vvarChar("zipCityStreet",110,EVflag.fIndexed+EVflag.fUnique) But it takes many place. many. (+ utf16..) If I could make a method, in my table streets : ZipCityStreet=new vvarChar("zipCityStreet",110,EVflag.fIndexed+EVflag.fUnique,"CONCAT(zipC ity.RECID,street)") I shall have "only" indexes. Many place would be saved. thank you olivier From AYU8 at CDC.GOV Fri Jan 21 14:09:39 2005 From: AYU8 at CDC.GOV (Sims, John) Date: Fri Jan 21 13:10:01 2005 Subject: Vset to VCursor Message-ID: Hi Ruslan, I apologize if this is in the documentation already but I just can't find it if it is. Let's say I am working with several unions/intersections of a few different Vsets. I finally get the Vset that represents the data I am interested in. Now what I would like to do is create a Vcursor containing the records I have found. Is it possible to do something like... SELECT FirstName, LastName, ZipCode FROM Person WHERE RecID IN (insert my final Vset here) The only way I can think to do it would be to turn the RecID's of the Vset into an array and use the Binds feature of the SELECT statement. That said, I think it would be a very cool feature to be able to just pass in the Vset... myCursor = SqlSelect("SELECT FirstName, LastName, ZipCode FROM Person WHERE RecID IN (:1)", myVset) If this is not currently possible and you agree that it would be cool and it is doable, let me know and I'll be glad to add it to Mantis. Thanks. -John -------------- next part -------------- An HTML attachment was scrubbed... URL: http://lists.macserve.net/pipermail/valentina-beta/attachments/20050121/4c92f866/attachment-0001.html From sunshine at public.kherson.ua Fri Jan 21 21:14:57 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Fri Jan 21 13:15:02 2005 Subject: Vset to VCursor In-Reply-To: Message-ID: On 1/21/05 9:09 PM, "Sims, John" wrote: Hi John, > I apologize if this is in the documentation already but I just can't find it > if it is. > > Let's say I am working with several unions/intersections of a few different > Vsets. I finally get the Vset that represents the data I am interested in. ok > Now what I would like to do is create a Vcursor containing the records I have > found. Is it possible to do something like? But you do not need cursor if you use API way and sets! You already have TABLE + SET. You can think about TABLE + SET = Cursor You can iterate set, get RecID, goto that record in the table, and done. > SELECT FirstName, LastName, ZipCode FROM Person WHERE RecID IN (insert my > final Vset here) > > The only way I can think to do it would be to turn the RecID's of the Vset > into an array and use the Binds feature of the SELECT statement. That said, I > think it would be a very cool feature to be able to just pass in the Vset? > > myCursor = SqlSelect("SELECT FirstName, LastName, ZipCode FROM Person WHERE > RecID IN (:1)", myVset) > > If this is not currently possible and you agree that it would be cool and it > is doable, let me know and I'll be glad to add it to Mantis. I agree that conversion Cursor <-> Set Can be useful. Although I see that this can be done only for SINGLE TABLE cursor. -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From sunshine at public.kherson.ua Fri Jan 21 22:03:00 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Fri Jan 21 14:03:06 2005 Subject: methods In-Reply-To: <617FB0BF-6BD5-11D9-96AD-000D93C35C66@yahoo.fr> Message-ID: On 1/21/05 7:53 PM, "Olivier" wrote: >> >> >> 3) what is your task? >> give us example of what you need ? > > I have a table of ZIP-Citys : > > zip=new vstring("zip",5,EVflag.fIndexed) > city=new vvarChar("city",100,EVflag.fIndexed+EVflag.findexByWords) > CityIdx=new vvarChar("cityIdx",100,EVflag.fIndexed,"city") > > and a table of streets : > > street=new vvarChar("street",100,EVflag.fIndexed+EVflag.fIndexByWords) > streetIdx=new vvarChar("streetIdx",25,EVflag.findexed,"street") > > Every street belongs to a couple zip / city : > > link_Street_ZipCity=new vbinaryLink("link_Street_ZipCity", streetTb, > ZipCityTb, EVlinkType.kOne, EVlinkType.kMany) I think you have made not correct settings. Must be ONE city -- MANY streets. Right? Then should be link_Street_ZipCity=new vbinaryLink( "link_Street_ZipCity", ZipCityTb, streetTb, EVlinkType.kOne, EVlinkType.kMany) (I have change order of tables) -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From jda at his.com Fri Jan 21 17:24:20 2005 From: jda at his.com (jda) Date: Fri Jan 21 16:24:41 2005 Subject: locked field bug? Message-ID: 1. I create a cursor (kReadOnly or kReadWrite, it doesn't matter). 2. I drop one record from the cursor. 3. The cursor is set to nil 4. Subsequent searches for those records -> field (null) locked. Anything I should be doing differently, or bug? Jon From ralf at end-if.de Sat Jan 22 01:19:12 2005 From: ralf at end-if.de (Ralf Sander) Date: Fri Jan 21 18:19:17 2005 Subject: OID field In-Reply-To: References: Message-ID: <3DE73AC7-6C0B-11D9-945E-000A95DF4532@end-if.de> Shouldn't a double do the job in RB? Cheers, Ralf > On 1/21/05 4:39 PM, "End If Software - Ralf Sander" > wrote: > >> Hi Ruslan, >> >> is the OID field only accessible via sql? No Api like the RecID field? > > Table.OID ? > > Problem is that REALbiasic and Director both do not support Llong > values. > > You can access this field as > > Table.Field("OID").GetString > > -- > Best regards, > Ruslan Zasukhin [ I feel the need...the need for speed ] > ------------------------------------------------------------- > e-mail: ruslan@paradigmasoft.com > web: http://www.paradigmasoft.com > > To subscribe to the Valentina mail list go to: > http://lists.macserve.net/mailman/listinfo/valentina > ------------------------------------------------------------- > > > _______________________________________________ > Valentina-beta mailing list > Valentina-beta@lists.macserve.net > http://lists.macserve.net/mailman/listinfo/valentina-beta > From sunshine at public.kherson.ua Sat Jan 22 02:46:34 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Fri Jan 21 18:46:41 2005 Subject: methods In-Reply-To: <617FB0BF-6BD5-11D9-96AD-000D93C35C66@yahoo.fr> Message-ID: On 1/21/05 7:53 PM, "Olivier" wrote: >> >> >> 3) what is your task? >> give us example of what you need ? > > I have a table of ZIP-Citys : > > zip=new vstring("zip",5,EVflag.fIndexed) > city=new vvarChar("city",100,EVflag.fIndexed+EVflag.findexByWords) > CityIdx=new vvarChar("cityIdx",100,EVflag.fIndexed,"city") > > and a table of streets : > > street=new vvarChar("street",100,EVflag.fIndexed+EVflag.fIndexByWords) > streetIdx=new vvarChar("streetIdx",25,EVflag.findexed,"street") > > Every street belongs to a couple zip / city : > > link_Street_ZipCity=new vbinaryLink("link_Street_ZipCity", streetTb, > ZipCityTb, EVlinkType.kOne, EVlinkType.kMany) ----- > In a couple zip-city, it cannot have two identical streets there. > I cannot use of "unique" flag on the column " street " because it can > have several streets with the same name but in different cities there. > > I thus added in my table "streets", a column " Zip+city+street " which > contains RecId of the couple zip / city + the name of the street. > This column has the "unique" flag. OK > > ZipCityStreet=new > vvarChar("zipCityStreet",110,EVflag.fIndexed+EVflag.fUnique) > > But it takes many place. many. (+ utf16..) > If I could make a method, in my table streets : > > ZipCityStreet=new > vvarChar("zipCityStreet",110,EVflag.fIndexed+EVflag.fUnique,"CONCAT(zipC > ity.RECID,street)") > > I shall have "only" indexes. Many place would be saved. I think in general case solution is TRIGGER. We have no triggers, so you can do this manually. As far as I see, BEFORE add new street to city, You need make search of street with such name. In SQL: FROM streets WHERE streename = 'aaa' and CityPtr = 1 If search returns zero then you ad this street. -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From vidal_olivier at yahoo.fr Sat Jan 22 07:06:19 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Sat Jan 22 00:06:16 2005 Subject: methods In-Reply-To: References: Message-ID: >> >> But it takes many place. many. (+ utf16..) >> If I could make a method, in my table streets : >> >> ZipCityStreet=new >> vvarChar("zipCityStreet",110,EVflag.fIndexed+EVflag.fUnique,"CONCAT(zi >> pC >> ity.RECID,street)") >> >> I shall have "only" indexes. Many place would be saved. > > I think in general case solution is TRIGGER. > > We have no triggers, so you can do this manually. > > As far as I see, BEFORE add new street to city, > You need make search of street with such name. > yes, thank you olivier From sunshine at public.kherson.ua Sat Jan 22 09:32:51 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Sat Jan 22 01:33:01 2005 Subject: OID field In-Reply-To: <3DE73AC7-6C0B-11D9-945E-000A95DF4532@end-if.de> Message-ID: On 1/22/05 2:19 AM, "Ralf Sander" wrote: > Shouldn't a double do the job in RB? No, double have other structure. So it will be wrong. We need again and again press REAL to introduce longlong at least in RB 6 And I do not understand them. This is so simple. This is native computer type. > Cheers, > Ralf > >> On 1/21/05 4:39 PM, "End If Software - Ralf Sander" >> wrote: >> >>> Hi Ruslan, >>> >>> is the OID field only accessible via sql? No Api like the RecID field? >> >> Table.OID ? >> >> Problem is that REALbiasic and Director both do not support Llong >> values. >> >> You can access this field as >> >> Table.Field("OID").GetString -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From vidal_olivier at yahoo.fr Sat Jan 22 13:03:03 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Sat Jan 22 06:03:08 2005 Subject: API methods on binary link In-Reply-To: <617FB0BF-6BD5-11D9-96AD-000D93C35C66@yahoo.fr> References: <617FB0BF-6BD5-11D9-96AD-000D93C35C66@yahoo.fr> Message-ID: <91EBBD80-6C6D-11D9-96AD-000D93C35C66@yahoo.fr> Hi Ruslan and list, > I have a table of ZIP-Citys : > > zip=new vstring("zip",5,EVflag.fIndexed) > city=new vvarChar("city",100,EVflag.fIndexed+EVflag.findexByWords) > CityIdx=new vvarChar("cityIdx",100,EVflag.fIndexed,"city") > > and a table of streets : > > street=new > vvarChar("street",100,EVflag.fIndexed+EVflag.fIndexByWords) > streetIdx=new vvarChar("streetIdx",25,EVflag.findexed,"street") > > Every street belongs to a couple zip / city : > > link_Street_ZipCity=new vbinaryLink("link_Street_ZipCity", ZipCityTb, > StreetTb, EVlinkType.kOne, EVlinkType.kMany) > > If I want to see all the streets the zip of which begins with '33', in SQL I makes: SELECT * FROM ZIPCity, Streets WHERE left(zip,2)='33' How to make easily the same thing with the API? thank you olivier From sunshine at public.kherson.ua Sat Jan 22 15:13:22 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Sat Jan 22 07:13:28 2005 Subject: API methods on binary link In-Reply-To: <91EBBD80-6C6D-11D9-96AD-000D93C35C66@yahoo.fr> Message-ID: On 1/22/05 2:03 PM, "Olivier" wrote: > Hi Ruslan and list, > >> I have a table of ZIP-Citys : >> >> zip=new vstring("zip",5,EVflag.fIndexed) >> city=new vvarChar("city",100,EVflag.fIndexed+EVflag.findexByWords) >> CityIdx=new vvarChar("cityIdx",100,EVflag.fIndexed,"city") >> >> and a table of streets : >> >> street=new >> vvarChar("street",100,EVflag.fIndexed+EVflag.fIndexByWords) >> streetIdx=new vvarChar("streetIdx",25,EVflag.findexed,"street") >> >> Every street belongs to a couple zip / city : >> >> link_Street_ZipCity=new vbinaryLink("link_Street_ZipCity", ZipCityTb, >> StreetTb, EVlinkType.kOne, EVlinkType.kMany) >> >> > > If I want to see all the streets the zip of which begins with '33', in > SQL I makes: > > SELECT * FROM ZIPCity, Streets WHERE left(zip,2)='33' You make at first search in table ZIP set1 = fldZip.Find( 33 ) Now you can iterate set and use Link.FindLinked( inRecID as Integer, inTableA as VTable, inTableB as VTable Hmm, we need enable into V4RB one more function that do set = Link.FindLinked( inSet, TA, TB ) We have such func in kernel. -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From vidal_olivier at yahoo.fr Sat Jan 22 14:21:01 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Sat Jan 22 07:21:04 2005 Subject: API methods on binary link In-Reply-To: References: Message-ID: <75D2D2B1-6C78-11D9-96AD-000D93C35C66@yahoo.fr> Le 22 janv. 05, ? 14:13, Ruslan Zasukhin a ?crit : > Hmm, we need enable into V4RB one more function that do > > set = Link.FindLinked( inSet, TA, TB ) Yes, it is of what I thought. when ? ;o) olivier From sunshine at public.kherson.ua Sat Jan 22 15:25:31 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Sat Jan 22 07:25:53 2005 Subject: API methods on binary link In-Reply-To: <75D2D2B1-6C78-11D9-96AD-000D93C35C66@yahoo.fr> Message-ID: On 1/22/05 3:21 PM, "Olivier" wrote: > > Le 22 janv. 05, ? 14:13, Ruslan Zasukhin a ?crit : > >> Hmm, we need enable into V4RB one more function that do >> >> set = Link.FindLinked( inSet, TA, TB ) > > Yes, it is of what I thought. > > when ? ;o) Not today -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From vidal_olivier at yahoo.fr Sat Jan 22 14:27:16 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Sat Jan 22 07:27:14 2005 Subject: API methods on binary link In-Reply-To: References: Message-ID: <555BC12A-6C79-11D9-96AD-000D93C35C66@yahoo.fr> >> If I want to see all the streets the zip of which begins with '33', in >> SQL I makes: >> >> SELECT * FROM ZIPCity, Streets WHERE left(zip,2)='33' > > You make at first search in table ZIP > > set1 = fldZip.Find( 33 ) > I want ONLY the zip BEGINNING by 33. ( left(zip,2)=33). how ? > Now you can iterate set and use > > Link.FindLinked( > inRecID as Integer, inTableA as VTable, inTableB as VTable > > I does not understand. I use the API because it should be faster than SQL (and because I don't like SQL! ;o-). If I has to use a Realbasic loop, it is going to be slower! RB is not C ++! Slower and more complicated. What interest? > Hmm, we need enable into V4RB one more function that do > > set = Link.FindLinked( inSet, TA, TB ) > > We have such func in kernel. > > > -- > Best regards, > Ruslan Zasukhin [ I feel the need...the need for speed ] > ------------------------------------------------------------- > e-mail: ruslan@paradigmasoft.com > web: http://www.paradigmasoft.com > > To subscribe to the Valentina mail list go to: > http://lists.macserve.net/mailman/listinfo/valentina > ------------------------------------------------------------- > > > _______________________________________________ > Valentina-beta mailing list > Valentina-beta@lists.macserve.net > http://lists.macserve.net/mailman/listinfo/valentina-beta > From sunshine at public.kherson.ua Sat Jan 22 15:36:35 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Sat Jan 22 07:36:42 2005 Subject: API methods on binary link In-Reply-To: <555BC12A-6C79-11D9-96AD-000D93C35C66@yahoo.fr> Message-ID: On 1/22/05 3:27 PM, "Olivier" wrote: >>> If I want to see all the streets the zip of which begins with '33', in >>> SQL I makes: >>> >>> SELECT * FROM ZIPCity, Streets WHERE left(zip,2)='33' >> >> You make at first search in table ZIP >> >> set1 = fldZip.Find( 33 ) >> > > I want ONLY the zip BEGINNING by 33. ( left(zip,2)=33). > > how ? Vfield.FindStartWith( "33" ) Although I am not sure we have finish it for numeric fields. Check it -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From sunshine at public.kherson.ua Sat Jan 22 15:38:15 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Sat Jan 22 07:38:17 2005 Subject: API methods on binary link In-Reply-To: <555BC12A-6C79-11D9-96AD-000D93C35C66@yahoo.fr> Message-ID: On 1/22/05 3:27 PM, "Olivier" wrote: >> Now you can iterate set and use >> >> Link.FindLinked( >> inRecID as Integer, inTableA as VTable, inTableB as VTable >> >> > > I does not understand. > I use the API because it should be faster than SQL (and because I don't > like SQL! ;o-). > If I has to use a Realbasic loop, it is going to be slower! > RB is not C ++! Not a fact. By my tests RB is 10% slower maximum > Slower and more complicated. > What interest? The correct way to have FindLinked( Set, Ta, Tb ) Agree? Then all your questions go away. -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From vidal_olivier at yahoo.fr Sat Jan 22 14:52:42 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Sat Jan 22 07:52:57 2005 Subject: API methods on binary link In-Reply-To: References: Message-ID: > > Vfield.FindStartWith( "33" ) > ok, sorry, I had not seen it. it's not in doc, But I come to see it in RB auto-completion, with findLike, regex, findEndsWith.... I am going to looking better at the auto-completion from now on. olivier From sunshine at public.kherson.ua Sat Jan 22 16:01:23 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Sat Jan 22 08:01:28 2005 Subject: API methods on binary link In-Reply-To: Message-ID: On 1/22/05 3:52 PM, "Olivier" wrote: >> Vfield.FindStartWith( "33" ) >> > > > ok, sorry, I had not seen it. > it's not in doc, But I come to see it in RB auto-completion, with > findLike, regex, findEndsWith.... You have latest docs ? I believe we already have add this to docs. May be not uploaded... > I am going to looking better at the auto-completion from now on. Check latest PDF -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From vidal_olivier at yahoo.fr Sat Jan 22 15:16:04 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Sat Jan 22 08:16:05 2005 Subject: API methods on binary link In-Reply-To: References: Message-ID: <26984A00-6C80-11D9-96AD-000D93C35C66@yahoo.fr> Le 22 janv. 05, ? 15:01, Ruslan Zasukhin a ?crit : > Check latest PDF Exact. I had not had to see the last upload of doc. sorry. From vidal_olivier at yahoo.fr Sat Jan 22 19:37:24 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Sat Jan 22 12:37:46 2005 Subject: API methods on binary link In-Reply-To: References: Message-ID: Hi Ruslan and list, I have the impression that it is not easy to use the API when we want to obtain sorts of temporary tables (consisted of fields of several linked tables). It is maybe because I still thinks in term of SQL. For example, always with my tables of zip / city and streets. I have a datagrid which has to display in the opening of the window a list with three columns. First column: zip (table ZipCity) The second column: city (table ZipCity) The third column: street (table Streets) It has to display all the streets with the zip beginning with 75. The list must be sorted out by zip, city, street. Ex: 75001 PARIS street of Kiev 75001 PARIS street Clemenceau 75001 PARIS place Vend?me 75002 PARIS impasse of Bordeaux The first one and the second column are thus two fields of the table "zip/city" The third column is a field of the table "streets". In SQL: SELECT zip, city, street FROM ZipCityTb, StreetTb WHERE left (zip, 2) = '75' ORDER by zip, city, street We obtain then a cursor with which I can easily fill my datagrid. Especially, the cursor keeps a number for all the line " zip, city, street ", even if originally they are two different tables. It is very practical to modify couples " zipCity-Streets " for example, because the couple is kept in 1 line of the cursor. But with the sets it is less evident because 1 set is attached to 1 single table. The sets are perfected to work on a table but on any? In my example, I do not manage to obtain a practical code. please, How would you make to make an equivalent of this SQL query? thank you very much olivier Le 22 janv. 05, ? 14:13, Ruslan Zasukhin a ?crit : > On 1/22/05 2:03 PM, "Olivier" wrote: > >> Hi Ruslan and list, >> >>> I have a table of ZIP-Citys : >>> >>> zip=new vstring("zip",5,EVflag.fIndexed) >>> city=new vvarChar("city",100,EVflag.fIndexed+EVflag.findexByWords) >>> CityIdx=new vvarChar("cityIdx",100,EVflag.fIndexed,"city") >>> >>> and a table of streets : >>> >>> street=new >>> vvarChar("street",100,EVflag.fIndexed+EVflag.fIndexByWords) >>> streetIdx=new vvarChar("streetIdx",25,EVflag.findexed,"street") >>> >>> Every street belongs to a couple zip / city : >>> >>> link_Street_ZipCity=new vbinaryLink("link_Street_ZipCity", ZipCityTb, >>> StreetTb, EVlinkType.kOne, EVlinkType.kMany) >>> >>> >> >> If I want to see all the streets the zip of which begins with '33', in >> SQL I makes: >> >> SELECT * FROM ZIPCity, Streets WHERE left(zip,2)='33' > > You make at first search in table ZIP > > set1 = fldZip.Find( 33 ) > > Now you can iterate set and use > > Link.FindLinked( > inRecID as Integer, inTableA as VTable, inTableB as VTable > > > Hmm, we need enable into V4RB one more function that do > > set = Link.FindLinked( inSet, TA, TB ) > > We have such func in kernel. > > > -- > Best regards, > Ruslan Zasukhin [ I feel the need...the need for speed ] > ------------------------------------------------------------- > e-mail: ruslan@paradigmasoft.com > web: http://www.paradigmasoft.com > > To subscribe to the Valentina mail list go to: > http://lists.macserve.net/mailman/listinfo/valentina > ------------------------------------------------------------- > > > _______________________________________________ > Valentina-beta mailing list > Valentina-beta@lists.macserve.net > http://lists.macserve.net/mailman/listinfo/valentina-beta > From vidal_olivier at yahoo.fr Sun Jan 23 07:57:27 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Sun Jan 23 00:57:48 2005 Subject: API methods on binary link In-Reply-To: References: Message-ID: <0B2411F5-6D0C-11D9-96AD-000D93C35C66@yahoo.fr> Ruslan, for the moment, The simplest API equivalent of the following SQL query is it ? : SELECT zip, city, street FROM ZipCityTb, StreetTb WHERE left (zip, 2) = '75' ORDER by zip, city, street ( The problem is that I do not manage to cast results Set in Arrayset. I do not thus have to use " findStartWith ") dim setStreetOneCity, setAll,setZipArray, set2 As varraySet dim a,b,i,j,k,m,n As integer setAll=new varraySet(50) setZipArray= base.zipCity.stringField("zipcode").findAsArraySet("75000") setzipArray=base.zipCity.sort(setzipArray,new vsortItem(base.zipCity.zipCode), new vsortItem(base.zipCity.CityName)) k=setZipArray.count for i=1 to k a=setZipArray.ItemAt(i) setstreetOneCity=base.linkCityStreet.findLinked(a,base.zipCity,base.stre et) m=setstreetOneCity.count for j=1 to m setAll.append setStreetOneCity.itemAt(j) next j next i setAll=base.street.sort(setall,base.street.namestreet) k=setall.count for i=1 to k listBox1.AddRow "" a=setAll.ItemAt(i) base.street.recID=a listBox1.cell(i-1,2)=base.street.varcharField("namestreet").getstring set2=base.linkCityStreet.findLinked(a,base.street,base.zipCity) a=set2.itemAt(1) base.zipCity.recID=a listbox1.cell(i-1,0)=base.zipCity.stringField("zipcode").getstring listbox1.cell(i-1,1)=base.zipCity.varCharField("cityname").getstring next i it's works well. If the user now wants to sort out all the list only by street (and not by city), it is necessary to begin again all the operation. right ? Two questions Ruslan: - What are exactly the advantages of the binary links on objectPtr? - The API asks for much more code. Is the result really faster? Only on of very big dbs? thank you very much Ruslan olivier Le 22 janv. 05, ? 19:37, Olivier a ?crit : > Hi Ruslan and list, > > I have the impression that it is not easy to use the API when we want > to obtain sorts of temporary tables (consisted of fields of several > linked tables). > It is maybe because I still thinks in term of SQL. > > For example, always with my tables of zip / city and streets. > > I have a datagrid which has to display in the opening of the window a > list with three columns. > First column: zip (table ZipCity) > The second column: city (table ZipCity) > The third column: street (table Streets) > > It has to display all the streets with the zip beginning with 75. > The list must be sorted out by zip, city, street. > > Ex: > 75001 PARIS street of Kiev > 75001 PARIS street Clemenceau > 75001 PARIS place Vend?me > 75002 PARIS impasse of Bordeaux > > The first one and the second column are thus two fields of the table > "zip/city" > The third column is a field of the table "streets". > > In SQL: > > SELECT zip, city, street FROM ZipCityTb, StreetTb WHERE left (zip, 2) > = '75' ORDER by zip, city, street > > We obtain then a cursor with which I can easily fill my datagrid. > Especially, the cursor keeps a number for all the line " zip, city, > street ", even if originally they are two different tables. > It is very practical to modify couples " zipCity-Streets " for > example, because the couple is kept in 1 line of the cursor. > > But with the sets it is less evident because 1 set is attached to 1 > single table. > > The sets are perfected to work on a table but on any? > In my example, I do not manage to obtain a practical code. > > please, How would you make to make an equivalent of this SQL query? > > thank you very much > > olivier > > > > Le 22 janv. 05, ? 14:13, Ruslan Zasukhin a ?crit : > >> On 1/22/05 2:03 PM, "Olivier" wrote: >> >>> Hi Ruslan and list, >>> >>>> I have a table of ZIP-Citys : >>>> >>>> zip=new vstring("zip",5,EVflag.fIndexed) >>>> city=new vvarChar("city",100,EVflag.fIndexed+EVflag.findexByWords) >>>> CityIdx=new vvarChar("cityIdx",100,EVflag.fIndexed,"city") >>>> >>>> and a table of streets : >>>> >>>> street=new >>>> vvarChar("street",100,EVflag.fIndexed+EVflag.fIndexByWords) >>>> streetIdx=new vvarChar("streetIdx",25,EVflag.findexed,"street") >>>> >>>> Every street belongs to a couple zip / city : >>>> >>>> link_Street_ZipCity=new vbinaryLink("link_Street_ZipCity", >>>> ZipCityTb, >>>> StreetTb, EVlinkType.kOne, EVlinkType.kMany) >>>> >>>> >>> >>> If I want to see all the streets the zip of which begins with '33', >>> in >>> SQL I makes: >>> >>> SELECT * FROM ZIPCity, Streets WHERE left(zip,2)='33' >> >> You make at first search in table ZIP >> >> set1 = fldZip.Find( 33 ) >> >> Now you can iterate set and use >> >> Link.FindLinked( >> inRecID as Integer, inTableA as VTable, inTableB as VTable >> >> >> Hmm, we need enable into V4RB one more function that do >> >> set = Link.FindLinked( inSet, TA, TB ) >> >> We have such func in kernel. >> >> >> -- >> Best regards, >> Ruslan Zasukhin [ I feel the need...the need for speed ] >> ------------------------------------------------------------- >> e-mail: ruslan@paradigmasoft.com >> web: http://www.paradigmasoft.com >> >> To subscribe to the Valentina mail list go to: >> http://lists.macserve.net/mailman/listinfo/valentina >> ------------------------------------------------------------- >> >> >> _______________________________________________ >> Valentina-beta mailing list >> Valentina-beta@lists.macserve.net >> http://lists.macserve.net/mailman/listinfo/valentina-beta >> > > _______________________________________________ > Valentina-beta mailing list > Valentina-beta@lists.macserve.net > http://lists.macserve.net/mailman/listinfo/valentina-beta > From jda at his.com Sun Jan 23 16:26:39 2005 From: jda at his.com (jda) Date: Sun Jan 23 15:26:50 2005 Subject: Yielding time during reindex Message-ID: Hi Ruslan, I seem to recall you saying that in V2 operations like reindex will yield time to the calling app, so that we can, among other things, update progress bars or the like in the user interface. Was that implemented? If so, do we have to put the call in thread (in RB) to have it do this? Jon From sunshine at public.kherson.ua Mon Jan 24 00:44:59 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Sun Jan 23 16:45:20 2005 Subject: Yielding time during reindex In-Reply-To: Message-ID: On 1/23/05 11:26 PM, "jda" wrote: Hi Jon, > I seem to recall you saying that in V2 operations like reindex will > yield time to the calling app, so that we can, among other things, > update progress bars or the like in the user interface. Was that > implemented? If so, do we have to put the call in thread (in RB) to > have it do this? We have told about TaskManager and Task classes. They are not finished as was planned. It should not require threads on RB side. -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From vidal_olivier at yahoo.fr Mon Jan 24 06:44:31 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Sun Jan 23 23:44:43 2005 Subject: SQL and API methods + benchs Message-ID: <053A7180-6DCB-11D9-96AD-000D93C35C66@yahoo.fr> Ruslan, The API methods are exciting and very interesting because we really have the impression to control our DB and because the API methods become integrated perfectly with the other methods RB. valentina is great ! It is true that they ask for more code. I made some benchs with the small project which I sent you. The API methods that I used are the same that in the sent project. The links are binary links. At the moment, SQL is much faster. Maybe that I have badly to optimize my code for the API. Some remarks: I have a crash if I use: SELECT zipcode, cityname, namestreet FROM zipcity, street where zipcode = '75000' With SELECT zipcode, cityname, namestreet FROM zipcity join street on linkCityStreet where zipcode = '75000' That works perfectly. I do not manage to cast the sets. Dim setBit have vbitset Dim setArray have varrayset Dim set have vset setBit=base.zipCity.stringField("zipcode").find("75000") //OK setArray=new varraySet(setBit) -> crash setBit=base.zipCity.stringField("zipcode").find("75000") //OK setArray=varrayset(setBit) -> "illegal cast exception" setBit=base.zipCity.stringField("zipcode").findStartsWith("75") -> "type mismatch error. expected vbitset but got vset" set=base.zipCity.stringField("zipcode").findStartsWith("75") setArray=new varraySet(set) -> "parameters are not cumpatible with this function" set=base.zipCity.stringField("zipcode").findStartsWith("75") setArray=varraySet(set) -> "illegal cast exception" bench tests : the two with valentina 2 b37, RB 5.5.4, os X eMac 1,25Ghz, cache 80 Mb. (with displaying of listbox) table zipCity : 40 000 records (among which 2 with zipCode='75000') table streets : 100 000 records 1 - 100 streets arre linked (with binarylink) with 1 zipcode '75000'. WITH SELECT zipcode, cityname,namestreet FROM zipcity join street on linkcitystreet where zipcode='75000' ORDER BY zipcode, cityname, namestreet first search : 3 or 4 ticks next searchs : 1 ticks, sometimes 2. WITH EQUIVALENT API and Sets first search : 8 ticks next searchs : 6 or 7 ticks 2 - 100 streets arre linked (with binarylink) with 1 zipcode '75000'. WITH SELECT zipcode, cityname,namestreet FROM zipcity join street on linkcitystreet where zipcode='75000' first search : 3 ticks next searchs : 1 ticks, sometimes 2. WITH EQUIVALENT API and Sets first search : 7 ticks next searchs : 3 ticks 3 - 1000 streets arre linked (with binarylink) with 1 zipcode '75000'. WITH SELECT zipcode, cityname,namestreet FROM zipcity join street on linkcitystreet where zipcode='75000' ORDER BY zipcode, cityname, namestreet first search : 25 ticks next searchs : 12 ticks. WITH EQUIVALENT API and Sets first search : 115 ticks next searchs : 33 ticks 4 - 1000 streets arre linked (with binarylink) with 1 zipcode '75000'. WITH SELECT zipcode, cityname,namestreet FROM zipcity join street on linkcitystreet where zipcode='75000' first search : 19 ticks next searchs : 11 ticks. WITH EQUIVALENT API and Sets first search : 33 ticks next searchs : 25 ticks thank you olivier -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: text/enriched Size: 3658 bytes Desc: not available Url : http://lists.macserve.net/pipermail/valentina-beta/attachments/20050124/d5217f37/attachment.bin From sunshine at public.kherson.ua Mon Jan 24 10:05:23 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Mon Jan 24 02:05:30 2005 Subject: SQL and API methods + benchs In-Reply-To: <053A7180-6DCB-11D9-96AD-000D93C35C66@yahoo.fr> Message-ID: On 1/24/05 7:44 AM, "Olivier" wrote: Hi Oliver, > The API methods are exciting and very interesting because we really have the > impression to control our DB and because the API methods become integrated > perfectly with the other methods RB. valentina is great ! > It is true that they ask for more code. > I made some benchs with the small project which I sent you. > The API methods that I used are the same that in the sent project. The links > are binary links. > At the moment, SQL is much faster. Maybe that I have badly to optimize my code > for the API. I also think so. It needs study your task. For now quick notes: 1) have you bench with DebugLevel = 0 ? 2) as far as I see, in the SQL you SELECT all fields of table Zipcity, Right? Actually SQL way have interesting feature: Let you have table with 20 fields, but you select only 4 fields. then Iteration of cursor will be 5 times faster of iteration of original table, because cursor will read only that 4 fields. For now we do not have way do the same trick in API. (in C++ we have in fact, just did not opened to V4RB.) 3) API way has other name -- Navigational way. this means that you should walk from record to record. this means that you can touch ONLY really required records. SQL way in contrast is oriented on the work with SETS of records. you find set of records WHERE. you UPDATE set of records WHERE. you join SET1 to SET2 of T1 and T2.. and so on. This is the main difference in the ideology of these ways. Valentina although try optimize API way to SETs also. The old Navigational dbs (like dbVista) cannot do this absolutely. -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From sunshine at public.kherson.ua Mon Jan 24 10:06:32 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Mon Jan 24 02:06:39 2005 Subject: SQL and API methods + benchs In-Reply-To: <053A7180-6DCB-11D9-96AD-000D93C35C66@yahoo.fr> Message-ID: On 1/24/05 7:44 AM, "Olivier" wrote: > Some remarks: > > I have a crash if I use: > SELECT zipcode, cityname, namestreet FROM zipcity, street where zipcode = > '75000' > > With > SELECT zipcode, cityname, namestreet FROM zipcity join street on > linkCityStreet where zipcode = '75000' > That works perfectly. Oliver, have you specify FOREIGN KEY for this tables? Or you use ObjectPtrs? Or BinaryLink? Please send to us DB and we will try the first query in the viSQL To fix crash. CC db to Ivan. -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From sunshine at public.kherson.ua Mon Jan 24 10:10:33 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Mon Jan 24 02:10:39 2005 Subject: SQL and API methods + benchs In-Reply-To: <053A7180-6DCB-11D9-96AD-000D93C35C66@yahoo.fr> Message-ID: On 1/24/05 7:44 AM, "Olivier" wrote: > I do not manage to cast the sets. > > Dim setBit have vbitset > Dim setArray have varrayset > Dim set have vset > > setBit=base.zipCity.stringField("zipcode").find("75000") //OK You use FindValue() so it returns BitSet. Oliver, if you want to get ArraySet then you should use simple FindValueAsArraySet() > setArray=new varraySet(setBit) > -> crash Last line try create new ArraySet using BitSet it should not crash. > setBit=base.zipCity.stringField("zipcode").find("75000") //OK > setArray=varrayset(setBit) > -> "illegal cast exception" RIGHT, because setBit is VBitSet. You cannot cast it to ArraySet. > setBit=base.zipCity.stringField("zipcode").findStartsWith("75") > -> "type mismatch error. expected vbitset but got vset" Strange. You use latest beta ? > set=base.zipCity.stringField("zipcode").findStartsWith("75") > setArray=new varraySet(set) > -> "parameters are not cumpatible with this function" > > set=base.zipCity.stringField("zipcode").findStartsWith("75") > setArray=varraySet(set) > -> "illegal cast exception" -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From sunshine at public.kherson.ua Mon Jan 24 10:12:21 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Mon Jan 24 02:12:25 2005 Subject: SQL and API methods + benchs In-Reply-To: <053A7180-6DCB-11D9-96AD-000D93C35C66@yahoo.fr> Message-ID: On 1/24/05 7:44 AM, "Olivier" wrote: > 1 - > > 100 streets arre linked (with binarylink) with 1 zipcode '75000'. > > WITH SELECT zipcode, cityname,namestreet FROM zipcity join street on > linkcitystreet where zipcode='75000' ORDER BY zipcode, cityname, namestreet > > first search : 3 or 4 ticks > next searchs : 1 ticks, sometimes 2. > > WITH EQUIVALENT API and Sets > > first search : 8 ticks > next searchs : 6 or 7 ticks You have sent me code on this so I will do testing self ? -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From vidal_olivier at yahoo.fr Mon Jan 24 09:13:48 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Mon Jan 24 02:14:11 2005 Subject: SQL and API methods + benchs In-Reply-To: References: Message-ID: > > Oliver, have you specify FOREIGN KEY for this tables? > Or you use ObjectPtrs? > Or BinaryLink? > binaryLink. I send project and DB in one half hour. thank you olivier From vidal_olivier at yahoo.fr Mon Jan 24 09:15:48 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Mon Jan 24 02:15:49 2005 Subject: SQL and API methods + benchs In-Reply-To: References: Message-ID: <2718C04C-6DE0-11D9-96AD-000D93C35C66@yahoo.fr> > I send you in half hour. olivier > You have sent me code on this so I will do testing self ? > > -- > Best regards, > Ruslan Zasukhin [ I feel the need...the need for speed ] > ------------------------------------------------------------- > e-mail: ruslan@paradigmasoft.com > web: http://www.paradigmasoft.com > > To subscribe to the Valentina mail list go to: > http://lists.macserve.net/mailman/listinfo/valentina > ------------------------------------------------------------- > > > _______________________________________________ > Valentina-beta mailing list > Valentina-beta@lists.macserve.net > http://lists.macserve.net/mailman/listinfo/valentina-beta > From vidal_olivier at yahoo.fr Mon Jan 24 09:19:39 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Mon Jan 24 02:19:39 2005 Subject: SQL and API methods + benchs In-Reply-To: References: Message-ID: Le 24 janv. 05, ? 09:10, Ruslan Zasukhin a ?crit : > On 1/24/05 7:44 AM, "Olivier" wrote: > >> I do not manage to cast the sets. >> >> Dim setBit have vbitset >> Dim setArray have varrayset >> Dim set have vset >> >> setBit=base.zipCity.stringField("zipcode").find("75000") //OK > > You use FindValue() so it returns BitSet. > > Oliver, if you want to get ArraySet then you should use simple > FindValueAsArraySet() yes, but it NOT exist FindStartWithAsArraySet or FindLikeAsArraySet or FindRegexAsArraySet... >> setBit=base.zipCity.stringField("zipcode").findStartsWith("75") >> -> "type mismatch error. expected vbitset but got vset" > > Strange. > > You use latest beta ? > yes, b37. findStartWith, findEndsWith... return a set, not a bitset >> set=base.zipCity.stringField("zipcode").findStartsWith("75") >> setArray=new varraySet(set) >> -> "parameters are not cumpatible with this function" >> >> set=base.zipCity.stringField("zipcode").findStartsWith("75") >> setArray=varraySet(set) >> -> "illegal cast exception" > > -- > Best regards, > Ruslan Zasukhin [ I feel the need...the need for speed ] > ------------------------------------------------------------- > e-mail: ruslan@paradigmasoft.com > web: http://www.paradigmasoft.com > > To subscribe to the Valentina mail list go to: > http://lists.macserve.net/mailman/listinfo/valentina > ------------------------------------------------------------- > > > _______________________________________________ > Valentina-beta mailing list > Valentina-beta@lists.macserve.net > http://lists.macserve.net/mailman/listinfo/valentina-beta > From sunshine at public.kherson.ua Mon Jan 24 10:25:14 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Mon Jan 24 02:25:26 2005 Subject: SQL and API methods + benchs In-Reply-To: Message-ID: On 1/24/05 10:19 AM, "Olivier" wrote: >> Strange. >> >> You use latest beta ? >> > > yes, b37. > findStartWith, findEndsWith... return a set, not a bitset Aha, I will fix this! -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From sunshine at public.kherson.ua Mon Jan 24 10:35:24 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Mon Jan 24 02:35:34 2005 Subject: SQL and API methods + benchs In-Reply-To: Message-ID: On 1/24/05 10:19 AM, "Olivier" wrote: >>> Dim setBit have vbitset >>> Dim setArray have varrayset >>> Dim set have vset >>> >>> setBit=base.zipCity.stringField("zipcode").find("75000") //OK >> >> You use FindValue() so it returns BitSet. >> >> Oliver, if you want to get ArraySet then you should use simple >> FindValueAsArraySet() > > yes, but it NOT exist FindStartWithAsArraySet or FindLikeAsArraySet or > FindRegexAsArraySet... Right! Again, ArraySet is preferred only of you KNOW that you will find few records from 40,000 table records. BitSet is preferred in all rest cases. ---------- FindValue() searches for ONE value. Exists many chances that will be not many records that have one the same value. So in this case ArraySet can be good. ---------- Next, in the SQL mashine, we use only bitSets, btw. ---------- If you going do Intersection for results, then you MUST use the same types BitSet + BitSet So in this case use bitSets always. ------------- FindLike() returns BitSet, because MOST PROBABLY it will find many records that match pattern. ---------- So Oliver, AS ARRAY SET functions, will be used very rare I think. Only for specific tasks, when I know that I will find dew records, When I will not intersect with BitSets. I.e. I just want find that 5 records, and iterate them in loop. -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From sunshine at public.kherson.ua Mon Jan 24 10:49:46 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Mon Jan 24 02:49:54 2005 Subject: STATUS Message-ID: Hi All, * starting from prev Wednesday we have stick in refactoring of our sources To break all dependencies between (V4RB/V4MD) and Kernel/Client dlls. This was unexpectedly hard task, so we all have fight here 5 days with this task, and only with this task. Now most things works again. Only MACHO target still broken (Vserver). * Jochen you can get from CVS sources for Windows target and continue Vstudio. Macho I will fix ASAP. * Now we can continue to fix bugs found in b37. -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From sunshine at public.kherson.ua Mon Jan 24 10:58:24 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Mon Jan 24 02:58:33 2005 Subject: SQL and API methods + benchs In-Reply-To: <053A7180-6DCB-11D9-96AD-000D93C35C66@yahoo.fr> Message-ID: On 1/24/05 7:44 AM, "Olivier" wrote: > I have a crash if I use: > SELECT zipcode, cityname, namestreet FROM zipcity, street where zipcode = > '75000' And you say that you use BinaryLink for tables? Then try also THIS syntax: SELECT zipcode, cityname, namestreet FROM zipcity JOIN street ON linkname where zipcode = '75000' I.e. Specify link name > With > SELECT zipcode, cityname, namestreet FROM zipcity join street on > linkCityStreet where zipcode = '75000' > That works perfectly. I think, here works old-good JOIN, based on 2 fields. This join works always, even if you did not make FOREIGN KEY. -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From vidal_olivier at yahoo.fr Mon Jan 24 10:03:57 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Mon Jan 24 03:04:03 2005 Subject: SQL and API methods + benchs In-Reply-To: References: Message-ID: Le 24 janv. 05, ? 09:10, Ruslan Zasukhin a ?crit : >> setArray=new varraySet(setBit) >> -> crash > > Last line try create new ArraySet using BitSet > it should not crash. Nevertheless, try in my project the simple code : dim setbit As vbitSet dim setarray As varraySet setbit=base.zipCity.stringField("zipcode").find("75000") // OK setarray=new varraySet(setbit) // CRASH crash on second line. olivier From vidal_olivier at yahoo.fr Mon Jan 24 10:07:07 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Mon Jan 24 03:07:09 2005 Subject: SQL and API methods + benchs In-Reply-To: References: Message-ID: <52765DCA-6DE7-11D9-96AD-000D93C35C66@yahoo.fr> Le 24 janv. 05, ? 09:58, Ruslan Zasukhin a ?crit : > On 1/24/05 7:44 AM, "Olivier" wrote: > >> I have a crash if I use: >> SELECT zipcode, cityname, namestreet FROM zipcity, street where >> zipcode = >> '75000' > > And you say that you use BinaryLink for tables? > > Then try also THIS syntax: > > SELECT zipcode, cityname, namestreet > FROM zipcity JOIN street ON linkname > where zipcode = '75000' > > I.e. Specify link name it's done. it's in sent project. > > olivier From vidal_olivier at yahoo.fr Mon Jan 24 10:14:00 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Mon Jan 24 03:14:01 2005 Subject: SQL and API methods + benchs In-Reply-To: References: Message-ID: <48CEBC58-6DE8-11D9-96AD-000D93C35C66@yahoo.fr> >> yes, but it NOT exist FindStartWithAsArraySet or FindLikeAsArraySet or >> FindRegexAsArraySet... > > Right! > Again, > > ArraySet is preferred only of you KNOW that you will find few records > from > 40,000 table records. > > BitSet is preferred in all rest cases. > > FindValue() searches for ONE value. > Exists many chances that will be not many records that have one the > same > value. So in this case ArraySet can be good. > ---------- > Next, in the SQL mashine, we use only bitSets, btw. ok > ---------- > If you going do Intersection for results, then you MUST use the same > types > BitSet + BitSet ok > So in this case use bitSets always. > So Oliver, AS ARRAY SET functions, will be used very rare I think. > Only for specific tasks, when I know that I will find dew records, > When I will not intersect with BitSets. > ok. but for random access ? it's not easy to work with iterator. often, in my app, I use einhugur's datagrid. I need random access. thank you olivier From sunshine at public.kherson.ua Mon Jan 24 11:26:27 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Mon Jan 24 03:26:33 2005 Subject: SQL and API methods + benchs In-Reply-To: <52765DCA-6DE7-11D9-96AD-000D93C35C66@yahoo.fr> Message-ID: On 1/24/05 11:07 AM, "Olivier" wrote: >>> I have a crash if I use: >>> SELECT zipcode, cityname, namestreet FROM zipcity, street where >>> zipcode = >>> '75000' >> >> And you say that you use BinaryLink for tables? >> >> Then try also THIS syntax: >> >> SELECT zipcode, cityname, namestreet >> FROM zipcity JOIN street ON linkname >> where zipcode = '75000' >> >> I.e. Specify link name > > it's done. it's in sent project. And it works or no ? -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From vidal_olivier at yahoo.fr Mon Jan 24 10:28:43 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Mon Jan 24 03:28:47 2005 Subject: SQL and API methods + benchs In-Reply-To: References: Message-ID: <570863F0-6DEA-11D9-96AD-000D93C35C66@yahoo.fr> Le 24 janv. 05, ? 10:26, Ruslan Zasukhin a ?crit : > On 1/24/05 11:07 AM, "Olivier" wrote: > >>>> I have a crash if I use: >>>> SELECT zipcode, cityname, namestreet FROM zipcity, street where >>>> zipcode = >>>> '75000' >>> >>> And you say that you use BinaryLink for tables? >>> >>> Then try also THIS syntax: >>> >>> SELECT zipcode, cityname, namestreet >>> FROM zipcity JOIN street ON linkname >>> where zipcode = '75000' >>> >>> I.e. Specify link name >> >> it's done. it's in sent project. > > And it works or no ? > yes, perfectly. > -- > Best regards, > Ruslan Zasukhin [ I feel the need...the need for speed ] > ------------------------------------------------------------- > e-mail: ruslan@paradigmasoft.com > web: http://www.paradigmasoft.com > > To subscribe to the Valentina mail list go to: > http://lists.macserve.net/mailman/listinfo/valentina > ------------------------------------------------------------- > > > _______________________________________________ > Valentina-beta mailing list > Valentina-beta@lists.macserve.net > http://lists.macserve.net/mailman/listinfo/valentina-beta > From sunshine at public.kherson.ua Mon Jan 24 11:28:31 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Mon Jan 24 03:28:55 2005 Subject: SQL and API methods + benchs In-Reply-To: <48CEBC58-6DE8-11D9-96AD-000D93C35C66@yahoo.fr> Message-ID: On 1/24/05 11:14 AM, "Olivier" wrote: >> ---------- >> Next, in the SQL mashine, we use only bitSets, btw. > ok >> ---------- >> If you going do Intersection for results, then you MUST use the same >> types >> BitSet + BitSet > ok >> So in this case use bitSets always. >> So Oliver, AS ARRAY SET functions, will be used very rare I think. >> Only for specific tasks, when I know that I will find dew records, >> When I will not intersect with BitSets. >> > ok. > but for random access ? > it's not easy to work with iterator. > often, in my app, I use einhugur's datagrid. > I need random access. correction: In the SQL machine we use BitSets while we resolve WHERE clause. When we do ORDER BY we convert final selection into ArraySet, Because only ArraySet can keep sort order. So finally, cursor contains ArraySet and we can do random access. Ok, we need fix converion BitSet to ArraySet. Major operation. -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From vbt at paradigma.ukrcom.kherson.ua Mon Jan 24 10:59:36 2005 From: vbt at paradigma.ukrcom.kherson.ua (vbt@paradigma.ukrcom.kherson.ua) Date: Mon Jan 24 03:29:10 2005 Subject: [V4RB 0000328]: VField.FindLink() now returns VBitSet Message-ID: The following issue has been SUBMITTED. ====================================================================== ====================================================================== Reported By: Ruslan Zasukhin Assigned To: Ruslan Zasukhin ====================================================================== Project: V4RB Issue ID: 328 Category: Fields Reproducibility: always Severity: minor Priority: normal Status: assigned ====================================================================== Date Submitted: 01-24-2005 10:59 EET Last Modified: 01-24-2005 10:59 EET ====================================================================== Summary: VField.FindLink() now returns VBitSet Description: this is good ====================================================================== Issue History Date Modified Username Field Change ====================================================================== 01-24-05 10:59 Ruslan ZasukhinNew Issue ====================================================================== From sunshine at public.kherson.ua Mon Jan 24 11:32:18 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Mon Jan 24 03:32:27 2005 Subject: SQL and API methods + benchs In-Reply-To: <570863F0-6DEA-11D9-96AD-000D93C35C66@yahoo.fr> Message-ID: On 1/24/05 11:28 AM, "Olivier" wrote: >>>> Then try also THIS syntax: >>>> >>>> SELECT zipcode, cityname, namestreet >>>> FROM zipcity JOIN street ON linkname >>>> where zipcode = '75000' >>>> >>>> I.e. Specify link name >>> >>> it's done. it's in sent project. >> >> And it works or no ? >> > > yes, perfectly. Ok, so crash in case we do not specify any link. Hmm, must work. Ivan, you need check this. -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From vidal_olivier at yahoo.fr Mon Jan 24 10:51:09 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Mon Jan 24 03:51:15 2005 Subject: SQL and API methods + benchs In-Reply-To: References: Message-ID: <7935338B-6DED-11D9-96AD-000D93C35C66@yahoo.fr> In my project, in API way, For the SORT, I made a mistake in my reasoning. But I do not see how, in my example, we can sort out the result' search by zip, cty, namestreet. (two tables) (In API) olivier From sunshine at public.kherson.ua Mon Jan 24 11:59:35 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Mon Jan 24 03:59:41 2005 Subject: SQL and API methods + benchs In-Reply-To: <7935338B-6DED-11D9-96AD-000D93C35C66@yahoo.fr> Message-ID: On 1/24/05 11:51 AM, "Olivier" wrote: > In my project, in API way, > > For the SORT, I made a mistake in my reasoning. > But I do not see how, in my example, we can sort out the result' search > by zip, cty, namestreet. (two tables) (In API) 2 tables no way. JOIN of several tables -- this is feature of SQL. For single table you can use Table.Sort() But explain me, what you want to sort in result ? -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From vidal_olivier at yahoo.fr Mon Jan 24 11:21:37 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Mon Jan 24 04:21:53 2005 Subject: SQL and API methods + benchs In-Reply-To: References: Message-ID: Le 24 janv. 05, ? 10:59, Ruslan Zasukhin a ?crit : > 2 tables no way. > > JOIN of several tables -- this is feature of SQL. > > For single table you can use Table.Sort() > > > But explain me, what you want to sort in result ? I have two tables : One table with zip codes and cities. A table with streets. the two tables are binary-linked. I make a search. For example I want to have all the streets of the country. I have a listbox where I display all the streets of the country. This listbox has three columns: zip, city, namestreet. I shall want that this listbox is sorted out 1-by zip, 2-by city, 3-by street, so that the reading is easy. 33000 BORDEAUX street of Odessa 75001 PARIS street of Kiev 75001 PARIS street of Paradigma 75002 PARIS street of Youri Gagarine I reports me that the API is useful only in some cases. I wished to modify my project to use only the API in ALL my project. It was exciting. I had begun to modify all my project for that. But I made an error, I believe that I am rather going to use the Valentina' SQL and to use maybe the API that in certain cases. For BINARY LINK, What are the advantages with regard to objectPtr? thank you olivier From ralf at end-if.de Mon Jan 24 11:29:06 2005 From: ralf at end-if.de (Ralf Sander) Date: Mon Jan 24 04:29:21 2005 Subject: Llong and ULlong datatypes ( was:Re: OID field) In-Reply-To: References: Message-ID: No feature request so far. So I created one. Please vote! Cheers, Ralf On 22.01.2005, at 08:32, Ruslan Zasukhin wrote: > On 1/22/05 2:19 AM, "Ralf Sander" wrote: > >> Shouldn't a double do the job in RB? > > No, double have other structure. > So it will be wrong. > > We need again and again press REAL to introduce longlong at least in > RB 6 > And I do not understand them. This is so simple. This is native > computer > type. > > > >> Cheers, >> Ralf >> >>> On 1/21/05 4:39 PM, "End If Software - Ralf Sander" >>> wrote: >>> >>>> Hi Ruslan, >>>> >>>> is the OID field only accessible via sql? No Api like the RecID >>>> field? >>> >>> Table.OID ? >>> >>> Problem is that REALbiasic and Director both do not support Llong >>> values. >>> >>> You can access this field as >>> >>> Table.Field("OID").GetString > > -- > Best regards, > Ruslan Zasukhin [ I feel the need...the need for speed ] > ------------------------------------------------------------- > e-mail: ruslan@paradigmasoft.com > web: http://www.paradigmasoft.com > > To subscribe to the Valentina mail list go to: > http://lists.macserve.net/mailman/listinfo/valentina > ------------------------------------------------------------- > > > _______________________________________________ > Valentina-beta mailing list > Valentina-beta@lists.macserve.net > http://lists.macserve.net/mailman/listinfo/valentina-beta > From vidal_olivier at yahoo.fr Mon Jan 24 11:30:26 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Mon Jan 24 04:30:28 2005 Subject: SQL and API methods + benchs In-Reply-To: References: Message-ID: >> At the moment, SQL is much faster. Maybe that I have badly to >> optimize my code >> for the API. > > I also think so. It needs study your task. > For now quick notes: > > 1) have you bench with DebugLevel = 0 ? > I try with. That changes nothing > 2) as far as I see, in the SQL you SELECT all fields of table Zipcity, > Right? > > Actually SQL way have interesting feature: > Let you have table with 20 fields, > but you select only 4 fields. > > then Iteration of cursor will be 5 times faster of iteration > of original table, because cursor will read only that 4 fields. > interesting, but Here that does not modify the bench. There are only two fields in the table. Both are selected. SQL and API are thus equally on this point. > For now we do not have way do the same trick in API. > (in C++ we have in fact, just did not opened to V4RB.) > > 3) API way has other name -- Navigational way. > this means that you should walk from record to record. > this means that you can touch ONLY really required records. > > SQL way in contrast is oriented on the work with SETS of records. > you find set of records WHERE. > you UPDATE set of records WHERE. > you join SET1 to SET2 of T1 and T2.. > and so on. > > This is the main difference in the ideology of these ways. > Valentina although try optimize API way to SETs also. > The old Navigational dbs (like dbVista) cannot do this absolutely. Thus if I want to seek and to display in a listbox a big set of records, I use rather SQL? What is strange in my example, it is that SQL is much faster. The API is thus faster only for certain processings? thank you olivier > From sunshine at public.kherson.ua Mon Jan 24 12:44:43 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Mon Jan 24 04:44:49 2005 Subject: Llong and ULlong datatypes ( was:Re: OID field) In-Reply-To: Message-ID: On 1/24/05 12:29 PM, "Ralf Sander" wrote: > No feature request so far. No no Ralf! There is such feature request!!! More than year. You should find old one and add vote > So I created one. Please vote! > > > Cheers, > Ralf -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From jda at his.com Mon Jan 24 07:47:02 2005 From: jda at his.com (jda) Date: Mon Jan 24 06:47:19 2005 Subject: STATUS In-Reply-To: References: Message-ID: Hi Ruslan, Thank you for the update. > > >starting from prev Wednesday we have stick in refactoring of our sources >To break all dependencies between (V4RB/V4MD) and Kernel/Client dlls. > Does this have any direct implications for developers (for example, letting us group the Valentina dlls in folders with our apps), or is it just a Good Thing? Jon From sunshine at public.kherson.ua Mon Jan 24 14:55:37 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Mon Jan 24 06:55:46 2005 Subject: STATUS In-Reply-To: Message-ID: On 1/24/05 2:47 PM, "jda" wrote: > Hi Ruslan, > > Thank you for the update. > >> >> >> starting from prev Wednesday we have stick in refactoring of our sources >> To break all dependencies between (V4RB/V4MD) and Kernel/Client dlls. >> > > Does this have any direct implications for developers (for example, > letting us group the Valentina dlls in folders with our apps), or is > it just a Good Thing? This is the PROMISED feature, that you can remove kernel OR client dll, And everything must work. Problem was that we could not remove kernel.dll This feature is important for Vserver + VClient. -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From ralf at end-if.de Mon Jan 24 13:57:35 2005 From: ralf at end-if.de (Ralf Sander) Date: Mon Jan 24 06:57:35 2005 Subject: Llong and ULlong datatypes ( was:Re: OID field) In-Reply-To: References: Message-ID: <84955E72-6E07-11D9-B026-000A95DF4532@end-if.de> OK, not easy to find, as they don't have a full text search. Here it is: Cheers, Ralf On 24.01.2005, at 11:44, Ruslan Zasukhin wrote: > On 1/24/05 12:29 PM, "Ralf Sander" wrote: > >> No feature request so far. > > No no Ralf! > > There is such feature request!!! > More than year. > > You should find old one and add vote > >> So I created one. Please vote! >> > reportid=huwhzwle> >> >> Cheers, >> Ralf > > -- > Best regards, > Ruslan Zasukhin [ I feel the need...the need for speed ] > ------------------------------------------------------------- > e-mail: ruslan@paradigmasoft.com > web: http://www.paradigmasoft.com > > To subscribe to the Valentina mail list go to: > http://lists.macserve.net/mailman/listinfo/valentina > ------------------------------------------------------------- > > > _______________________________________________ > Valentina-beta mailing list > Valentina-beta@lists.macserve.net > http://lists.macserve.net/mailman/listinfo/valentina-beta > -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: text/enriched Size: 1240 bytes Desc: not available Url : http://lists.macserve.net/pipermail/valentina-beta/attachments/20050124/2e423a26/attachment.bin From sunshine at public.kherson.ua Mon Jan 24 15:24:07 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Mon Jan 24 07:24:12 2005 Subject: Llong and ULlong datatypes ( was:Re: OID field) In-Reply-To: <84955E72-6E07-11D9-B026-000A95DF4532@end-if.de> Message-ID: On 1/24/05 2:57 PM, "Ralf Sander" wrote: > OK, not easy to find, as they don't have a full text search. > Here it is: > > Cheers, > Ralf Right. 86 votes! Interesting how many they need to see it ?! Btw, you can promote this link on RB lists again. :-) -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From vidal_olivier at yahoo.fr Mon Jan 24 14:28:03 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Mon Jan 24 07:28:20 2005 Subject: BinaryLink vs ObectPtr Message-ID: Hi Ruslan and list, In my tests to know the advantages of the binary links on objetPtr, I made a small bench. (in doc, the advantages and Inconvenience of the various ways are not explained) DB with BinaryLink: SELECT * FROM zipcity join street we linkcitystreet where zipcode = '75000' DB with ObjectPtr: SELECT * FROM zipcity, street where street.zipcityptr=zipcity.recid and zipcode = '75000' ( If I does not put "street.zipcityptr=zipcity.recid", there is a CRASH) BinaryLink is faster of approximately 10 %. It is little. When Valentina 2 will be optimized, both ways will approximately be so fast or binarylink will be even faster? Later, to use binaryLink will allow more possibilities? thank you olivier From vidal_olivier at yahoo.fr Mon Jan 24 14:37:27 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Mon Jan 24 07:37:29 2005 Subject: BinaryLink vs ObectPtr In-Reply-To: References: Message-ID: <16BF9A31-6E0D-11D9-96AD-000D93C35C66@yahoo.fr> > > BinaryLink is faster of approximately 10 %. I wanted to say 15 %. 15 %, not 10 %. olivier From sunshine at public.kherson.ua Mon Jan 24 16:03:22 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Mon Jan 24 08:03:30 2005 Subject: BinaryLink vs ObectPtr In-Reply-To: Message-ID: On 1/24/05 3:28 PM, "Olivier" wrote: > Hi Ruslan and list, > > In my tests to know the advantages of the binary links on objetPtr, I > made a small bench. > (in doc, the advantages and Inconvenience of the various ways are not > explained) > > DB with BinaryLink: > SELECT * FROM zipcity join street we linkcitystreet where zipcode = > '75000' > > DB with ObjectPtr: > SELECT * FROM zipcity, street where street.zipcityptr=zipcity.recid and > zipcode = '75000' > ( If I does not put "street.zipcityptr=zipcity.recid", there is a CRASH) > > BinaryLink is faster of approximately 10 %. It is little. > When Valentina 2 will be optimized, both ways will approximately be so > fast or binarylink will be even faster? > Later, to use binaryLink will allow more possibilities? I have describe this to Lynn, not so far. Let me check my letters. I will copy here parts of my texts: ----------------------------------------------------------------------- On 12/30/04 11:49 PM, "Lynn Fredricks" wrote: > Hi Ruslan, > > There is a big section in your notes about Abstraction Links, which mentions > Foreign Keys, ObjPtr Fields and Binary Links. Are there any other kinds of > links or keys? Of course, we support a primary key, right? Yes we support PRIMARY KEY. This is Relational model. PK + FOREIGN KEY = this is Relational model. This is the only what have all other RDBMS. Valentina 1.x also did have ObjectPtr, Although 1.x did not support PK+FK completely. Valentina 2.0 support 3 kinds of links. FK ObjectPtr BinaryLink. ObjectPtr -- this is analog of DIRECT POINTER from Navigational model, although it differ. FK -- this is pointer by value. BianryLink -- this is totally new thing. Nobody have it. it is best then both above ----------------------------------------------------------------------- On 12/30/04 11:55 PM, "Lynn Fredricks" wrote: > Hi Ruslan, > > In your doc, you mention that ObjectPtr fields are up to 4 x faster than > foreign keys. But what are the advantages of using a Binary Link (is there > anything other than M:M)? * M : M this is already HUGE advantage! * I am going write many text on this in docs, because this thing is not describe in books. * another main advantage is DESIGN of db structure. Look, with FK and ObjectPtr to link 2 tables you must insert FK field into one of them, which play role of pointer to the parent record. The you have T1 which point T2. T1 have 3 own fields that describe it info (its real world object). Now you need add 1 pointer field to establish link. If around T1 you have 5 tables which must be linked, then you will have 3 + 5 fields in T1. Look, 3 fields which keep ESSENTIAL DATA of this table, And 5 fields which keep LINK information. LINK information is NOT a property of T1 object, But Relational model push you insert this info into T1. This means that IF you change db structure, you will MUST change structure of already existed tables to add links. It is obvious that this is bad. All programming languages try isolate one part of code from other. BinaryLink (as well as Link abstraction) DO THIS for DBMS world. You will have in T1 3 fields. Always. (of course if you use BinaryLinks only :-) If you add/remove linked tables, then you just add T3 and new BinaryLink. T1 is not touched. ----------------------------------------------------------------------- >> FK >> ObjectPtr >> BinaryLink. >> >> ObjectPtr -- this is analog of DIRECT POINTER from Navigational model, >> although it differ. >> >> FK -- this is pointer by value. >> >> BianryLink -- this is totally new thing. Nobody have it. >> it is best then both above > > Best for what? Speed? That is what I need to understand -- in marketing 2.0, > I need to convey why this new type is better. SPEED also. And BinaryLink will use about 2 times less disk space then trick with third table to establish M : M link as do RDBMS world. ----------------------------------------------------------------------- >> The you have T1 which point T2. >> T1 have 3 own fields that describe it info (its real world object). >> >> Now you need add 1 pointer field to establish link. >> If around T1 you have 5 tables which must be linked, then you >> will have 3 + 5 fields in T1. >> >> Look, 3 fields which keep ESSENTIAL DATA of this table, And 5 >> fields which keep LINK information. > > There is where Im getting confused. > > There are three fields of real world data in T1. > > There are five tables which must be linked to T1, so therefore, T1 has 3 > fields (of real data) + 5 link fields (to connect to each of the five other > tables) = 8 fields. Correct. And in RDBMS, these 5 fields are FOREIGN KEY fields. Also note, may happens that 5 tables POINT TO T1, In this case each of them get this additional FK field, but T1 have 3 fields. In any case, db have 5 more fields in total. >> LINK information is NOT a property of T1 object, But >> Relational model push you insert this info into T1. > > With Binary Links, where is the information stored to make the link, if it > is not in T1? It is stored in the NEW SPECIAL structure -- Binary Link. Internally this is NOT a table. It is quite other format (similar to indexes). -------------------------------------- NOTE: in RDBS, developer also can have T1 with 3 fields, and T2..T6 tables around it. to get better design developer can make links as additional tables with 2 fields each: T12 -- link T1 and T2 T13 -- link T1 and T3 ... in case of M : M developer simply MUST to use additional table. NOTE2: in pure Relational model exists ONLY tables. even indexes not exists, so strictly speaking all RDBMS as MS, mySQL are not pure Relational. * So we do step forward and add one more abstraction LINK. * and we have 3 kinds of links. So as I have told, BinaryLink is not a T12 table. This is special kind structure. It is smaller and faster. And the main, it act as abstraction LINK. >> This means that IF you change db structure, you will MUST >> change structure of already existed tables to add links. >> >> It is obvious that this is bad. >> All programming languages try isolate one part of code from other. >> >> >> BinaryLink (as well as Link abstraction) DO THIS for DBMS world. >> You will have in T1 3 fields. Always. >> (of course if you use BinaryLinks only :-) >> >> If you add/remove linked tables, then you just add T3 and new >> BinaryLink. T1 is not touched. > > So is the Binary Link data stored in its own sort of table and contain all > information related to the linked tables? Yes. Close to true. Just it is not table. You can think about Binary link as array of pairs { 3, 5 } rec 3 of T1 is linked to rec 5 of T2. { 3, 7 } > And so if this is true, then no > linking information needs to be stored in T1, right? Right. From vidal_olivier at yahoo.fr Mon Jan 24 15:29:04 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Mon Jan 24 08:29:12 2005 Subject: BinaryLink vs ObectPtr In-Reply-To: References: Message-ID: <4C90D52E-6E14-11D9-96AD-000D93C35C66@yahoo.fr> OK, thank you VERY much for all this precision! It is important to know WHERE we go. I am going to use the binary links in my project. Good work and still thank you. olivier Le 24 janv. 05, ? 15:03, Ruslan Zasukhin a ?crit : > On 1/24/05 3:28 PM, "Olivier" wrote: > >> Hi Ruslan and list, >> >> In my tests to know the advantages of the binary links on objetPtr, I >> made a small bench. >> (in doc, the advantages and Inconvenience of the various ways are not >> explained) >> >> DB with BinaryLink: >> SELECT * FROM zipcity join street we linkcitystreet where zipcode = >> '75000' >> >> DB with ObjectPtr: >> SELECT * FROM zipcity, street where street.zipcityptr=zipcity.recid >> and >> zipcode = '75000' >> ( If I does not put "street.zipcityptr=zipcity.recid", there is a >> CRASH) >> >> BinaryLink is faster of approximately 10 %. It is little. >> When Valentina 2 will be optimized, both ways will approximately be so >> fast or binarylink will be even faster? >> Later, to use binaryLink will allow more possibilities? > > I have describe this to Lynn, not so far. Let me check my letters. > I will copy here parts of my texts: > > > ----------------------------------------------------------------------- > On 12/30/04 11:49 PM, "Lynn Fredricks" > wrote: > >> Hi Ruslan, >> >> There is a big section in your notes about Abstraction Links, which >> mentions >> Foreign Keys, ObjPtr Fields and Binary Links. Are there any other >> kinds of >> links or keys? Of course, we support a primary key, right? > > Yes we support PRIMARY KEY. This is Relational model. > > PK + FOREIGN KEY = this is Relational model. > This is the only what have all other RDBMS. > > > Valentina 1.x also did have ObjectPtr, > Although 1.x did not support PK+FK completely. > > Valentina 2.0 support 3 kinds of links. > > FK > ObjectPtr > BinaryLink. > > ObjectPtr -- this is analog of DIRECT POINTER from Navigational model, > although it differ. > > FK -- this is pointer by value. > > BianryLink -- this is totally new thing. Nobody have it. > it is best then both above > > ----------------------------------------------------------------------- > > On 12/30/04 11:55 PM, "Lynn Fredricks" > wrote: > >> Hi Ruslan, >> >> In your doc, you mention that ObjectPtr fields are up to 4 x faster >> than >> foreign keys. But what are the advantages of using a Binary Link (is >> there >> anything other than M:M)? > > * M : M this is already HUGE advantage! > > * I am going write many text on this in docs, because this thing is not > describe in books. > > * another main advantage is DESIGN of db structure. > > Look, with FK and ObjectPtr to link 2 tables you must insert FK field > into > one of them, which play role of pointer to the parent record. > > The you have T1 which point T2. > T1 have 3 own fields that describe it info (its real world object). > > Now you need add 1 pointer field to establish link. > If around T1 you have 5 tables which must be linked, > then you will have 3 + 5 fields in T1. > > Look, 3 fields which keep ESSENTIAL DATA of this table, > And 5 fields which keep LINK information. > > LINK information is NOT a property of T1 object, > But Relational model push you insert this info into T1. > > This means that IF you change db structure, you will MUST change > structure > of already existed tables to add links. > > It is obvious that this is bad. > All programming languages try isolate one part of code from other. > > > BinaryLink (as well as Link abstraction) DO THIS for DBMS world. > You will have in T1 3 fields. Always. > (of course if you use BinaryLinks only :-) > > If you add/remove linked tables, then you just add T3 and new > BinaryLink. > T1 is not touched. > > ----------------------------------------------------------------------- >>> FK >>> ObjectPtr >>> BinaryLink. >>> >>> ObjectPtr -- this is analog of DIRECT POINTER from Navigational >>> model, >>> although it differ. >>> >>> FK -- this is pointer by value. >>> >>> BianryLink -- this is totally new thing. Nobody have it. >>> it is best then both above >> >> Best for what? Speed? That is what I need to understand -- in >> marketing 2.0, >> I need to convey why this new type is better. > > SPEED also. > > And BinaryLink will use about 2 times less disk space then trick with > third table to establish M : M link as do RDBMS world. > > > ----------------------------------------------------------------------- >>> The you have T1 which point T2. >>> T1 have 3 own fields that describe it info (its real world object). >>> >>> Now you need add 1 pointer field to establish link. >>> If around T1 you have 5 tables which must be linked, then you >>> will have 3 + 5 fields in T1. >>> >>> Look, 3 fields which keep ESSENTIAL DATA of this table, And 5 >>> fields which keep LINK information. >> >> There is where Im getting confused. >> >> There are three fields of real world data in T1. >> >> There are five tables which must be linked to T1, so therefore, T1 >> has 3 >> fields (of real data) + 5 link fields (to connect to each of the five >> other >> tables) = 8 fields. > > Correct. And in RDBMS, these 5 fields are FOREIGN KEY fields. > > Also note, may happens that 5 tables POINT TO T1, > In this case each of them get this additional FK field, > but T1 have 3 fields. In any case, db have 5 more fields in total. > > >>> LINK information is NOT a property of T1 object, But >>> Relational model push you insert this info into T1. >> >> With Binary Links, where is the information stored to make the link, >> if it >> is not in T1? > > It is stored in the NEW SPECIAL structure -- Binary Link. > Internally this is NOT a table. It is quite other format (similar to > indexes). > > -------------------------------------- > NOTE: in RDBS, developer also can have > > T1 with 3 fields, and T2..T6 tables around it. > > to get better design developer can make links as > additional tables with 2 fields each: > T12 -- link T1 and T2 > T13 -- link T1 and T3 > ... > > in case of M : M developer simply MUST to use > additional table. > > NOTE2: > in pure Relational model exists ONLY tables. > even indexes not exists, so strictly speaking > all RDBMS as MS, mySQL are not pure Relational. > > > * So we do step forward and add one more abstraction LINK. > * and we have 3 kinds of links. > > > So as I have told, BinaryLink is not a T12 table. > This is special kind structure. It is smaller and faster. > And the main, it act as abstraction LINK. > >>> This means that IF you change db structure, you will MUST >>> change structure of already existed tables to add links. >>> >>> It is obvious that this is bad. >>> All programming languages try isolate one part of code from other. >>> >>> >>> BinaryLink (as well as Link abstraction) DO THIS for DBMS world. >>> You will have in T1 3 fields. Always. >>> (of course if you use BinaryLinks only :-) >>> >>> If you add/remove linked tables, then you just add T3 and new >>> BinaryLink. T1 is not touched. >> >> So is the Binary Link data stored in its own sort of table and >> contain all >> information related to the linked tables? > > Yes. Close to true. Just it is not table. > > You can think about Binary link as array of pairs > > { 3, 5 } rec 3 of T1 is linked to rec 5 of T2. > { 3, 7 } > > >> And so if this is true, then no >> linking information needs to be stored in T1, right? > > Right. > > > _______________________________________________ > Valentina-beta mailing list > Valentina-beta@lists.macserve.net > http://lists.macserve.net/mailman/listinfo/valentina-beta > From AYU8 at CDC.GOV Mon Jan 24 15:23:00 2005 From: AYU8 at CDC.GOV (Sims, John) Date: Mon Jan 24 14:23:42 2005 Subject: Binary Link Owner Message-ID: Hi Ruslan, I hope you are having a good day. In the documentation regarding Binary Links I found this statement regarding ownership of records as it pertains to deletion: In rest cases (1:1 and M:M) developer can assign to some table to be owner. However, I do not see in the documentation exactly how we do this. In these cases is the Left Table always the owner or is there something missing from the documentation (or am I just looking right over it --- highly possible)? Thanks. -John -------------- next part -------------- An HTML attachment was scrubbed... URL: http://lists.macserve.net/pipermail/valentina-beta/attachments/20050124/a918b3dc/attachment.html From sunshine at public.kherson.ua Mon Jan 24 23:09:34 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Mon Jan 24 15:09:39 2005 Subject: Binary Link Owner In-Reply-To: Message-ID: On 1/24/05 10:23 PM, "Sims, John" wrote: > Hi Ruslan, > > I hope you are having a good day. > > In the documentation regarding Binary Links I found this statement regarding > ownership of records as it pertains to deletion: > > In rest cases (1:1 and M:M) developer can assign to some table to be > owner. > > However, I do not see in the documentation exactly how we do this. Vlink.owner = tblA > In these > cases is the Left Table always the owner or is there something missing from > the documentation (or am I just looking right over it --- highly possible)? Owner is important IF you want to use e.g. CASCADE deletion. Then when you delete record of owner, the child records are deleted. Since, 1 : 1 and M : M are symmetric, We need assign owner manually. But Hmm, I do not remember if the first table is marked to be owner on default....we need check this. -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From sunshine at public.kherson.ua Tue Jan 25 01:02:39 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Mon Jan 24 17:02:55 2005 Subject: [V4RB] 2.0 b38 uploaded Message-ID: kernel - 2.0 b38 ================ - 0000333: [API] Valentina.GetVersion() for 2.0 should return 0x200. (Ruslan Zasukhin) - 0000322: [SQL Parser] SQL syntax for Diagnose comand. (Sergey Duminskiy) - 0000332: [SQL nodes] Crash if join-query where clause does not contain join specification. (Ivan Smahin) - 0000329: [Fields] ObjectPtr register self as Link in tables only - not in db. (Ivan Smahin) - 0000323: [Methods (calculated fields)] There is no unique check for unique methods during insert/update. (Ivan Smahin) - 0000321: [Tables] We get a crash trying to drop a TEMPORARY table. (Ivan Smahin) - 0000319: [Fields] OID now returns values as expected. (Ruslan Zasukhin) - 0000317: [Database] Wrong segment size after Convert_1_2. (Ivan Smahin) V4RB - 2.0 b38 ============== - 0000330: [SQL] crash on SQL query without Link conditions (Ivan Smahin) - 0000326: [SQL] droprecord -> locked records (Ivan Smahin) - 0000281: [API] UpdateRecord Crash on large table (Igor Gomon) - 0000328: [Fields] VField.FindLink() now returns VBitSet (Ruslan Zasukhin) - 0000324: [Fields] Flag "UNIQUE" doesn't work (Ruslan Zasukhin) -------------- next part -------------- An HTML attachment was scrubbed... URL: http://lists.macserve.net/pipermail/valentina-beta/attachments/20050125/ddd39455/attachment-0001.html From sunshine at public.kherson.ua Tue Jan 25 01:12:30 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Mon Jan 24 17:12:33 2005 Subject: [V4RB] 2.0 b38 uploaded In-Reply-To: Message-ID: On 1/25/05 1:02 AM, "Ruslan Zasukhin" wrote: Forget one more thing: Oliver, * added VLink( inset as Vset, Ta, TB ) * fixed crashes in the constructors VBitSet( ArraySet ), AraySet( BitSet ) -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From sunshine at public.kherson.ua Tue Jan 25 01:33:51 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Mon Jan 24 17:33:57 2005 Subject: V4RB b38 + RB Stadnard Message-ID: Hi All, If somebody use RB Standard he will not be able compile app with b38 Because we have plug support of RBDB API. In b39 we will give you way TURN OFF the RBDB API in the plugn. -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From jda at his.com Mon Jan 24 19:33:14 2005 From: jda at his.com (jda) Date: Mon Jan 24 18:33:25 2005 Subject: V4RB crashes on launch In-Reply-To: References: Message-ID: I trashed my preferences and all old databases. Here's the crash report: Thread 0 Crashed: 0 vshared_carbon_cw.shlb 0x000f3bb4 0xf2000 + 0x1bb4 1 Kernel_Carbon_Final.shlb 0x0205472c get_Path__Q23fbl20Location_Disk_FSSpecCFv + 0x74 2 Kernel_Carbon_Final.shlb 0x023d8090 Valentina_Init_ICU__3fblFv + 0x24 3 Kernel_Carbon_Final.shlb 0x023d7bd4 ValentinaInit__3fblFUlPcPc + 0x3c 4 Kernel_Carbon_Final.shlb 0x023d7b58 ValentinaInitL__3fblFUlPcPc + 0x10 5 Plugin 0x00528bf0 V4RB_RegisterErrorCodes__Fv + 0x2c 6 Plugin 0x0050f638 PluginEntry + 0x24c Jon From macsforever2000 at axint.net Mon Jan 24 19:01:25 2005 From: macsforever2000 at axint.net (Frank Schima) Date: Mon Jan 24 20:01:42 2005 Subject: [V4RB] 2.0 b38 uploaded In-Reply-To: References: Message-ID: <04D7D1CA-6E75-11D9-BFA0-0003939246BC@axint.net> Hi Ruslan, On Jan 24, 2005, at 4:02 PM, Ruslan Zasukhin wrote: > - 0000281: [API] UpdateRecord Crash on large table (Igor Gomon) It still crashes for me. I tried my sample test project and it totally locks up my computer when it crashes! Are you sure it works for you? Best regards, Frank Schima From vidal_olivier at yahoo.fr Tue Jan 25 07:36:22 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Tue Jan 25 00:36:40 2005 Subject: [V4RB] 2.0 b38 uploaded In-Reply-To: References: Message-ID: <6E14864A-6E9B-11D9-96AD-000D93C35C66@yahoo.fr> Le 25 janv. 05, ? 00:12, Ruslan Zasukhin a ?crit : > On 1/25/05 1:02 AM, "Ruslan Zasukhin" > wrote: > > Forget one more thing: > > Oliver, > > * added VLink( inset as Vset, Ta, TB ) > perfect ! > * fixed crashes in the constructors > VBitSet( ArraySet ), > AraySet( BitSet ) > perfect ! thank you olivier From vidal_olivier at yahoo.fr Tue Jan 25 07:52:17 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Tue Jan 25 00:52:21 2005 Subject: problem with SQL query on binaryLink Message-ID: Hi Ruslan and list, The mantis is inalienable this morning, then I detail this bug (on b37 AND b38) here: In a DB with two tables linked by binarylinks (create by the small project sent yesterday). City table : 40 002 records Streets table : 100 000 records 30 000 records of table street are linked to 1 record of table city. 2 records of table street are linked to 1 OTHER record of table city. This 2 records of table city have a stringfield (" zipcode ") = '75000' If I launch the following SQL query (with my app or ViSQL), the FIRST time, after the opening of the DB: SELECT * FROM ZipCity join street on linkcitystreet where zipcode = '75000' Result: 2 records ( FALSE) If I relaunch the same SQL query SECOND time (and ALL the times later): Result: 30 002 records ( TRUE) I suppose that it is a problem with the cache. What is surprising, it is that I tried to reproduce the problem with a small DB. There everything works without problem. I send DB to Ruslan. thank you Olivier From vidal_olivier at yahoo.fr Tue Jan 25 08:07:58 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Tue Jan 25 01:08:03 2005 Subject: ViSQL much faster ! good ! Message-ID: Hi, I have the impression that ViSQL (in folder of b38) is much faster now to open a DB. And maybe also for the display of the results. Very good olivier From sunshine at public.kherson.ua Tue Jan 25 09:50:55 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Tue Jan 25 01:51:03 2005 Subject: [V4RB] 2.0 b38 uploaded In-Reply-To: <04D7D1CA-6E75-11D9-BFA0-0003939246BC@axint.net> Message-ID: On 1/25/05 4:01 AM, "Frank Schima" wrote: > Hi Ruslan, > > > On Jan 24, 2005, at 4:02 PM, Ruslan Zasukhin wrote: > >> - 0000281: [API] UpdateRecord Crash on large table (Igor Gomon) > > It still crashes for me. I tried my sample test project and it totally > locks up my computer when it crashes! > > Are you sure it works for you? Yes it did work. Although I have comment the first 2 loops, because they gone fine... Strange. -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From sunshine at public.kherson.ua Tue Jan 25 09:58:17 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Tue Jan 25 01:58:29 2005 Subject: V4RB crashes on launch In-Reply-To: Message-ID: On 1/25/05 2:33 AM, "jda" wrote: > I trashed my preferences and all old databases. Here's the crash report: > > Thread 0 Crashed: > 0 vshared_carbon_cw.shlb 0x000f3bb4 0xf2000 + 0x1bb4 > 1 Kernel_Carbon_Final.shlb 0x0205472c > get_Path__Q23fbl20Location_Disk_FSSpecCFv + 0x74 > 2 Kernel_Carbon_Final.shlb 0x023d8090 > Valentina_Init_ICU__3fblFv + 0x24 > 3 Kernel_Carbon_Final.shlb 0x023d7bd4 > ValentinaInit__3fblFUlPcPc + 0x3c > 4 Kernel_Carbon_Final.shlb 0x023d7b58 > ValentinaInitL__3fblFUlPcPc + 0x10 > 5 Plugin 0x00528bf0 > V4RB_RegisterErrorCodes__Fv + 0x2c > 6 Plugin 0x0050f638 PluginEntry + 0x24c > > > Jon As far as I see, Frank and Oliver was able to run b38... -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From sunshine at public.kherson.ua Tue Jan 25 10:27:21 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Tue Jan 25 02:27:29 2005 Subject: API methods on binary link // Philosophy (!!) In-Reply-To: Message-ID: On 1/22/05 8:37 PM, "Olivier" wrote: Hi Oliver, It seems to me, that your task better solve in the SQL. You need do joins, you need sort joins, you need display results of joins to user in the table. Navigational model, as I have told, oriented on other tasks. It is good for tasks when algorithm require navigation from record to record, to do some calculations... Or may be you need change your mind a little, and way how work your app. *********************************************** ** Now I want to make VERY IMPORTANT points: ** *********************************************** * the main idea of relational model is that you split data into different tables, then you combine them back using JOINs. * how looks join usually? One record of parent table is MANY times joined with records from child table. T1.f1 T1.f2 T2.aaa1 T2.aaa2 T1.f1 T1.f2 T2.bbb1 T2.bbb2 T1.f1 T1.f2 T2.ccc1 T2.ccc2 T1.f1 T1.f2 T2.ddd1 T2.ddd2 * What you get as join table? apples + oranges Look on your table: Paris + street But may be it is better to show to user list of cities: AND the parent list of streets? In this case PARIS will present only once on monitor. And streets will be show in other datagrid. This solution CAN BE much better because 1) you do not spend time to build JOIN table. 2) you do not eat any bytes for JOIN. 3) on monitor you eat less space for duplicates of PARIS. 4) it will work much faster, because you - display at first cities. - IF user click some city then you FIND only linked streets. You see? You do much less of work!!!! > Hi Ruslan and list, > > I have the impression that it is not easy to use the API when we want to > obtain sorts of temporary tables (consisted of fields of several linked > tables). It is maybe because I still thinks in term of SQL. > > For example, always with my tables of zip / city and streets. > > I have a datagrid which has to display in the opening of the window a list > with three columns. First column: zip (table ZipCity) The second column: city > (table ZipCity) The third column: street (table Streets) > > It has to display all the streets with the zip beginning with 75. The list > must be sorted out by zip, city, street. > > Ex: 75001 PARIS street of Kiev 75001 PARIS street Clemenceau 75001 > PARIS place Vend?me 75002 PARIS impasse of Bordeaux > > The first one and the second column are thus two fields of the table > "zip/city" The third column is a field of the table "streets". > > In SQL: > > SELECT zip, city, street FROM ZipCityTb, StreetTb WHERE left (zip, 2) = '75' > ORDER by zip, city, street > > We obtain then a cursor with which I can easily fill my datagrid. Especially, > the cursor keeps a number for all the line " zip, city, street ", even if > originally they are two different tables. It is very practical to modify > couples " zipCity-Streets " for example, because the couple is kept in 1 line > of the cursor. > > But with the sets it is less evident because 1 set is attached to 1 single > table. > > The sets are perfected to work on a table but on any? In my example, I do not > manage to obtain a practical code. > > please, How would you make to make an equivalent of this SQL query? -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From sunshine at public.kherson.ua Tue Jan 25 10:38:32 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Tue Jan 25 02:38:37 2005 Subject: SQL and API methods + benchs // Philosophy of API way. More In-Reply-To: Message-ID: On 1/24/05 12:30 PM, "Olivier" wrote: > What is strange in my example, it is that SQL is much faster. > The API is thus faster only for certain processings? Oliver, In my prev letter I have show the main difference of SQL and Navigational worlds. SQL *MUST* do joins. JOIN in fact is very heavy operation! If you will correctly make GUI of app. Then Navigational way can do much less of work to get the same data to display for user. btw, have you see 4D ? 4D and FM have so-called Related-View. Remember? It is easy to see that this is because both 4D and FM, do not use SQL way. ** Your main problem is that you really try find *equivalent* of SQL In the API way. But this is bad idea. API way this is OTHER world. It require different GUI, different coding, different thinking. -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From sunshine at public.kherson.ua Tue Jan 25 11:29:40 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Tue Jan 25 03:29:52 2005 Subject: [V4RB] 2.0 b38 uploaded In-Reply-To: <04D7D1CA-6E75-11D9-BFA0-0003939246BC@axint.net> Message-ID: On 1/25/05 4:01 AM, "Frank Schima" wrote: > Hi Ruslan, > > > On Jan 24, 2005, at 4:02 PM, Ruslan Zasukhin wrote: > >> - 0000281: [API] UpdateRecord Crash on large table (Igor Gomon) > > It still crashes for me. I tried my sample test project and it totally > locks up my computer when it crashes! > > Are you sure it works for you? Frank, Once again I have run your project AS IS, With the release version of b38. It works here. May be you try to run your SECOND project ? -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From vidal_olivier at yahoo.fr Tue Jan 25 11:05:31 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Tue Jan 25 04:05:54 2005 Subject: API methods on binary link // Philosophy (!!) In-Reply-To: References: Message-ID: Thank you very much Ruslan for all these explanations. I have to think really about my application. It is VERY attractive.. But I have to think seriously about all this. I shall not like, with THIS model, having to redo all structures of my tables or to have now REDUNDANT data. In fact, I am afraid of being more limited than in SQL. > > * the main idea of relational model is that you split data into > different > tables, not in navigationnal model ? > then you combine them back using JOINs. > > * how looks join usually? One record of parent table is MANY times > joined > with records from child table. > > Look on your table: > Paris + street > > But may be it is better to show to user > list of cities: > AND the parent list of streets? yes, it's possible. But if, for example, I have to print thousands of labels of addresses (to make a mailing)? The join is necessary here because every address is linked to a zip-city and because the zip-city must be printed on every address? Unless we have redundant tables ?! I have to think about all this. still thank you, olivier -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: text/enriched Size: 1162 bytes Desc: not available Url : http://lists.macserve.net/pipermail/valentina-beta/attachments/20050125/b1b7d21d/attachment.bin From sunshine at public.kherson.ua Tue Jan 25 12:39:53 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Tue Jan 25 04:40:05 2005 Subject: API methods on binary link // Philosophy (!!) In-Reply-To: Message-ID: On 1/25/05 12:05 PM, "Olivier" wrote: > Thank you very much Ruslan for all these explanations. > > I have to think really about my application. > It is VERY attractive.. > But I have to think seriously about all this. > I shall not like, with THIS model, having to redo all structures of my tables > or to have now REDUNDANT data. REDO of course is hard, so may be no need. > yes, it's possible. > But if, for example, I have to print thousands of labels of addresses (to make > a mailing)? > The join is necessary here because every address is linked to a zip-city and > because the zip-city must be printed on every address? No. I give example > Unless we have redundant tables ?! No need!!!! Where you see need in redunant tables?! Its SQL will have them for JOIN. To print labels you write (syntax just to show idea) PrintLabels( inCities as Vset ) { for_each cityRecID from inCites { PrintLabelsForCity( cityRecId ) } } PrintLabelsForCity( cityRecID as Integer ) { tblCity.RecID = cityRecID // goto City record cityName = tblCity.Name cityZip = tblCity.Zip streetsSet = LinkStreets.FindLinked( cityRecID ) for_each streetRecID from streetsSet { tblStreet.RecId = streetRecID // go to street PrintLabel( cityName, cityZip, tblStreet.Name ) } } PrintLabel( cityName, cityZip, streetName ) { // PRINT code here. } What you think now? :-) Is it so hard ? And this way in general erquire less RAM/DISK space, Because we simply use original tables. No any redundant data/tables. SQL way have them -- JOIN TABLE. -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From vidal_olivier at yahoo.fr Tue Jan 25 12:21:11 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Tue Jan 25 05:21:24 2005 Subject: API methods on binary link // Philosophy (!!) In-Reply-To: References: Message-ID: <37E1871C-6EC3-11D9-96AD-000D93C35C66@yahoo.fr> ok, Ruslan, I am convinced ! I go use API in my app. thank you olivier Le 25 janv. 05, ? 11:39, Ruslan Zasukhin a ?crit : > On 1/25/05 12:05 PM, "Olivier" wrote: > >> Thank you very much Ruslan for all these explanations. >> >> I have to think really about my application. >> It is VERY attractive.. >> But I have to think seriously about all this. >> I shall not like, with THIS model, having to redo all structures of >> my tables >> or to have now REDUNDANT data. > > REDO of course is hard, so may be no need. > >> yes, it's possible. >> But if, for example, I have to print thousands of labels of addresses >> (to make >> a mailing)? > >> The join is necessary here because every address is linked to a >> zip-city and >> because the zip-city must be printed on every address? > > No. I give example > >> Unless we have redundant tables ?! > > No need!!!! > > Where you see need in redunant tables?! > Its SQL will have them for JOIN. > > To print labels you write (syntax just to show idea) > > PrintLabels( inCities as Vset ) > { > for_each cityRecID from inCites > { > PrintLabelsForCity( cityRecId ) > } > } > > > PrintLabelsForCity( cityRecID as Integer ) > { > tblCity.RecID = cityRecID // goto City record > > cityName = tblCity.Name > cityZip = tblCity.Zip > > streetsSet = LinkStreets.FindLinked( cityRecID ) > > for_each streetRecID from streetsSet > { > tblStreet.RecId = streetRecID // go to street > PrintLabel( cityName, cityZip, tblStreet.Name ) > } > } > > > PrintLabel( cityName, cityZip, streetName ) > { > // PRINT code here. > } > > > What you think now? :-) > Is it so hard ? > > And this way in general erquire less RAM/DISK space, > Because we simply use original tables. > No any redundant data/tables. > SQL way have them -- JOIN TABLE. > > > -- > Best regards, > Ruslan Zasukhin [ I feel the need...the need for speed ] > ------------------------------------------------------------- > e-mail: ruslan@paradigmasoft.com > web: http://www.paradigmasoft.com > > To subscribe to the Valentina mail list go to: > http://lists.macserve.net/mailman/listinfo/valentina > ------------------------------------------------------------- > > > _______________________________________________ > Valentina-beta mailing list > Valentina-beta@lists.macserve.net > http://lists.macserve.net/mailman/listinfo/valentina-beta > From jda at his.com Tue Jan 25 06:43:47 2005 From: jda at his.com (jda) Date: Tue Jan 25 05:43:57 2005 Subject: V4RB crashes on launch In-Reply-To: References: Message-ID: >On 1/25/05 2:33 AM, "jda" wrote: > >> I trashed my preferences and all old databases. Here's the crash report: >> >> Thread 0 Crashed: >> 0 vshared_carbon_cw.shlb 0x000f3bb4 0xf2000 + 0x1bb4 >> 1 Kernel_Carbon_Final.shlb 0x0205472c >> get_Path__Q23fbl20Location_Disk_FSSpecCFv + 0x74 >> 2 Kernel_Carbon_Final.shlb 0x023d8090 >> Valentina_Init_ICU__3fblFv + 0x24 >> 3 Kernel_Carbon_Final.shlb 0x023d7bd4 >> ValentinaInit__3fblFUlPcPc + 0x3c >> 4 Kernel_Carbon_Final.shlb 0x023d7b58 >> ValentinaInitL__3fblFUlPcPc + 0x10 >> 5 Plugin 0x00528bf0 >> V4RB_RegisterErrorCodes__Fv + 0x2c >> 6 Plugin 0x0050f638 PluginEntry + 0x24c >> >> >> Jon > >As far as I see, Frank and Oliver was able to run b38... > Hi Ruslan, You think I wrote the crash report? Please run my project -- you have the latest build. Jon From sunshine at public.kherson.ua Tue Jan 25 14:22:23 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Tue Jan 25 06:22:30 2005 Subject: API methods on binary link // Philosophy (!!) In-Reply-To: <37E1871C-6EC3-11D9-96AD-000D93C35C66@yahoo.fr> Message-ID: On 1/25/05 1:21 PM, "Olivier" wrote: > ok, Ruslan, I am convinced ! Be careful :-) > I go use API in my app. Only I agree that SetIterators are not comfortable somehow... May be let's invent more comfortable syntax... But I doubt this is possible. -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From jda at his.com Tue Jan 25 07:22:54 2005 From: jda at his.com (jda) Date: Tue Jan 25 06:23:04 2005 Subject: V4RB crashes on launch In-Reply-To: References: Message-ID: > > I trashed my preferences and all old databases. Here's the crash report: >> >> Thread 0 Crashed: >> 0 vshared_carbon_cw.shlb 0x000f3bb4 0xf2000 + 0x1bb4 >> 1 Kernel_Carbon_Final.shlb 0x0205472c >> get_Path__Q23fbl20Location_Disk_FSSpecCFv + 0x74 >> 2 Kernel_Carbon_Final.shlb 0x023d8090 >> Valentina_Init_ICU__3fblFv + 0x24 >> 3 Kernel_Carbon_Final.shlb 0x023d7bd4 >> ValentinaInit__3fblFUlPcPc + 0x3c >> 4 Kernel_Carbon_Final.shlb 0x023d7b58 >> ValentinaInitL__3fblFUlPcPc + 0x10 >> 5 Plugin 0x00528bf0 >> V4RB_RegisterErrorCodes__Fv + 0x2c >> 6 Plugin 0x0050f638 PluginEntry + 0x24c >> >> >> Jon > >As far as I see, Frank and Oliver was able to run b38... > Ruslan, I see what the bug is, now. I have two copies of the Valentinat VComponents files, one in CFMSupport and one in the folder with my app. You told me to do this so I can run my project in RB and also my standalone (for distribution the files will be in the same folder as the app). This worked for b37. In b38 this -> crash. I removed all the support VComponent *files* from the app folder -> still crashed. I removed the Resources folder (now *everything* was out of the app folder) -> runs. You should be able to reproduce this with my project. Jon From sunshine at public.kherson.ua Tue Jan 25 14:24:17 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Tue Jan 25 06:24:21 2005 Subject: V4RB crashes on launch In-Reply-To: Message-ID: On 1/25/05 1:43 PM, "jda" wrote: >> On 1/25/05 2:33 AM, "jda" wrote: >> >>> I trashed my preferences and all old databases. Here's the crash report: >>> >>> Thread 0 Crashed: >>> 0 vshared_carbon_cw.shlb 0x000f3bb4 0xf2000 + 0x1bb4 >>> 1 Kernel_Carbon_Final.shlb 0x0205472c >>> get_Path__Q23fbl20Location_Disk_FSSpecCFv + 0x74 >>> 2 Kernel_Carbon_Final.shlb 0x023d8090 >>> Valentina_Init_ICU__3fblFv + 0x24 >>> 3 Kernel_Carbon_Final.shlb 0x023d7bd4 >>> ValentinaInit__3fblFUlPcPc + 0x3c >>> 4 Kernel_Carbon_Final.shlb 0x023d7b58 >>> ValentinaInitL__3fblFUlPcPc + 0x10 >>> 5 Plugin 0x00528bf0 >>> V4RB_RegisterErrorCodes__Fv + 0x2c >>> 6 Plugin 0x0050f638 PluginEntry + 0x24c >>> >>> >>> Jon >> >> As far as I see, Frank and Oliver was able to run b38... >> > > Hi Ruslan, > > You think I wrote the crash report? > > Please run my project -- you have the latest build. Stop. So crash only with your project ? Are you able to run examples ? 2) I need your fresh project. If you can run examples, but not your project then I need test this. -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From sunshine at public.kherson.ua Tue Jan 25 14:24:42 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Tue Jan 25 06:24:50 2005 Subject: V4RB crashes on launch In-Reply-To: Message-ID: On 1/25/05 1:43 PM, "jda" wrote: >> On 1/25/05 2:33 AM, "jda" wrote: >> >>> I trashed my preferences and all old databases. Here's the crash report: >>> >>> Thread 0 Crashed: >>> 0 vshared_carbon_cw.shlb 0x000f3bb4 0xf2000 + 0x1bb4 >>> 1 Kernel_Carbon_Final.shlb 0x0205472c >>> get_Path__Q23fbl20Location_Disk_FSSpecCFv + 0x74 >>> 2 Kernel_Carbon_Final.shlb 0x023d8090 >>> Valentina_Init_ICU__3fblFv + 0x24 >>> 3 Kernel_Carbon_Final.shlb 0x023d7bd4 >>> ValentinaInit__3fblFUlPcPc + 0x3c >>> 4 Kernel_Carbon_Final.shlb 0x023d7b58 >>> ValentinaInitL__3fblFUlPcPc + 0x10 >>> 5 Plugin 0x00528bf0 >>> V4RB_RegisterErrorCodes__Fv + 0x2c >>> 6 Plugin 0x0050f638 PluginEntry + 0x24c >>> >>> >>> Jon >> >> As far as I see, Frank and Oliver was able to run b38... >> > > Hi Ruslan, > > You think I wrote the crash report? > > Please run my project -- you have the latest build. Do you put DLLs into ROOT path ? -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From sunshine at public.kherson.ua Tue Jan 25 14:35:15 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Tue Jan 25 06:35:19 2005 Subject: V4RB crashes on launch In-Reply-To: Message-ID: On 1/25/05 2:22 PM, "jda" wrote: > Ruslan, > > I see what the bug is, now. > > I have two copies of the Valentinat VComponents files, one in > CFMSupport and one in the folder with my app. You told me to do this > so I can run my project in RB and also my standalone (for > distribution the files will be in the same folder as the app). > > This worked for b37. > > In b38 this -> crash. > > I removed all the support VComponent *files* from the app folder -> > still crashed. > > I removed the Resources folder (now *everything* was out of the app > folder) -> runs. > > You should be able to reproduce this with my project. * I know. It seems I have not finish APP-folder way for b39. * "resources" folder is indicator where DLLs are. this is important. * so yes for now you need simply move all to ROOT. And Jown, while you work in DEBUG and in REALbasic Then no need move Valentina DLLs into app folder! You need do this steps I think: * when you compile RELESE APP, you get app file. * now I recommend to have some script that will build YOUR archive that you upload to people. And this script will copy your app and Vcomponents dlls into that new folder. -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From sunshine at public.kherson.ua Tue Jan 25 14:37:04 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Tue Jan 25 06:37:07 2005 Subject: Jda, check b39 Message-ID: Hi Jon, So please check b39 against your bugs * big index fails. it may work now. * any else bugs you have report. I think only SORT of NULLs exactly now work yet. Let me know please which bugs still are actual. Okay ? -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From jda at his.com Tue Jan 25 07:47:37 2005 From: jda at his.com (jda) Date: Tue Jan 25 06:47:47 2005 Subject: Jda, check b39 In-Reply-To: References: Message-ID: >Hi Jon, > >So please check b39 against your bugs You mean b38, right? >* big index fails. > it may work now. No, it fails. I just tested it. > >* any else bugs you have report. > I think only SORT of NULLs exactly now work yet. > > >Let me know please which bugs still are actual. >Okay ? > Other issues not resolved (I just tested these with b38, the latest download): #244 (unicode/long file names) #149 (blob fields lose contents after Vexception) #99 (convert 1.x to 2) -- I just tried and it crashed. I don't have time to investigate more right now. I'll get back to it later. Jon From vidal_olivier at yahoo.fr Tue Jan 25 14:24:36 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Tue Jan 25 07:24:54 2005 Subject: API methods + b38 faster ? In-Reply-To: References: Message-ID: <754246DE-6ED4-11D9-96AD-000D93C35C66@yahoo.fr> > >> I go use API in my app. > > Only I agree that SetIterators are not comfortable somehow... > Yes, it push to use systematically arraysets (for the random access)... My application seems to open much more quickly tables from their first opening. In b38, indexs are cached ? olivier From sunshine at public.kherson.ua Tue Jan 25 15:27:48 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Tue Jan 25 07:28:00 2005 Subject: API methods + b38 faster ? In-Reply-To: <754246DE-6ED4-11D9-96AD-000D93C35C66@yahoo.fr> Message-ID: On 1/25/05 3:24 PM, "Olivier" wrote: >> >>> I go use API in my app. >> >> Only I agree that SetIterators are not comfortable somehow... >> > > Yes, it push to use systematically arraysets (for the random access)... > > My application seems to open much more quickly tables from their first > opening. > In b38, indexs are cached ? Not yet Oliver. -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From sunshine at public.kherson.ua Tue Jan 25 15:35:31 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Tue Jan 25 07:35:36 2005 Subject: Jda, check b39 In-Reply-To: Message-ID: On 1/25/05 2:47 PM, "jda" wrote: >> * big index fails. >> it may work now. > > No, it fails. I just tested it. > >> >> * any else bugs you have report. >> I think only SORT of NULLs exactly now work yet. >> >> >> Let me know please which bugs still are actual. >> Okay ? >> > > Other issues not resolved (I just tested these with b38, the latest download): > > #244 (unicode/long file names) > #149 (blob fields lose contents after Vexception) > #99 (convert 1.x to 2) -- I just tried and it crashed. I don't have > time to investigate more right now. I'll get back to it later. So at least 4 bugs yet... Please send me your latest project. -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From vidal_olivier at yahoo.fr Wed Jan 26 05:37:27 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Tue Jan 25 22:37:42 2005 Subject: flush binary link Message-ID: Hi Ruslan and list, a small notice : When we make an operation on a record (update, delete), it is advised to flush only the table of the record (rather than the whole base) so that it is faster. But I noticed that with the binary links, if, for example, we add a new record in a table, and that it is necessary to link this record to another table (e.g. add a new street in a table streets, and link this new street to a city of the table city), it will be necessary then to flush ALL the base because the information of links is not in the table but in the base. Olivier From sunshine at public.kherson.ua Wed Jan 26 08:06:44 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Wed Jan 26 00:07:05 2005 Subject: flush binary link In-Reply-To: Message-ID: On 1/26/05 6:37 AM, "Olivier" wrote: > Hi Ruslan and list, > > a small notice : > > When we make an operation on a record (update, delete), it is advised > to flush only the table of the record (rather than the whole base) so > that it is faster. Agree Oliver! > But I noticed that with the binary links, if, for example, we add a new > record in a table, and that it is necessary to link this record to > another table (e.g. add a new street in a table streets, and link this > new street to a city of the table city), it will be necessary then to > flush ALL the base because the information of links is not in the table > but in the base. So you offer, Link.flush() -- flushes only link information. have sense for binary link only. Right ? -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From vidal_olivier at yahoo.fr Wed Jan 26 07:24:59 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Wed Jan 26 00:25:01 2005 Subject: flush binary link In-Reply-To: References: Message-ID: <00D18F7E-6F63-11D9-A289-000D93C35C66@yahoo.fr> Le 26 janv. 05, ? 07:06, Ruslan Zasukhin a ?crit : > Link.flush() -- flushes only link information. > have sense for binary link only. > > Right ? yes, I have already tried. Link.flush is not in auto-completion of RB. It is refused at the compilation. I think that it is not implemented in valentina for RB. olivier From vidal_olivier at yahoo.fr Wed Jan 26 07:31:51 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Wed Jan 26 00:31:51 2005 Subject: flush binary link In-Reply-To: References: Message-ID: > Sorry, I believe that I badly understood your answer. English is not my first language. Thus yes, it would be perfect that a function as " link.flush " exists! thank you olivier > So you offer, > > Link.flush() -- flushes only link information. > have sense for binary link only. > > Right ? > > > -- > Best regards, > Ruslan Zasukhin [ I feel the need...the need for speed ] > ------------------------------------------------------------- > e-mail: ruslan@paradigmasoft.com > web: http://www.paradigmasoft.com > > To subscribe to the Valentina mail list go to: > http://lists.macserve.net/mailman/listinfo/valentina > ------------------------------------------------------------- > > > _______________________________________________ > Valentina-beta mailing list > Valentina-beta@lists.macserve.net > http://lists.macserve.net/mailman/listinfo/valentina-beta > From AYU8 at CDC.GOV Wed Jan 26 08:35:54 2005 From: AYU8 at CDC.GOV (Sims, John) Date: Wed Jan 26 07:36:47 2005 Subject: flush binary link Message-ID: > Sorry, I believe that I badly understood your answer. English > is not my > first language. > Thus yes, it would be perfect that a function as " link.flush > " exists! > > thank you > > olivier > > > So you offer, > > > > Link.flush() -- flushes only link information. > > have sense for binary link only. > > > > Right ? > > > > > > -- > > Best regards, > > Ruslan Zasukhin [ I feel the need...the need for speed ] Ruslan, If you add the new feature as suggested above (which I agree is a good idea), would calling Link.Flush automatically flush the tables involved? If not, what would happen if a developer calls Link.Flush without flushing the tables. I know the cache would be flushed on close but what if there was a "catastrophic shutdown", ie. a power failure and the computer the program is running on is not plugged into a battery backup. So, now the binarylink thinks it is linking 2 records that do not actually exist in the tables, right? BTW, development using V2 is going very well so far (early stages of course) and I have to say that V2 is shaping up to be a fantastic product. Thanks Ruslan. -John From vidal_olivier at yahoo.fr Wed Jan 26 14:36:57 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Wed Jan 26 07:37:22 2005 Subject: delete a record binary linked to another record : crash Message-ID: <598975AA-6F9F-11D9-A289-000D93C35C66@yahoo.fr> Hi, I cannot delete records of a table linked to another table by a binary link. I have every time a crash ( no Exception) in the line: table.deleteRecord () Example: Table street: 4 records City table: 4 records Link_ZipCity_Street=new vbinaryLink ("link_ZipCity_street", zipCity, Street, EVlinkType.kOne, EVlinkType.kMany, evOnDeletion.kCascade) Link_ZipCity_Street.linkRecords (array (2,2)) Later: Street.recID=2 LinkCityStreet.unlinkRecords (array (2,2)) DicoTable.deleteRecord () even if I remove the line "unlikrecords ", Street.recID=2 DicoTable.deleteRecord () I has also a crash there. olivier From sunshine at public.kherson.ua Wed Jan 26 15:52:07 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Wed Jan 26 07:52:18 2005 Subject: flush binary link In-Reply-To: Message-ID: On 1/26/05 3:35 PM, "Sims, John" wrote: Hi John, > If you add the new feature as suggested above (which I agree is a good > idea), would calling Link.Flush automatically flush the tables involved? I think about this also. > If not, what would happen if a developer calls Link.Flush without > flushing the tables. Then records of tables are still in cache, But info of BinaryLink (i.e. Pair of RecIDs) will be flushed. > I know the cache would be flushed on close but > what if there was a "catastrophic shutdown", ie. a power failure and the > computer the program is running on is not plugged into a battery backup. Then you loose any not flushed info. > So, now the binarylink thinks it is linking 2 records that do not > actually exist in the tables, right? If that was 2 absolutely new records, then yes. So we have come to idea, that Link MUST flush that records of tables also. > BTW, development using V2 is going very well so far (early stages of > course) and I have to say that V2 is shaping up to be a fantastic > product. Thank you! -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From sunshine at public.kherson.ua Wed Jan 26 15:52:36 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Wed Jan 26 07:52:46 2005 Subject: delete a record binary linked to another record : crash In-Reply-To: <598975AA-6F9F-11D9-A289-000D93C35C66@yahoo.fr> Message-ID: On 1/26/05 3:36 PM, "Olivier" wrote: > Hi, > > I cannot delete records of a table linked to another table by a binary > link. > I have every time a crash ( no Exception) in the line: > table.deleteRecord () > > Example: > > Table street: 4 records > City table: 4 records > Link_ZipCity_Street=new vbinaryLink ("link_ZipCity_street", zipCity, > Street, EVlinkType.kOne, EVlinkType.kMany, evOnDeletion.kCascade) > > Link_ZipCity_Street.linkRecords (array (2,2)) > > Later: > > Street.recID=2 > LinkCityStreet.unlinkRecords (array (2,2)) > DicoTable.deleteRecord () > > even if I remove the line "unlikrecords ", > > Street.recID=2 > DicoTable.deleteRecord () > > I has also a crash there. Can you reproduce for us in project? -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From vidal_olivier at yahoo.fr Wed Jan 26 16:59:27 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Wed Jan 26 09:59:32 2005 Subject: delete a record binary linked to another record : crash In-Reply-To: References: Message-ID: <41C047DA-6FB3-11D9-A289-000D93C35C66@yahoo.fr> That was difficult to reproduce the problem. In fact, the problem is not due to the links. Simply, we cannot delete a record of a table if this table contains a method' field. constructor of table street : name="Street" namestreet=new vvarchar("nameStreet",100,EVFlag.fIndexed+evFlag.fIndexByWords) name2=new vvarchar("name2",100,EVFlag.fIndexed,"namestreet") -------> CRASH if record of table street deleted. constructor of table street : name="Street" namestreet=new vvarchar("nameStreet",100,EVFlag.fIndexed+evFlag.fIndexByWords) -------> OK if record deleted ! little project sent to Ruslan. olivier Le 26 janv. 05, ? 14:52, Ruslan Zasukhin a ?crit : > On 1/26/05 3:36 PM, "Olivier" wrote: > >> Hi, >> >> I cannot delete records of a table linked to another table by a binary >> link. >> I have every time a crash ( no Exception) in the line: >> table.deleteRecord () >> >> Example: >> >> Table street: 4 records >> City table: 4 records >> Link_ZipCity_Street=new vbinaryLink ("link_ZipCity_street", zipCity, >> Street, EVlinkType.kOne, EVlinkType.kMany, evOnDeletion.kCascade) >> >> Link_ZipCity_Street.linkRecords (array (2,2)) >> >> Later: >> >> Street.recID=2 >> LinkCityStreet.unlinkRecords (array (2,2)) >> DicoTable.deleteRecord () >> >> even if I remove the line "unlikrecords ", >> >> Street.recID=2 >> DicoTable.deleteRecord () >> >> I has also a crash there. > > Can you reproduce for us in project? > > -- > Best regards, > Ruslan Zasukhin [ I feel the need...the need for speed ] > ------------------------------------------------------------- > e-mail: ruslan@paradigmasoft.com > web: http://www.paradigmasoft.com > > To subscribe to the Valentina mail list go to: > http://lists.macserve.net/mailman/listinfo/valentina > ------------------------------------------------------------- > > > _______________________________________________ > Valentina-beta mailing list > Valentina-beta@lists.macserve.net > http://lists.macserve.net/mailman/listinfo/valentina-beta > From vidal_olivier at yahoo.fr Wed Jan 26 17:05:17 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Wed Jan 26 10:05:16 2005 Subject: flush binary link In-Reply-To: References: Message-ID: <124D33D6-6FB4-11D9-A289-000D93C35C66@yahoo.fr> Le 26 janv. 05, ? 14:52, Ruslan Zasukhin a ?crit : > So we have come to idea, that Link MUST flush that records of tables > also. yes, good idea ! olivier From sunshine at public.kherson.ua Wed Jan 26 19:28:20 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Wed Jan 26 11:28:26 2005 Subject: flush binary link In-Reply-To: <124D33D6-6FB4-11D9-A289-000D93C35C66@yahoo.fr> Message-ID: On 1/26/05 6:05 PM, "Olivier" wrote: > > Le 26 janv. 05, ? 14:52, Ruslan Zasukhin a ?crit : > >> So we have come to idea, that Link MUST flush that records of tables >> also. > > yes, good idea ! We will make it as: Vlink.Flush( inFlushTables as Boolean = true ) -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From sunshine at public.kherson.ua Thu Jan 27 14:12:54 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Thu Jan 27 06:13:21 2005 Subject: Frank, I have fix bug in indexes! Message-ID: So now, Oliver with report on wrong BinaryLink results John with report on big index bad result Please check this reports with next b39. May happens that it will work. -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From jordan at digi-net.com Thu Jan 27 10:03:49 2005 From: jordan at digi-net.com (Ken Jordan) Date: Thu Jan 27 09:00:44 2005 Subject: [V4RB] 2.0 b38 uploaded In-Reply-To: References: Message-ID: <41F902D5.3050405@digi-net.com> How is the work on the Win version progressing? Ken Ruslan Zasukhin wrote: > > kernel - 2.0 b38 > ================ > - 0000333: *[API]* Valentina.GetVersion() for 2.0 should return 0x200. > (Ruslan Zasukhin) > - 0000322: *[SQL Parser]* SQL syntax for Diagnose comand. (Sergey > Duminskiy) > - 0000332: *[SQL nodes]* Crash if join-query where clause does not > contain join specification. (Ivan Smahin) > - 0000329: *[Fields]* ObjectPtr register self as Link in tables only - > not in db. (Ivan Smahin) > - 0000323: *[Methods (calculated fields)]* There is no unique check > for unique methods during insert/update. (Ivan Smahin) > - 0000321: *[Tables]* We get a crash trying to drop a TEMPORARY table. > (Ivan Smahin) > - 0000319: *[Fields]* OID now returns values as expected. (Ruslan > Zasukhin) > - 0000317: *[Database]* Wrong segment size after Convert_1_2. (Ivan > Smahin) > > > > > V4RB - 2.0 b38 > ============== > - 0000330: *[SQL]* crash on SQL query without Link conditions (Ivan > Smahin) > - 0000326: *[SQL]* droprecord -> locked records (Ivan Smahin) > - 0000281: *[API]* UpdateRecord Crash on large table (Igor Gomon) > - 0000328: *[Fields]* VField.FindLink() now returns VBitSet (Ruslan > Zasukhin) > - 0000324: *[Fields]* Flag "UNIQUE" doesn't work (Ruslan Zasukhin) > >------------------------------------------------------------------------ > >_______________________________________________ >Valentina-beta mailing list >Valentina-beta@lists.macserve.net >http://lists.macserve.net/mailman/listinfo/valentina-beta > > -------------- next part -------------- An HTML attachment was scrubbed... URL: http://lists.macserve.net/pipermail/valentina-beta/attachments/20050127/659919aa/attachment.html From sunshine at public.kherson.ua Thu Jan 27 23:50:45 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Thu Jan 27 15:50:54 2005 Subject: [V4RB] 2.0. b39 uploaded Message-ID: kernel - 2.0 b39 ================ - 0000337: [Database] It's hard to find Table, Link, Field which is absent in the scema but present in "classes way" definitions. (Ivan Smahin) - 0000336: [Database] Random field order creation in "classes way" getting a crash. (Ivan Smahin) - 0000338: [API] xLinkError new error code is added - ERR_LINK_WRONG_TYPE. (Ivan Smahin) - 0000341: [Fields] There is zero values in ObjectPtr field after db.Open(). (Ivan Smahin) - 0000339: [Links] Add I_Link::Flush() and its implementations. (Ivan Smahin) V4RB - 2.0 b39 ============== - 0000343: [Links] [NEW] VLink.Flush( inFlushTables as Boolean = true ) (Ruslan Zasukhin) - 0000149: [API] Blob fields seem to lose content if there is an exception with addrecord attempt (Ivan Smahin) -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- -------------- next part -------------- An HTML attachment was scrubbed... URL: http://lists.macserve.net/pipermail/valentina-beta/attachments/20050127/0a1d8094/attachment.html From vidal_olivier at yahoo.fr Fri Jan 28 06:02:17 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Thu Jan 27 23:02:24 2005 Subject: [V4RB] 2.0. b39 uploaded In-Reply-To: References: Message-ID: Le 27 janv. 05, ? 22:50, Ruslan Zasukhin a ?crit : > > ============== > - 0000343: [Links] [NEW] VLink.Flush( inFlushTables as Boolean = true > ) (Ruslan Zasukhin) > good ! -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: text/enriched Size: 521 bytes Desc: not available Url : http://lists.macserve.net/pipermail/valentina-beta/attachments/20050128/bd1331c8/attachment.bin From vidal_olivier at yahoo.fr Fri Jan 28 06:32:18 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Thu Jan 27 23:32:22 2005 Subject: Frank, I have fix bug in indexes! In-Reply-To: References: Message-ID: Hi Ruslan and list, > Oliver with report on wrong BinaryLink results > There is always a problem. First, I had difficulties to reproduce the problem with b38. But I isolated the problem and it reproduces with b39. I send you (to ruslan) thus a small project where always reproduces the problem. We thus have two tables : ZipCity table Street table Both tables are linked by a binary link: LinkCityStreet=new vbinaryLink (" linkCityStreet ", zipCity, street, EVLinkType.kOne, EvLinkType.kmany) - > launch the project - > click on " create DB ": WAIT for some minutes which the DB builds up itself: 40 002 citys create 100 000 streets create For i=1 to 30000 Base.linkCityStreet.linkRecords (array (1, i)) // 30 000 streets linked with the city "1" which along Zipcode = '75000' Next i For i=30001 to 30002 Base.linkCityStreet.linkRecords (array (2, i)) // 2 streets linked with the city "2" which along Zipcode ALSO = '75000' Next i - > click " SQL Start ": Cur=new vcursor (base, " SELECT * FROM zipcity join street we linkcitystreet where zipcode = '75000' ", evCursorLocation.kServerSide, evLockType.kReadOnly, evCursorDirection.kRandom) - > result: 30 002 found records, OK ! TRUE - > quit the project - > restart the project - > click on " OPEN DB " (NOT "create DB", click "Open DB" button) - > click " SQL Start " - > result: only 2 records found! FALSE - > click again one " SQL Start " - > result: 30 002 founds records! TRUE Thus, in the first opening of the DB, on the first SQL query, the result is false. NOW, quit app, and in "PushButton2.action" (" create DB " button), you change the following lines: For i=1 to 30000 Base.linkCityStreet.linkRecords (array (1, i)) Next i For i=30001 to 30002 Base.linkCityStreet.linkRecords (array (2, i)) Next i BY For i=1 to 30000 Base.linkCityStreet.linkRecords (array (2, i)) // 30 000 records are linked NOW to city 2 Next i For i=30001 to 30002 Base.linkCityStreet.linkRecords (array (1, i)) // this 2 records are linked NOW to city 1 Next i There, there are not problems anymore! Olivier From jda at his.com Fri Jan 28 12:18:51 2005 From: jda at his.com (jda) Date: Fri Jan 28 11:19:03 2005 Subject: Converted chars > 127 not correct In-Reply-To: References: Message-ID: Hi Ruslan, I'm converting V1.x dbs to V2. In one case I have the delta character ? (Option-J) When it is converted (V1.x db was set to "English") to V2 (where the field is UTF-16 for now, eventually UTF-8) the character is converted to 3 character garbage (I assume it's UTF-8 represented as ASCII): ??? Jon From sunshine at public.kherson.ua Fri Jan 28 20:35:16 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Fri Jan 28 12:35:23 2005 Subject: Converted chars > 127 not correct In-Reply-To: Message-ID: On 1/28/05 7:18 PM, "jda" wrote: > Hi Ruslan, > > I'm converting V1.x dbs to V2. In one case I have the delta character > > ? (Option-J) > > When it is converted (V1.x db was set to > "English") to V2 (where the field is UTF-16 for > now, eventually UTF-8) the character is converted > to 3 character garbage (I assume it's UTF-8 > represented as ASCII): > > ??? Add this to mantis please -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From sunshine at public.kherson.ua Fri Jan 28 23:03:20 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Fri Jan 28 15:03:27 2005 Subject: Large index bug In-Reply-To: Message-ID: On 1/28/05 1:05 PM, "jda" wrote: > Hi Ruslan, > > New info on the large index bug. > > > > I do the search for 'lobster' in the "Keywords" field (left()) -> 54 hits > I repeat the search -> 0 hits > > I do a search for a different word ('physiology') -> 72 hits > I repeat the search -> 72 hits > Again -> 72 hits > > I do the search for 'lobster' in the "Keywords" field (left()) -> 54 hits > I repeat the search -> 0 hits > > > > > Maybe Valentina is dropping some text from the index (?) after the > first search. Other text is not affected. This is just a guess. Check next build. I have fix similar bug for Oliver -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From sunshine at public.kherson.ua Fri Jan 28 23:31:59 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Fri Jan 28 15:32:05 2005 Subject: [V4RB] 2.0b40 Message-ID: b40 0000340: bad results of SQL query on binary link --------------- Actually fixed bug in index which could produce result in first time different from second. -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From jda at his.com Fri Jan 28 16:44:05 2005 From: jda at his.com (jda) Date: Fri Jan 28 15:44:21 2005 Subject: [V4RB] 2.0b40 In-Reply-To: References: Message-ID: >b40 > > 0000340: bad results of SQL query on binary link > > >--------------- >Actually fixed bug in index which could produce result in first time >different from second. > > Sorry, my index bug is still there (behaves a bit differently from before, but still there). Jon From sunshine at public.kherson.ua Fri Jan 28 23:53:11 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Fri Jan 28 15:53:17 2005 Subject: [V4RB] 2.0b40 In-Reply-To: Message-ID: On 1/28/05 11:44 PM, "jda" wrote: >> b40 >> >> 0000340: bad results of SQL query on binary link >> >> >> --------------- >> Actually fixed bug in index which could produce result in first time >> different from second. >> >> > > Sorry, my index bug is still there (behaves a bit differently from > before, but still there). Bad. Oliver, please confirm that BinaryLink works for you know. -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From sunshine at public.kherson.ua Sat Jan 29 01:06:12 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Fri Jan 28 17:06:19 2005 Subject: [V4RB] 2.0b40 In-Reply-To: Message-ID: On 1/28/05 11:31 PM, "Ruslan Zasukhin" wrote: > > b40 > > 0000340: bad results of SQL query on binary link * also fixed problem in the Varchar field, so now diagnose will not complain on all tables that have a varchar field. -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From vidal_olivier at yahoo.fr Sat Jan 29 07:00:36 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Sat Jan 29 00:00:44 2005 Subject: [V4RB] 2.0b40 In-Reply-To: References: Message-ID: <189282A2-71BB-11D9-A2B4-000D93C35C66@yahoo.fr> Le 28 janv. 05, ? 22:53, Ruslan Zasukhin a ?crit : > On 1/28/05 11:44 PM, "jda" wrote: > >>> b40 >>> >>> 0000340: bad results of SQL query on binary link >>> >>> >>> --------------- >>> Actually fixed bug in index which could produce result in first time >>> different from second. >>> >>> >> >> Sorry, my index bug is still there (behaves a bit differently from >> before, but still there). > > Bad. > > Oliver, please confirm that BinaryLink works for you know. > > Hi Ruslan, yes, for me, that's works ! thank you olivier From vidal_olivier at yahoo.fr Sat Jan 29 08:06:20 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Sat Jan 29 01:06:28 2005 Subject: [V4RB] 2.0b40 In-Reply-To: References: Message-ID: <472FD03C-71C4-11D9-A2B4-000D93C35C66@yahoo.fr> Le 28 janv. 05, ? 22:53, Ruslan Zasukhin a ?crit : > On 1/28/05 11:44 PM, "jda" wrote: > >>> b40 >>> >>> 0000340: bad results of SQL query on binary link >>> >>> >>> --------------- >>> Actually fixed bug in index which could produce result in first time >>> different from second. >>> >>> >> >> Sorry, my index bug is still there (behaves a bit differently from >> before, but still there). > > Bad. > > Oliver, please confirm that BinaryLink works for you know. > > Ruslan, I have a problem with b40. If I open my DB with vStudio a77, everything works. If I launch a SQL query, everything works. If I open my DB with ViSQL b40, same thing. SQL query: same thing. ( With ViSQL b39: crash) But if I launch my app: - Opening of the DB: OK - Opening of tables without links: OK (the sets and the sorts work) - Opening of tables WITH links: OK BUT crash as soon as I try to sort out it : setall=ZipCityTable.selectAllRecords set=ZipCityTable.sort(setall,ZipCityTable.field("zip")) However NOW, I recreate a NEW db (with my app and b40), my app seems to work (with the sorts). Olivier From sunshine at public.kherson.ua Sat Jan 29 09:16:55 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Sat Jan 29 01:17:05 2005 Subject: [V4RB] 2.0b40 In-Reply-To: <472FD03C-71C4-11D9-A2B4-000D93C35C66@yahoo.fr> Message-ID: On 1/29/05 9:06 AM, "Olivier" wrote: > Ruslan, > > I have a problem with b40. > > If I open my DB with vStudio a77, everything works. > If I launch a SQL query, everything works. > > If I open my DB with ViSQL b40, same thing. > SQL query: same thing. > ( With ViSQL b39: crash) > > But if I launch my app: > > - Opening of the DB: OK > - Opening of tables without links: OK (the sets and the sorts work) > - Opening of tables WITH links: OK BUT crash as soon as I try to sort > out it : > > setall=ZipCityTable.selectAllRecords > set=ZipCityTable.sort(setall,ZipCityTable.field("zip")) > > However NOW, I recreate a NEW db (with my app and b40), my app seems to > work (with the sorts). May be something was changed. So if new db works that is ok. -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From vidal_olivier at yahoo.fr Mon Jan 31 15:22:08 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Mon Jan 31 08:22:21 2005 Subject: API methods + binary link + M:M tables Message-ID: <7D55C61D-7393-11D9-A2B4-000D93C35C66@yahoo.fr> Hi Ruslan, I try to think only in navigationnal way to use the API methods. But in the following problem, I do not see how I can have an effective API code: I shall like knowing how much there is of persons having the same surname in every city. (Statistics). For example: PARIS: Name Counts name1 500 name2 300 name3 100 (thus, there 100 persons at Paris having the surname "name3") BORDEAUX: name1 300 name2 150 name3 80 .............. I think thus of having: - A city table (there are 36 000 cities) - A table of surnames (several thousand surnames) - A table (tbQuantity) making the link between the previous two tables AND having a field: "counts" ("counts"=Number of persons having this surname in a city) - A binary link 1:M between the table "quantity" and the table "name" - A binary link 1:M between the table "quantity" and the table "city" I think that it is necessary to make the third table (quantity) because of this field "counts". It is not maybe a good structure to use the API methods, but in that case, can we better make? Apparently, the best solution here are join them and SQL? thank you olivier From sunshine at public.kherson.ua Mon Jan 31 17:47:19 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Mon Jan 31 09:47:32 2005 Subject: API methods + binary link + M:M tables In-Reply-To: <7D55C61D-7393-11D9-A2B4-000D93C35C66@yahoo.fr> Message-ID: On 1/31/05 4:22 PM, "Olivier" wrote: > Hi Ruslan, > > I try to think only in navigationnal way to use the API methods. > But in the following problem, I do not see how I can have an effective > API code: > > > I shall like knowing how much there is of persons having the same > surname in every city. (Statistics). > > For example: > > PARIS: > > Name Counts > > name1 500 > name2 300 > name3 100 > > (thus, there 100 persons at Paris having the surname "name3") > > BORDEAUX: > > name1 300 > name2 150 > name3 80 I think FOR THIS task it is better to use GROUP BY of SQL. But I will think more on this In C++ we have low level methods which can solve this... But it is not easy open them for V4RB... It needs to think. -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From sunshine at public.kherson.ua Mon Jan 31 17:59:53 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Mon Jan 31 10:00:02 2005 Subject: API methods + binary link + M:M tables In-Reply-To: <7D55C61D-7393-11D9-A2B4-000D93C35C66@yahoo.fr> Message-ID: On 1/31/05 4:22 PM, "Olivier" wrote: > Hi Ruslan, > > I try to think only in navigationnal way to use the API methods. > But in the following problem, I do not see how I can have an effective > API code: > > > I shall like knowing how much there is of persons having the same > surname in every city. (Statistics). > > For example: > > PARIS: > > Name Counts > > name1 500 > name2 300 > name3 100 > > (thus, there 100 persons at Paris having the surname "name3") > > BORDEAUX: > > name1 300 > name2 150 > name3 80 > > .............. > > I think thus of having: > > - A city table (there are 36 000 cities) > - A table of surnames (several thousand surnames) > - A table (tbQuantity) making the link between the previous two tables > AND having a field: "counts" ("counts"=Number of persons having this > surname in a city) I do not think you need COUNTS table. I see only reason to have it, if you prepare some DB for CD and you want optimize of for speed as much as possible. If you develop regular db that will change runtime then I think this will make things more complex then needed. > - A binary link 1:M between the table "quantity" and the table "name" > - A binary link 1:M between the table "quantity" and the table "city" No no no. All should be natural and mirror the real world entities. > I think that it is necessary to make the third table (quantity) because > of this field "counts". > > It is not maybe a good structure to use the API methods, but in that > case, can we better make? > > Apparently, the best solution here are join them and SQL? Aha! Look into ValentinaSQL.pdf Here exists new group of "Link Methods". It seems to me you need the function COUNT( link_name, fld_B ) But I am not sure... ----------------- Aha! Guys! I think we need yet extend syntax of GROUP BY in the SQL to support Binary Link. Because now there is no way to say GROUP records of PERSON that are linked to some CITY With ObjectPtr-link field we can do this grouping by ObjectPtr itself. With binaryLink there is no such ability. So we must have syntax as Select From GROUP BY link_name Sergey, Ivan, you see idea? -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From vidal_olivier at yahoo.fr Mon Jan 31 17:00:17 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Mon Jan 31 10:01:05 2005 Subject: API methods + binary link + M:M tables In-Reply-To: References: Message-ID: <3335D74C-73A1-11D9-A2B4-000D93C35C66@yahoo.fr> Le 31 janv. 05, ? 16:47, Ruslan Zasukhin a ?crit : > I think FOR THIS task it is better to use GROUP BY of SQL. > > But I will think more on this > In C++ we have low level methods which can solve this... Yes, if it could have a solution there, that would be great! > But it is not easy open them for V4RB... > It needs to think. I hope that vou will find something! thank you, olivier From sunshine at public.kherson.ua Mon Jan 31 18:13:12 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Mon Jan 31 10:13:21 2005 Subject: API methods + binary link + M:M tables In-Reply-To: <3335D74C-73A1-11D9-A2B4-000D93C35C66@yahoo.fr> Message-ID: On 1/31/05 6:00 PM, "Olivier" wrote: > > Le 31 janv. 05, ? 16:47, Ruslan Zasukhin a ?crit : > >> I think FOR THIS task it is better to use GROUP BY of SQL. >> >> But I will think more on this >> In C++ we have low level methods which can solve this... > > Yes, if it could have a solution there, that would be great! > >> But it is not easy open them for V4RB... >> It needs to think. > > I hope that vou will find something! In any case SQL is not the bad way for such complex tasks that require a lots of calculation. BAD IS that as far as I see, you will not be able with Binary Link Build GROUP BY query. So we need resolve this ASAP. And in any case study that new Link Methods. They work in both API and SQL way. They will be very useful. -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From vbt at paradigma.ukrcom.kherson.ua Mon Jan 31 18:06:22 2005 From: vbt at paradigma.ukrcom.kherson.ua (vbt@paradigma.ukrcom.kherson.ua) Date: Mon Jan 31 10:18:06 2005 Subject: [kernel 0000289]: Make sure that works SQL commands DIAGNOSE, COMPACT, DEFRAGMENT. Message-ID: The following issue has been RESOLVED. ====================================================================== ====================================================================== Reported By: Ruslan Zasukhin Assigned To: Ivan Smahin ====================================================================== Project: kernel Issue ID: 289 Category: SQL nodes Reproducibility: always Severity: minor Priority: normal Status: resolved Resolution: fixed Fixed in Version: 2.0 b40 ====================================================================== Date Submitted: 01-17-2005 00:46 EET Last Modified: 01-31-2005 18:06 EET ====================================================================== Summary: Make sure that works SQL commands DIAGNOSE, COMPACT, DEFRAGMENT. Description: diagnose must produce file with all levels descriptions. compact and deframent should call all levels down. ====================================================================== Issue History Date Modified Username Field Change ====================================================================== 01-17-05 00:46 Ruslan ZasukhinNew Issue 01-31-05 18:06 Ivan Smahin Status assigned => resolved 01-31-05 18:06 Ivan Smahin Fixed in Version => 2.0 b40 01-31-05 18:06 Ivan Smahin Resolution open => fixed ====================================================================== From vbt at paradigma.ukrcom.kherson.ua Mon Jan 31 18:08:03 2005 From: vbt at paradigma.ukrcom.kherson.ua (vbt@paradigma.ukrcom.kherson.ua) Date: Mon Jan 31 10:18:18 2005 Subject: [kernel 0000290]: Check Binary fields in 2.0. Tests. Message-ID: The following issue has been RESOLVED. ====================================================================== ====================================================================== Reported By: Ruslan Zasukhin Assigned To: Ivan Smahin ====================================================================== Project: kernel Issue ID: 290 Category: Fields Reproducibility: always Severity: minor Priority: normal Status: resolved Resolution: fixed Fixed in Version: 2.0 b40 ====================================================================== Date Submitted: 01-17-2005 00:46 EET Last Modified: 01-31-2005 18:08 EET ====================================================================== Summary: Check Binary fields in 2.0. Tests. Description: We need make sure they work ====================================================================== Issue History Date Modified Username Field Change ====================================================================== 01-17-05 00:46 Ruslan ZasukhinNew Issue 01-31-05 18:08 Ivan Smahin Status assigned => resolved 01-31-05 18:08 Ivan Smahin Fixed in Version => 2.0 b40 01-31-05 18:08 Ivan Smahin Resolution open => fixed ====================================================================== From vbt at paradigma.ukrcom.kherson.ua Mon Jan 31 18:14:45 2005 From: vbt at paradigma.ukrcom.kherson.ua (vbt@paradigma.ukrcom.kherson.ua) Date: Mon Jan 31 10:18:27 2005 Subject: [V4RB 0000353]: [NEW] Advanced/ImportExport example. Message-ID: The following issue has been SUBMITTED. ====================================================================== ====================================================================== Reported By: Ruslan Zasukhin Assigned To: Ruslan Zasukhin ====================================================================== Project: V4RB Issue ID: 353 Category: Examples Reproducibility: always Severity: minor Priority: normal Status: assigned ====================================================================== Date Submitted: 01-31-2005 18:14 EET Last Modified: 01-31-2005 18:14 EET ====================================================================== Summary: [NEW] Advanced/ImportExport example. Description: This example show 2 quite complex import/export dialogs. they use a couple of new functions of V4RB that was made specially for such tasks. ====================================================================== Issue History Date Modified Username Field Change ====================================================================== 01-31-05 18:14 Ruslan ZasukhinNew Issue ====================================================================== From vbt at paradigma.ukrcom.kherson.ua Mon Jan 31 18:08:38 2005 From: vbt at paradigma.ukrcom.kherson.ua (vbt@paradigma.ukrcom.kherson.ua) Date: Mon Jan 31 10:18:33 2005 Subject: [kernel_test 0000291]: Test on FixedBinary and Varbinary fields Message-ID: The following issue has been DELETED. ====================================================================== Reported By: Ruslan Zasukhin Assigned To: Ivan Smahin ====================================================================== Project: kernel_test Issue ID: 291 Category: Fields Reproducibility: always Severity: minor Priority: normal Status: assigned ====================================================================== Date Submitted: 01-17-2005 00:48 EET Last Modified: 01-17-2005 00:48 EET ====================================================================== Summary: Test on FixedBinary and Varbinary fields Description: todo ====================================================================== Issue History Date Modified Username Field Change ====================================================================== 01-17-05 00:48 Ruslan ZasukhinNew Issue 01-31-05 18:08 Ivan Smahin Issue Deleted: 0000291 ====================================================================== From vbt at paradigma.ukrcom.kherson.ua Mon Jan 31 18:17:00 2005 From: vbt at paradigma.ukrcom.kherson.ua (vbt@paradigma.ukrcom.kherson.ua) Date: Mon Jan 31 10:18:39 2005 Subject: [V4RB 0000354]: [NEW] Valentina.GetEncodings( inIndex as integer ) as String Message-ID: The following issue has been SUBMITTED. ====================================================================== ====================================================================== Reported By: Ruslan Zasukhin Assigned To: Ruslan Zasukhin ====================================================================== Project: V4RB Issue ID: 354 Category: API Reproducibility: always Severity: minor Priority: normal Status: assigned ====================================================================== Date Submitted: 01-31-2005 18:17 EET Last Modified: 01-31-2005 18:17 EET ====================================================================== Summary: [NEW] Valentina.GetEncodings( inIndex as integer ) as String Description: Emulates array-way access to list of avaiable encodings in the ICU DLL that present on your HDD. ====================================================================== Issue History Date Modified Username Field Change ====================================================================== 01-31-05 18:17 Ruslan ZasukhinNew Issue ====================================================================== From vbt at paradigma.ukrcom.kherson.ua Mon Jan 31 18:19:29 2005 From: vbt at paradigma.ukrcom.kherson.ua (vbt@paradigma.ukrcom.kherson.ua) Date: Mon Jan 31 10:21:51 2005 Subject: [V4RB 0000355]: [Chg] VCursor.Import function now have new parameter inMaxRecordsToImport Message-ID: The following issue has been SUBMITTED. ====================================================================== ====================================================================== Reported By: Ruslan Zasukhin Assigned To: Ruslan Zasukhin ====================================================================== Project: V4RB Issue ID: 355 Category: Import/Export Reproducibility: always Severity: minor Priority: normal Status: assigned ====================================================================== Date Submitted: 01-31-2005 18:19 EET Last Modified: 01-31-2005 18:19 EET ====================================================================== Summary: [Chg] VCursor.Import function now have new parameter inMaxRecordsToImport Description: We need this to make nice import dialog. ====================================================================== Issue History Date Modified Username Field Change ====================================================================== 01-31-05 18:19 Ruslan ZasukhinNew Issue ====================================================================== From vbt at paradigma.ukrcom.kherson.ua Mon Jan 31 18:20:50 2005 From: vbt at paradigma.ukrcom.kherson.ua (vbt@paradigma.ukrcom.kherson.ua) Date: Mon Jan 31 10:21:59 2005 Subject: [V4RB 0000356]: [NEW] Valentina.ImportTextAnalyse() Message-ID: The following issue has been SUBMITTED. ====================================================================== ====================================================================== Reported By: Ruslan Zasukhin Assigned To: Ruslan Zasukhin ====================================================================== Project: V4RB Issue ID: 356 Category: Import/Export Reproducibility: always Severity: minor Priority: normal Status: assigned ====================================================================== Date Submitted: 01-31-2005 18:20 EET Last Modified: 01-31-2005 18:20 EET ====================================================================== Summary: [NEW] Valentina.ImportTextAnalyse() Description: This function answer us how many columns have the file we going to import. ====================================================================== Issue History Date Modified Username Field Change ====================================================================== 01-31-05 18:20 Ruslan ZasukhinNew Issue ====================================================================== From vidal_olivier at yahoo.fr Mon Jan 31 18:01:26 2005 From: vidal_olivier at yahoo.fr (Olivier) Date: Mon Jan 31 11:01:40 2005 Subject: API methods + binary link + M:M tables In-Reply-To: References: Message-ID: > > I do not think you need COUNTS table. > > I see only reason to have it, if you prepare some DB for CD and you > want > optimize of for speed as much as possible. > > If you develop regular db that will change runtime then I think this > will > make things more complex then needed. > > No no no. > > All should be natural and mirror the real world entities. > I believe that I badly expressed myself. Originally I import in my database the data such which (a sort of array for two dimensions) : City Surname1 Surname2 Surname3 Surname1522... PARIS 300 220 145 BORDEAUX 122 240 90 City36600 ...... thus 300 persons having the surname1 at Paris. The data are thus already grouped by surname and city. It is from these data that I work. Thus, I have to translate this enormous array into tables. For example, my application verifies the surnames, the first names and the addresses of the directory of the customer. It compares the surnames of the directory with the surnames of the database (it is a database (described higher) which contains all the French surnames and which contains for every surname the number of times when it is present in every city). If the surname of the directory is included in the database: OK. If it is not included, the application indicates that it does not know this surname, that there is credibly a spelling mistake. According to the city, the application proposes a true name which looks like the most possible of that of the directory (the one who has a spelling mistake). If there are several surnames looking like that of the directory, the application will propose then in the first one, the surname the most used in this city, then the following ones, in the order. (used array (tables) described higher) Here we are, for example, to what will be of use this DB. How to make than without "counts" table ? olivier -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: text/enriched Size: 1963 bytes Desc: not available Url : http://lists.macserve.net/pipermail/valentina-beta/attachments/20050131/e0dd0422/attachment.bin From sunshine at public.kherson.ua Mon Jan 31 20:37:51 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Mon Jan 31 12:38:00 2005 Subject: Test ignore Message-ID: -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From vbt at paradigma.ukrcom.kherson.ua Mon Jan 31 18:06:22 2005 From: vbt at paradigma.ukrcom.kherson.ua (vbt@paradigma.ukrcom.kherson.ua) Date: Mon Jan 31 12:39:57 2005 Subject: [kernel 0000289]: Make sure that works SQL commands DIAGNOSE, COMPACT, DEFRAGMENT. Message-ID: The following issue has been RESOLVED. ====================================================================== ====================================================================== Reported By: Ruslan Zasukhin Assigned To: Ivan Smahin ====================================================================== Project: kernel Issue ID: 289 Category: SQL nodes Reproducibility: always Severity: minor Priority: normal Status: resolved Resolution: fixed Fixed in Version: 2.0 b40 ====================================================================== Date Submitted: 01-17-2005 00:46 EET Last Modified: 01-31-2005 18:06 EET ====================================================================== Summary: Make sure that works SQL commands DIAGNOSE, COMPACT, DEFRAGMENT. Description: diagnose must produce file with all levels descriptions. compact and deframent should call all levels down. ====================================================================== Issue History Date Modified Username Field Change ====================================================================== 01-17-05 00:46 Ruslan ZasukhinNew Issue 01-31-05 18:06 Ivan Smahin Status assigned => resolved 01-31-05 18:06 Ivan Smahin Fixed in Version => 2.0 b40 01-31-05 18:06 Ivan Smahin Resolution open => fixed ====================================================================== From sunshine at public.kherson.ua Mon Jan 31 20:43:54 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Mon Jan 31 12:44:02 2005 Subject: FW: [kernel 0000289]: Make sure that works SQL commands DIAGNOSE, COMPACT, DEFRAGMENT. In-Reply-To: <5791f172ef82620b1ad1fdd7e416cda9@paradigma.ukrcom.kheron.ua> Message-ID: The following issue has been RESOLVED. ====================================================================== ====================================================================== Reported By: Ruslan Zasukhin Assigned To: Ivan Smahin ====================================================================== Project: kernel Issue ID: 289 Category: SQL nodes Reproducibility: always Severity: minor Priority: normal Status: resolved Resolution: fixed Fixed in Version: 2.0 b40 ====================================================================== Date Submitted: 01-17-2005 00:46 EET Last Modified: 01-31-2005 18:06 EET ====================================================================== Summary: Make sure that works SQL commands DIAGNOSE, COMPACT, DEFRAGMENT. Description: diagnose must produce file with all levels descriptions. compact and deframent should call all levels down. ====================================================================== Issue History Date Modified Username Field Change ====================================================================== 01-17-05 00:46 Ruslan ZasukhinNew Issue 01-31-05 18:06 Ivan Smahin Status assigned => resolved 01-31-05 18:06 Ivan Smahin Fixed in Version => 2.0 b40 01-31-05 18:06 Ivan Smahin Resolution open => fixed ====================================================================== ------ End of Forwarded Message From sunshine at public.kherson.ua Mon Jan 31 20:45:05 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Mon Jan 31 12:45:12 2005 Subject: FW: [kernel 0000290]: Check Binary fields in 2.0. Tests. In-Reply-To: <917be32be999a9de59740eb8df9d5c99@paradigma.ukrcom.kheron.ua> Message-ID: The following issue has been RESOLVED. ====================================================================== ====================================================================== Reported By: Ruslan Zasukhin Assigned To: Ivan Smahin ====================================================================== Project: kernel Issue ID: 290 Category: Fields Reproducibility: always Severity: minor Priority: normal Status: resolved Resolution: fixed Fixed in Version: 2.0 b40 ====================================================================== Date Submitted: 01-17-2005 00:46 EET Last Modified: 01-31-2005 18:08 EET ====================================================================== Summary: Check Binary fields in 2.0. Tests. Description: We need make sure they work ====================================================================== Issue History Date Modified Username Field Change ====================================================================== 01-17-05 00:46 Ruslan ZasukhinNew Issue 01-31-05 18:08 Ivan Smahin Status assigned => resolved 01-31-05 18:08 Ivan Smahin Fixed in Version => 2.0 b40 01-31-05 18:08 Ivan Smahin Resolution open => fixed ====================================================================== ------ End of Forwarded Message From sunshine at public.kherson.ua Mon Jan 31 20:45:13 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Mon Jan 31 12:45:23 2005 Subject: FW: [V4RB 0000353]: [NEW] Advanced/ImportExport example. In-Reply-To: <96e80fb7770077537289a844988a1957@paradigma.ukrcom.kheron.ua> Message-ID: The following issue has been SUBMITTED. ====================================================================== ====================================================================== Reported By: Ruslan Zasukhin Assigned To: Ruslan Zasukhin ====================================================================== Project: V4RB Issue ID: 353 Category: Examples Reproducibility: always Severity: minor Priority: normal Status: assigned ====================================================================== Date Submitted: 01-31-2005 18:14 EET Last Modified: 01-31-2005 18:14 EET ====================================================================== Summary: [NEW] Advanced/ImportExport example. Description: This example show 2 quite complex import/export dialogs. they use a couple of new functions of V4RB that was made specially for such tasks. ====================================================================== Issue History Date Modified Username Field Change ====================================================================== 01-31-05 18:14 Ruslan ZasukhinNew Issue ====================================================================== ------ End of Forwarded Message From sunshine at public.kherson.ua Mon Jan 31 20:45:19 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Mon Jan 31 12:45:31 2005 Subject: FW: [V4RB 0000354]: [NEW] Valentina.GetEncodings( inIndex as integer ) as String In-Reply-To: Message-ID: The following issue has been SUBMITTED. ====================================================================== ====================================================================== Reported By: Ruslan Zasukhin Assigned To: Ruslan Zasukhin ====================================================================== Project: V4RB Issue ID: 354 Category: API Reproducibility: always Severity: minor Priority: normal Status: assigned ====================================================================== Date Submitted: 01-31-2005 18:17 EET Last Modified: 01-31-2005 18:17 EET ====================================================================== Summary: [NEW] Valentina.GetEncodings( inIndex as integer ) as String Description: Emulates array-way access to list of avaiable encodings in the ICU DLL that present on your HDD. ====================================================================== Issue History Date Modified Username Field Change ====================================================================== 01-31-05 18:17 Ruslan ZasukhinNew Issue ====================================================================== ------ End of Forwarded Message From sunshine at public.kherson.ua Mon Jan 31 20:45:26 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Mon Jan 31 12:45:40 2005 Subject: FW: [V4RB 0000355]: [Chg] VCursor.Import function now have new parameter inMaxRecordsToImport In-Reply-To: Message-ID: The following issue has been SUBMITTED. ====================================================================== ====================================================================== Reported By: Ruslan Zasukhin Assigned To: Ruslan Zasukhin ====================================================================== Project: V4RB Issue ID: 355 Category: Import/Export Reproducibility: always Severity: minor Priority: normal Status: assigned ====================================================================== Date Submitted: 01-31-2005 18:19 EET Last Modified: 01-31-2005 18:19 EET ====================================================================== Summary: [Chg] VCursor.Import function now have new parameter inMaxRecordsToImport Description: We need this to make nice import dialog. ====================================================================== Issue History Date Modified Username Field Change ====================================================================== 01-31-05 18:19 Ruslan ZasukhinNew Issue ====================================================================== ------ End of Forwarded Message From sunshine at public.kherson.ua Mon Jan 31 20:45:32 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Mon Jan 31 12:45:47 2005 Subject: FW: [V4RB 0000356]: [NEW] Valentina.ImportTextAnalyse() In-Reply-To: <2e38edeb02cb8b1de24ebc18dde399c6@paradigma.ukrcom.kheron.ua> Message-ID: The following issue has been SUBMITTED. ====================================================================== ====================================================================== Reported By: Ruslan Zasukhin Assigned To: Ruslan Zasukhin ====================================================================== Project: V4RB Issue ID: 356 Category: Import/Export Reproducibility: always Severity: minor Priority: normal Status: assigned ====================================================================== Date Submitted: 01-31-2005 18:20 EET Last Modified: 01-31-2005 18:20 EET ====================================================================== Summary: [NEW] Valentina.ImportTextAnalyse() Description: This function answer us how many columns have the file we going to import. ====================================================================== Issue History Date Modified Username Field Change ====================================================================== 01-31-05 18:20 Ruslan ZasukhinNew Issue ====================================================================== ------ End of Forwarded Message From sunshine at public.kherson.ua Mon Jan 31 20:45:40 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Mon Jan 31 12:45:54 2005 Subject: FW: [V4RB 0000357]: [Chg] ImportText() and ExportText() now have new parameter 'Encoding' In-Reply-To: <6e897f4541a8230cffce556ea7584648@paradigma.ukrcom.kheron.ua> Message-ID: The following issue has been SUBMITTED. ====================================================================== ====================================================================== Reported By: Ruslan Zasukhin Assigned To: Ruslan Zasukhin ====================================================================== Project: V4RB Issue ID: 357 Category: Import/Export Reproducibility: always Severity: minor Priority: normal Status: assigned ====================================================================== Date Submitted: 01-31-2005 18:48 EET Last Modified: 01-31-2005 18:48 EET ====================================================================== Summary: [Chg] ImportText() and ExportText() now have new parameter 'Encoding' Description: this allow you control which encoding have the text file. ====================================================================== Issue History Date Modified Username Field Change ====================================================================== 01-31-05 18:48 Ruslan ZasukhinNew Issue ====================================================================== ------ End of Forwarded Message From sunshine at public.kherson.ua Mon Jan 31 23:26:06 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Mon Jan 31 15:26:16 2005 Subject: [V4RB] 2.0 b41 uploaded. Message-ID: kernel - 2.0 b41 ================ - 0000290: [Fields] Check Binary fields in 2.0. Tests. (Ivan Smahin) - 0000289: [SQL nodes] Make sure that works SQL commands DIAGNOSE, COMPACT, DEFRAGMENT. (Ivan Smahin) - 0000350: [Import/Export] Skip text file signature on import. (Igor Gomon) - 0000349: [Import/Export] Write new function ImportAnalyze(). (Igor Gomon) - 0000348: [Import/Export] Add additional parameter in Import() function - inRecCount. (Igor Gomon) - 0000346: [Locale] We need function that returns list of available Encodings (Igor Gomon) V4RB - 2.0 b41 ============== - 0000357: [Import/Export] [Chg] ImportText() and ExportText() now have new parameter 'Encoding' (Ruslan Zasukhin) - 0000347: [Import/Export] ImportText fills fields that are not in the Import cursor (Ruslan Zasukhin) - 0000356: [Import/Export] [NEW] Valentina.ImportTextAnalyse() (Ruslan Zasukhin) - 0000355: [Import/Export] [Chg] VCursor.Import function now have new parameter inMaxRecordsToImport (Ruslan Zasukhin) - 0000354: [API] [NEW] Valentina.GetEncodings( inIndex as integer ) as String (Ruslan Zasukhin) - 0000353: [Examples] [NEW] Advanced/ImportExport example. (Ruslan Zasukhin) -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- -------------- next part -------------- An HTML attachment was scrubbed... URL: http://lists.macserve.net/pipermail/valentina-beta/attachments/20050131/d580a7cd/attachment.html From jordan at digi-net.com Mon Jan 31 16:47:37 2005 From: jordan at digi-net.com (Ken Jordan) Date: Mon Jan 31 15:44:25 2005 Subject: [V4RB] 2.0 b41 uploaded. In-Reply-To: References: Message-ID: <41FEA779.40004@digi-net.com> How is the work on the Windows version progressing? Ken Ruslan Zasukhin wrote: > > kernel - 2.0 b41 > ================ > - 0000290: *[Fields]* Check Binary fields in 2.0. Tests. (Ivan Smahin) > - 0000289: *[SQL nodes]* Make sure that works SQL commands DIAGNOSE, > COMPACT, DEFRAGMENT. (Ivan Smahin) > - 0000350: *[Import/Export]* Skip text file signature on import. (Igor > Gomon) > - 0000349: *[Import/Export]* Write new function ImportAnalyze(). (Igor > Gomon) > - 0000348: *[Import/Export]* Add additional parameter in Import() > function - inRecCount. (Igor Gomon) > - 0000346: *[Locale]* We need function that returns list of available > Encodings (Igor Gomon) > > > > V4RB - 2.0 b41 > ============== > - 0000357: *[Import/Export]* [Chg] ImportText() and ExportText() now > have new parameter 'Encoding' (Ruslan Zasukhin) > - 0000347: *[Import/Export]* ImportText fills fields that are not in > the Import cursor (Ruslan Zasukhin) > - 0000356: *[Import/Export]* [NEW] Valentina.ImportTextAnalyse() > (Ruslan Zasukhin) > - 0000355: *[Import/Export]* [Chg] VCursor.Import function now have > new parameter inMaxRecordsToImport (Ruslan Zasukhin) > - 0000354: *[API]* [NEW] Valentina.GetEncodings( inIndex as integer ) > as String (Ruslan Zasukhin) > - 0000353: *[Examples]* [NEW] Advanced/ImportExport example. (Ruslan > Zasukhin) > > > -- > Best regards, > Ruslan Zasukhin [ I feel the need...the need for speed ] > ------------------------------------------------------------- > e-mail: ruslan@paradigmasoft.com > web: http://www.paradigmasoft.com > > To subscribe to the Valentina mail list go to: > http://lists.macserve.net/mailman/listinfo/valentina > ------------------------------------------------------------- > >------------------------------------------------------------------------ > >_______________________________________________ >Valentina-beta mailing list >Valentina-beta@lists.macserve.net >http://lists.macserve.net/mailman/listinfo/valentina-beta > > -------------- next part -------------- An HTML attachment was scrubbed... URL: http://lists.macserve.net/pipermail/valentina-beta/attachments/20050131/f1b3bd46/attachment.html From sunshine at public.kherson.ua Mon Jan 31 23:50:30 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Mon Jan 31 15:50:39 2005 Subject: V4RB b41, Example Advanced/ImportExport. Message-ID: Hi All, I _very_ recommend to all check this example. * it show several new functions that was added to V4RB. * it show the new features of import/export * it show practically ready for use in real world dialgos of import/export. Jochen, you should practically copy this dialogs into Vstudio, and algorithms. Everything is quite simple. This is according to import/export of text files. * it show the use of RAM-based tmp table (!!!) cool idea which did allow us quite easy resolve task how to show the first N records of import in the dialog, like FM and VAPP do. ------ NOTE: 1) we have yet 2 bugs: import of UTF16 is broken import with custom Record Delimiter broken 2) to run example, make new db, this create 2 tables, first table have 2 records. you can export T1 then import that back into T1 or T2. -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From sunshine at public.kherson.ua Mon Jan 31 23:51:24 2005 From: sunshine at public.kherson.ua (Ruslan Zasukhin) Date: Mon Jan 31 15:51:30 2005 Subject: [V4RB] 2.0 b41 uploaded. In-Reply-To: <41FEA779.40004@digi-net.com> Message-ID: On 1/31/05 11:47 PM, "Ken Jordan" wrote: > > How is the work on the Windows version progressing? Sorry Ken, Each time not enough time to finish it. I promise that I will try again now, I even have turn on my PC :-) -- Best regards, Ruslan Zasukhin [ I feel the need...the need for speed ] ------------------------------------------------------------- e-mail: ruslan@paradigmasoft.com web: http://www.paradigmasoft.com To subscribe to the Valentina mail list go to: http://lists.macserve.net/mailman/listinfo/valentina ------------------------------------------------------------- From jda at his.com Mon Jan 31 17:53:53 2005 From: jda at his.com (jda) Date: Mon Jan 31 16:54:09 2005 Subject: [V4RB] 2.0 b41 uploaded. In-Reply-To: References: Message-ID: > >- 0000289: [SQL nodes] Make sure that works SQL commands DIAGNOSE, >COMPACT, DEFRAGMENT. (Ivan Smahin) Hi Ruslan, Was this (sql compact) supposed to fix the large file problem after conversion? If so, it doesn't. I can send the three files to you again that show this if you trashed them. And I would add this to mantis. Jon