[kernel 0000521]: Converting Routine can loose datas

[Claudius Sailer]

Hi Ruslan,

now problem for converting seems to be fixed, but when I understand 
solution every field could have after converting different length?

I am using Classes to define tables and I have always 504 by VarChar. 
What shall I do now here as value? 1022 and everything is fine?

Thanks for help


Claudius



Am 08. Mrz 2005 um 15:34 Uhr schrieb vbt at paradigma.ukrcom.kherson.ua:

>
> Das folgende Problem wurde behoben.
> ======================================================================
> <http://paradigma.ukrcom.kherson.ua/bt/view.php?id=521>
> ======================================================================
> Berichtet von:              Claudius Sailer
> Zugewiesen an:              Igor Gomon
> ======================================================================
> Projekt:                    kernel
> Problem ID:                 521
> Kategorie:                  Other
> Reproduzierbarkeit:         immer
> Auswirkung:                 Feature-Wunsch
> Priorität:                  normal
> Status:                     behoben
> Lösung:                     behoben
> Behoben in Version:         2.0 b56
> ======================================================================
> Erstellt am:                03-04-2005 23:50 EET
> Letzte Aktualisierung:      03-08-2005 16:34 EET
> ======================================================================
> Zusammenfassung:            Converting Routine can loose datas
> Beschreibung:
> by using converting routine of V4RB VarChar fields will be converted to
> UTF16 with same length as in V4RB 1.X. This brings the possibility of
> loosing data. It would be good, when converting routine converts field
> length too or gives the possibility with a Setup-Variable to say
> Valentina.VarcharLengthNewDatabaseIsOldDatabase = true
> Valentina.VarCharLengthNewDatabaseIsOldDatabase=false --> converting
> length from old to new that no data is lost.
> So everybody can decide whether converting routine should convert 
> length
> too or keep old values.
> ======================================================================
>
> ----------------------------------------------------------------------
>  Igor Gomon - 03-08-05 16:34
> ----------------------------------------------------------------------
> Actually when we create UTF-16 VarChar field then we specify its len in
> unicode characters, not bytes. So all the converting from 1.x to 2.x
> should take place without any data loosing. Note, that for all other
> encodings we specify the len of field in bytes not in chars.
>
> Problem-Historie
> Änderungsdatum Benutzername   Feld                     Änderung
> ======================================================================
> 03-04-05 23:50 Claudius SailerNeues Problem
> 03-05-05 20:28 Ruslan ZasukhinStatus                   neu => 
> zugewiesen
> 03-05-05 20:28 Ruslan ZasukhinBearbeitung durch         => Ivan Smahin
> 03-06-05 23:05 Ruslan ZasukhinBearbeitung durch        Ivan Smahin => 
> Igor Gomon
> 03-08-05 16:34 Igor Gomon     Status                   zugewiesen => 
> behoben
> 03-08-05 16:34 Igor Gomon     Behoben in Version        => 2.0 b56
> 03-08-05 16:34 Igor Gomon     Lösung                   offen => behoben
> 03-08-05 16:34 Igor Gomon     Problemnotiz hinzugefügt: 0000356
>
> ======================================================================
>