V4RB - Version 1.10 vs 1.11 and Encryption

[Russ Tyndall]

On Wednesday, October 26, 2005, at 04:36 PM, Ruslan Zasukhin wrote:

> Hi Russ,
>
>> Experienced a bizarre problem, thought I would post it for archive 
>> purposes.
>> Since this is just 1.1x, not sure we should agonize over it too much.
>>
>> An RB project using an encrypted Valentina DB and the 1.11 V4RB 
>> plugin began
>> inexplicably generating a Nil vCursor, even though the associated 
>> query has
>> run ok a million times.  Suspecting corruption, I restored the DB 
>> from a
>> backup.
>>
>> Unfortunately, I continued to get the Nil cursor at the same query.
>
> Strange. What if to create new fresh small database ?
>

I did not try with empty/small db.  I'm willing to try to send you 
another package that will help you duplicate the Nil cursor in 1.1x, 
although it will take me a little while.  Do you think this will 
somehow help in 2.x?  Otherwise, I am reluctant to distract you from 
2.x. Also, this would be the fourth package I have sent you in a month, 
I think.  At some point, you will start to delete my emails upon 
receipt.....  : – )

>> I ran DIAGNOSE on the DB, and the report was filled with errors.
>> I restored from an even older DB.  No change.
>
> You diagnose ECNRYPTED db ?
>
> It is know issue that for encrypted db Diagnose reports error
> Of kind: "wrong length of file ..."
>
> In fact db is okay.

Ah. Good.  I have great skill at ignoring error messages.  I'm as good 
as anybody.

>
>> Tried lots of rebuilding, exporting, importing, etc. with no luck.
>>
>> I remembered that I had upgraded V4RB from 1.10 to 1.11 as part of an
>> ill-fated attempt to work around problems upgrading project to V4RB 
>> 2.x.  On
>> a hunch, I switched out the plugins, so that RB was now using 1.10.
>>
>> This fixed the cursor, the project runs happily.
>
> May be you have some problems with record locks?
>
> I can assume that 1.11 is more strict on record locks, than 1.10
>

Parameters were never changed before or during switching out of plugin.

Here is line that makes vCursor:

   rstDB = dbAvx.SqlSelect("Select common_name, sort_id, species_no from 
species where active = 'Y' order by sort_id", kV_Client, kV_ReadOnly, 
kV_Random)

I don't know why fourth parameter is Random, could just be ForwardOnly. 
  Random access is not necessary.  Still, that *should* not cause Nil 
vCursor, should it? Even with more strict cursors in 1.11?

The app runs at least one other query automatically with the same 
parameters, from a different table.  The first query runs fine.

Same vCursor property is used in both first and second (Nil) cursor, 
and I am sure it is Nil before second cursor.  So, I do not *think* 
locks are a problem, if I understand you correctly.

Russ Tyndall
Wake Forest, NC