[V4Rb] VDatabase reference problem

[Charles Yeomans]

On Tuesday, June 17, 2003, at 12:48 PM, Ruslan Zasukhin wrote:

> on 6/17/03 19:07, Charles Yeomans at yeomans at desuetude.com wrote:
>
>>
>> On Tuesday, June 17, 2003, at 03:42 AM, Ruslan Zasukhin wrote:
>>
>>> on 6/17/03 2:38, Charles Yeomans at yeomans at desuetude.com wrote:
>>>
>>>> I finally tracked down the source of an intermittent error in my
>>>> application.  Unfortunately, it took me quite a while because
>>>> VBaseObject.Database.ErrNumber was not being set; looking in another
>>>> VDatabase object that pointed to the same database revealed the
>>>> problem. I can understand why this function returns a different
>>>> object,
>>>> but it would be nice if the object properties were kept in sync so
>>>> that
>>>
>>> Hmm, actually objects must be the same, otherwise we cannot avoid
>>> problems
>>> like you describe.
>>>
>>> I need check this.
>>> I was sure objects are the same.
>>
>> As it turns out, they are.  Let me offer more detail.
>>
>> The problem was that V4Rb was returning error 666; my registration 
>> code
>> obfuscation was returning a bad result.  What I am seeing is this:  
>> the
>> global database reference I hold and the reference obtained from
>> VBaseObject.Database point to the same object.  But when I test the
>> ErrNumber property, my global database reference returns 666, while 
>> the
>> reference obtained from VBaseObject.Database returns 0.  This doesn't
>> seem possible, but I am looking at it in the debugger right now.
>
> I think I know where is problem.
>
> Probably VBaseObject.Database clear error.
>
> I think after it and after you see error = 0,
> If you look again into global object it also will be zero.
> Check this please and let me know.

You are right; a subsequent check of the global reference's ErrNumber 
property returned 0.  This is the current beta of V4Rb, by the way.

Charles Yeomans