Point heard and if someone will root this out and show me where IBO needs to
be changed I will fully consider it. I'll log this as an issue for future
consideration so that if I get time to investigate more closely I will do
so. Please understand this isn't a high priority item for the time being.

Regards,
Jason Wharton
CPS - Mesa AZ
http://www.ibobjects.com

----- Original Message -----
From: "Nando Dessena" <nandod@...>
To: <IBObjects@yahoogroups.com>
Sent: Tuesday, March 12, 2002 3:49 AM
Subject: Re: [IBO] CancelUpdates


> Jason,
>
> > I clearly see your point but I consider that a problem with the BDE
code. I
> > prefer to have bad code dealt with than to candy coat over it.
> >
> > Does anyone else feel as Hans that the change to ignore dsInactive state
in
> > the dataset when assigning a field value is fine? Obviously, I was not
aware
> > the BDE would allow this. Perhaps a look in the VCL again would make
Hans
> > case better...
>
> for what I understand, the problem lies in the fact that the BDE never
> calls a data-aware control's OnChange event handler when the dataset is
> not in dsEditModes, whereas it seems that IBO is doing just that (please
> ignore me if I don't understand right).
>
> May I ask what is the exact combo object type Hans is using? if it's
> TDBLookupComboBox, then the problem may be the TIBOQuery datalink
> sending change notifications to the controls when the dataset is not in
> edit mode, if it is another type of control then let's see what.
>
> I, for one, would prefer BDE-compatibility, otherwise many applications
> that don't check the state of the dataset in OnChange handlers would be
> broken. This scares me because I know I have a lot of BDE code that
> works this way (with fields events, not controls) and I will eventually
> have to port it all to IBO-TDataSet, for which I need perfect
> compatibility (for better or for worse).
> Ciao
> --
> ____
> _/\/ando