Subject Re: IBO 4.6Bc
Author Marco Menardi
--- In, Helen Borrie <helebor@...> wrote:
> At 02:10 AM 5/08/2006, Sean Palmer wrote:
> >Well spotted, it is a typo in the message post, the actual keylinks
> >in TIBOQuery are correct.
> Typos apart :-) I confirm that you are seeing a bug, from a false
> negative being returned from CursorFields.GetByName() in
> IB_Components.pas. I've made a demo app for Jason to look at when he
> returns from family holidays...
> Meanwhile, as long as you are not selecting *both* the foreign key
> field and the linking primary key field, you *can* work around it by
> specifying the keylinks fields *without* table qualifiers. This is
> also what KeyLinksAutoDefine is doing and it's wrong. It's not a
> viable permanent solution, as it will cause ambiguity once GetByName
is fixed.
> Helen

Btw, can this be related with another bug I asked for time ago?
I had this problems in migration between 4.3.Aa and 4.5B, and I had to
"fix" my queries, making them more ambiguous (probably Firebird 2.0
will complain a lot, sigh).
Marco Menardi