Dave Page wrote:
>
>
> On 14/12/05 9:12 pm, "Dave Page" <dpage@vale-housing.co.uk> wrote:
>
>
>>
>>
>>On 14/12/05 7:05 pm, "Jerry LeVan" <jerry.levan@eku.edu> wrote:
>>
>>
>>>I guess the subject says it all. I replaced 1.4.0 with
>>>1.4.1 and many, many commands kill the application.
>>>(even select * from foo limit 100).
>>
>>Eeep, that's not good :-(. I'm investigating now (thanks for the report)
>>
>
>
> Now this is strange - I can reproduce this in 1.4.1 with any query in the
> query tool, including SELECT 1; or SELECT * fROM pg_class; However, SVN
> trunk, and 1.4.0 work just fine (I've done complete rebuilds of trunk and
> 1.4.1 on the same box. Didn't bother with 1.4.0).
>
> It bails out with the following at the tip of the back trace:
>
> #0 0x00008c40 in pgConnBase::RegisterNoticeProcessor(void (*)(void*, char
> const*), void*) ()
> #1 0x0000ab08 in pgQueryThreadBase::pgQueryThreadBase(pgConnBase*, wxString
> const&, int) ()
> #2 0x00014a04 in pgQueryThread::pgQueryThread(pgConnBase*, wxString const&,
> int) ()
> #3 0x000bb714 in ctlSQLResult::Execute(wxString const&, int) ()
> #4 0x000e20bc in frmQuery::execQuery(wxString const&, int, bool, int, bool)
> ()
> #5 0x000e1b44 in frmQuery::OnExecute(wxCommandEvent&) ()
>
> Now I can't see any changes at all that should affect
> RegisterNoticeProcessor - which is even wierder when you consider that all
> the changes to 1.4.x are also present in trunk. On a hunch I rolled back to
> PostgreSQL 8.1.0, but that didn't make any difference.
>
> Anyone got any ideas? Florian, could you try a build of 1.4.1 and trunk, and
> if so, do they work for you?
Try using the previous ctlComboBox.h version. GetSelection is a virtual
method, maybe it's some compiler problem. I noticed the last wxComboBox
problem just in frmQuery, maybe the connections stored in the combobox
are screwed up.
Regards,
Andresa