Re: - SQLCloseCursor

Peter Deacon wrote:

> On Tue, 20 Jan 2004, Paul wrote:
>
>  
>
> > On Tue, 2004-01-20 at 04:30, ZIGLIO, Frediano, VF-IT wrote:
> >    
>
>  
>
> > > > Hello, first I'd like to thank all the developers for a great program
> > > > for *nix developers. It's great to know that the C code I'm
> > > > writing will
> > > > work on Windows boxes as well as Linux.
> > > >        
>
> I remember trying to trace a problem with unixODBC that is still
> outstanding for me.  Havn't tried the latest version yet.  I was able to
> workaround it by changing my use of functions somewhat like whats
> described below.
>
> From what I saw the cause was the message/error chain would never be freed
> however the DM still appears to keep references to it which makes the
> problem hard to spot using a memory profiler.
>
>  
Thats been changed in the last release, I have given up on the old old 
ODBC 2 action of queing up error messages. They "should" be cleared now 
at each non-diag function call.

--
Nick