Dave Page wrote:
>
> > -----Original Message-----
> > From: Jim Hines [mailto:jhines@wdtv.com]
> > Sent: 04 November 2002 20:06
> > To: pgsql-odbc@postgresql.org; Dave Page
> > Subject: Re: [ODBC] CTID: Anyway to lock it?
> >
> > The folks at Chyron say they are not specifiing CTID and that
> > it must be
> > within the ODBC driver. Could this be the case? If so, what
> > setting would
> > control this?
>
> Hmm, yes - it appears to be in inner_process_tokens(QueryParse *qp,
> QueryBuild *qb) in convert.c. I don't know what options might affect
> that though I'm afraid - do you know Hiroshi?
The driver uses CTID and OID to implement updatable cursors.
Please turn off the *Updatable cursors* DSN option if you
don't like the behavior.
regards,
Hiroshi Inoue
http://w2422.nsk.ne.jp/~inoue/