Message-ID: <38E7B63B.C184B7B1@alumni.caltech.edu>
Date: Sun, 02 Apr 2000 21:06:03 +0000
From: Thomas Lockhart <lockhart@alumni.caltech.edu>
MIME-Version: 1.0
To: Byron Nikolaidis <byron.nikolaidis@home.com>
CC: Hiroshi Inoue <Inoue@tpf.co.jp>, Tibor Laszlo <ltibor@mail.tiszanet.hu>,
	pgsql-hackers <pgsql-hackers@postgresql.org>,
	pgsql-interfaces@postgresql.org
Subject: Re: "list index out of range" in C++ Builder 4
References: <001501bf949d$38ff6060$2801007e@tpf.co.jp>
	<38DA3EB6.1448F04C@alumni.caltech.edu> <38DD546C.10127953@home.com>
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit

> > OK, here is a patch to help fix up trouble with indices and primary
> > keys in the ODBC interface.
> From what I remember, not too many things use SQLPrimaryKeys, instead
> they use SQLStatistics to get the key info along with other more
> detailed information about an index.  Originally, we never even
> supported SQLPrimaryKeys.   Only recently, we added support for it along
> with SQLForeignKeys.  But that was then, this is now. :)
> And Thomas, is your patch compatible with the other versions of Postgres
> the odbc driver is supposed to support (it still supports all the way
> back to 6.2), or do we no longer care about that?

Hmm. The patch is not backward compatible. Presumably there is
provision in the code to know what version of DBMS to connect to? In
that case, I can probably carry along two different queries and a few
differences in handling the OID results.

Suggestions?

> If you think your patch is working, I could check out the cvs code,
> compile it on windows, and post the dll.

I'll try poking at it some more. Let me know if you have some tips or
want to do this yourself (we're trying to get a v7.0 release in the
next couple of weeks).

Regards.

                       - Thomas

-- 
Thomas Lockhart				lockhart@alumni.caltech.edu
South Pasadena, California