Re: multiple out parameters implementation

I agree with Kris' reasoning. I haven't liked this hack from the day it
was first proposed by Redhat/Fijitsu. I had come to accept it as a
necessary evil since I had given up on the backend ever supporting this
natively.

The best situation would be to have this in the backend. Second best
would be to implement Dave's changes and have the backend never support
it. Worst would be to have two incompatible implementations (in both
the driver and the backend) (think bytea vs LO but worse since I am not
sure it would be possible to toggle between two different ways of
dealing with multiple out parameters).

--Barry

-----Original Message-----
From: Kris Jurka [mailto:books(at)ejurka(dot)com]
Sent: Thursday, September 16, 2004 2:01 AM
To: Dave Cramer
Cc: pgsql-jdbc(at)postgresql(dot)org
Subject: Re: [JDBC] multiple out parameters implementation

On Wed, 1 Sep 2004, Dave Cramer wrote:

> I've managed to hack into the jdbc driver the capability for multiple
> out parameters from a postgresql function.
>
> I would like to query the list as to their thoughts, is this a useful
> feature for the driver ?
>

This is a useful feature because it's currently the only way to port
certain things, but it's clearly not very straightforward and it causes
some non trivial thinking on the parts of the database and Java
developers to get their code to match up.

Until just today I was in favor of this as we don't have any other
options or any hope for a real procedure interface in the near future,
but today on -hackers Gavin Sherry and Neil Conway indicated that they
have a proposal for implementing this feature in the backend. I'd say
lets take a look at the proposal and if we believe they've got a decent
shot at getting this in 8.1 then we hold off on this hack. The method
you proposed is pretty complicated and would introduce a pretty bad
backwards compatibility problem.

Kris Jurka

---------------------------(end of broadcast)---------------------------
TIP 3: if posting/reading through Usenet, please send an appropriate
subscribe-nomail command to majordomo(at)postgresql(dot)org so that your
message can get through to the mailing list cleanly