Обсуждение: 7.3.5

On Tue, Nov 04, 2003 at 23:16:15 -0000,
  Donald Fraser <demolish@cwgsy.net> wrote:
> Will there be a 7.3.5 version released, containing relevant patches, before the release of the 7.4 version?

That seems unlikely. It seems likely that there will be such a release, but
most likely it will be after the 7.4 release.

You can get a copy of 7.3 from CVS. I don't remember the appropiate tag
name, but it was posted on one of the lists in the last couple of days.

Bruno Wolff III <bruno@wolff.to> writes:
> On Tue, Nov 04, 2003 at 23:16:15 -0000,
>   Donald Fraser <demolish@cwgsy.net> wrote:
>> Will there be a 7.3.5 version released, containing relevant patches, before the release of the 7.4 version?

> That seems unlikely. It seems likely that there will be such a release, but
> most likely it will be after the 7.4 release.

Yes.  I'd expect we'd wait till a little while after 7.4, since often
interesting bugs are first reported just after a release --- there may
be things we learn that ought to be back-patched into 7.3.5.  There is
not currently anything in the 7.3 branch that looks like an immediate
release forcer to me.  (I posted a summary of recent 7.3 commit messages
to pgsql-hackers a couple days ago, if you're interested.)

> You can get a copy of 7.3 from CVS. I don't remember the appropiate tag
> name, but it was posted on one of the lists in the last couple of days.

The tip of the REL7_3_STABLE branch is the closest present approximation
to what 7.3.5 might look like.

            regards, tom lane

----- Original Message -----
> >> Will there be a 7.3.5 version released, containing relevant patches,
before the release of the 7.4 version?
>
> > That seems unlikely. It seems likely that there will be such a release, but
> > most likely it will be after the 7.4 release.
>
> Yes.  I'd expect we'd wait till a little while after 7.4, since often
> interesting bugs are first reported just after a release --- there may
> be things we learn that ought to be back-patched into 7.3.5.  There is
> not currently anything in the 7.3 branch that looks like an immediate
> release forcer to me.  (I posted a summary of recent 7.3 commit messages
> to pgsql-hackers a couple days ago, if you're interested.)

Hi Tom,
I systematically went through every message you posted on pgsql-hackers
back dating to the 28th Oct.
I didn't find the summary you mentioned - there were a lot of messages
(how you manage to get work done between messages must be a small
mirical) so there is a chance I missed it.

The fix I am looking for in 7.3.5, which is in the 7.4 branch, is for the
regproc-reverse-conversion problem (the error comes when the planner tries
to convert that string back to OID form).

Regards
Donald Fraser

"Donald Fraser" <demolish@cwgsy.net> writes:
> The fix I am looking for in 7.3.5, which is in the 7.4 branch, is for the
> regproc-reverse-conversion problem (the error comes when the planner tries
> to convert that string back to OID form).

IIRC there is no solution to that without moving to 7.4 --- the fix
required a change in the representation of data values in pg_statistic.

The best workaround I can recommend in 7.3 is to delete the pg_statistic
rows that relate to regproc columns.

            regards, tom lane