PostgreSQL v7.2rc2 Released

Good morning ...

On the 14th of January, 2002, the PostgreSQL Global Development
Group packaged up and released v7.2b5 (and, hopefully, our last beta) of
our upcoming release.

Release Candidate 1, baring any major bug reports from Beta5, is
schedualed for packaging and release on Friday, January 18th, with Full
Release a week later if all remains quiet.

Although primarily a documentation and packaging related release,
there are several fixes for bugs reported in v7.2b4.

For a complete listing of changes since v7.2b4, as well as the
newest source code, please go to:

ftp://ftp.postgresql.org/pub/beta

!!!!WARNING!!!! !!!!WARNING!!!! !!!!WARNING!!!!

Altho we try hard to keep this from happening, due to some
internal changes, this release *requires* a full
dump/initdb/reload from previous v7.2 betas ...

!!!!WARNING!!!! !!!!WARNING!!!! !!!!WARNING!!!!

Bug reports, as always, should be directed to
pgsql-bugs(at)postgresql(dot)org, and the severity of all bugs reported will
determine whether we move to the release cycle, or do another Beta, so we
encourage as many administrators as possible to test this current release.

Marc G. Fournier
Coordinator, PGDG

"Marc G. Fournier" <scrappy(at)hub(dot)org> writes:

> Good morning ...
>
> On the 14th of January, 2002, the PostgreSQL Global Development
> Group packaged up and released v7.2b5 (and, hopefully, our last beta) of
> our upcoming release.

RPMs for Red Hat Linux 7.2 are available at
http://people.redhat.com/teg/pg/

--
Trond Eivind Glomsrød
Red Hat, Inc.

Good morning ...

On the 26th of January, 2002, the PostgreSQL Global Development
Group packaged up and released v7.2rc2 of our upcoming release. rc1 was
released as a 'hackers-only' release, since there were several bugs quickly
found ...

Although primarily a documentation and packaging related release,
there are several fixes for bugs reported in v7.2b5 and v7.2rc1 (not formally
released).

For a complete listing of changes since v7.2b5, as well as the
newest source code, please go to:

ftp://ftp.postgresql.org/pub/beta

Note that this release does not require a full reload from b5, but,
as b5 did require one, anyone upgrading from an earlier version *will*
require one to go to RC1 ...

Bug reports, as always, should be directed to
pgsql-bugs(at)postgresql(dot)org, and the severity of all bugs reported will
determine whether we move to the release cycle, or do another Beta, so we
encourage as many administrators as possible to test this current release.

Marc G. Fournier
Coordinator, PGDG

Marc G. Fournier wrote:

> Good morning ...
>
> On the 26th of January, 2002, the PostgreSQL Global Development
> Group packaged up and released v7.2rc2 of our upcoming release. rc1 was
> released as a 'hackers-only' release, since there were several bugs quickly
> found ...

I have to admit to a certain curiosity. Who attended the Ouiji-board
session that predicted that several bugs would be found quickly,
requiring RC1 to be a "hackers-only" release instead of a real one?

OK, I can't type understandable prose any better than anyone else here,
but I did find the wording entertaining.

(And, Marc, please don't take offense, that's not the intent).

--
Don Baccus
Portland, OR
http://donb.photo.net, http://birdnotes.net, http://openacs.org

On Sat, 26 Jan 2002, Don Baccus wrote:

> Marc G. Fournier wrote:
>
> > Good morning ...
> >
> > On the 26th of January, 2002, the PostgreSQL Global Development
> > Group packaged up and released v7.2rc2 of our upcoming release. rc1 was
> > released as a 'hackers-only' release, since there were several bugs quickly
> > found ...
>
>
> I have to admit to a certain curiosity. Who attended the Ouiji-board
> session that predicted that several bugs would be found quickly,
> requiring RC1 to be a "hackers-only" release instead of a real one?
>
> OK, I can't type understandable prose any better than anyone else here,
> but I did find the wording entertaining.
>
> (And, Marc, please don't take offense, that's not the intent).

No offense .. the usual release procedure is to roll the tar ball, put it
on hackers for ~24hrs for limited testing, before a more formal release
... some of the folks at Redhat picks up a couple of issues within that
24hr period that needs to be fixed before that more formal release, so we
just held off, fixed those and re-rolled as rc2 in case some ppl had
picked up rc1 without the release announcements :)

On Sunday 27 January 2002 12:49 am, Marc G. Fournier wrote:
> On the 26th of January, 2002, the PostgreSQL Global Development
> Group packaged up and released v7.2rc2 of our upcoming release. rc1 was
> released as a 'hackers-only' release, since there were several bugs
> quickly found ...

> Note that this release does not require a full reload from b5, but,
> as b5 did require one, anyone upgrading from an earlier version *will*
> require one to go to RC1 ...

Pre-release RPMS are available for this release candidate at
ftp://ftp.postgresql.org/pub/binary/beta/RPMS

Binary RPMs built on Red Hat 7.2/Intel.

As the last beta RPMs WERE beta 3, a dump/reload is *required*.

These are pre-release RPMs. Regression passes (with the 'expected' failures
due to RHL's locale settings and the manner in which the regression tests are
built and run in this environment.).

Questions to pgsql-ports(at)postgresql(dot)org(dot)
--
Lamar Owen
WGCR Internet Radio
1 Peter 4:11

"Marc G. Fournier" <scrappy(at)hub(dot)org> writes:

> Good morning ...
>
> On the 26th of January, 2002, the PostgreSQL Global Development
> Group packaged up and released v7.2rc2 of our upcoming release. rc1 was
> released as a 'hackers-only' release, since there were several bugs quickly
> found ...
>
> Although primarily a documentation and packaging related release,
> there are several fixes for bugs reported in v7.2b5 and v7.2rc1 (not formally
> released).

Builds fine, all regression tests succeed on a Rawhideish system
(bleeding edge components) as as on Red Hat Linux 7.2.

--
Trond Eivind Glomsrød
Red Hat, Inc.

Lamar Owen writes:

> Pre-release RPMS are available for this release candidate at
> ftp://ftp.postgresql.org/pub/binary/beta/RPMS

* Still way old JDBC drivers in there.

* You're defining PGLIB in the bash profile. PGLIB doesn't do anything.

* postgresql.init:

Why do you need networking for postmaster?

Using `pidof postmaster` is wrong. There could be other postmasters
running just fine. Should use the lock file, or not care because the
postmaster or pg_ctl can handle this just fine.

Might want to implement the "reload" option.

* What happened to the contrib handling patch?

* What is going to happen to NLS?

--
Peter Eisentraut peter_e(at)gmx(dot)net

On Monday 28 January 2002 07:29 pm, Peter Eisentraut wrote:
> Lamar Owen writes:
> > Pre-release RPMS are available for this release candidate at
> > ftp://ftp.postgresql.org/pub/binary/beta/RPMS

> * Still way old JDBC drivers in there.

Yes. I still haven't gotten a good answer as to a good development
environment for the java build.

> * You're defining PGLIB in the bash profile. PGLIB doesn't do anything.

As of?

> * postgresql.init:

> Why do you need networking for postmaster?

Legacy. Shouldn't be an issue now as long as TCP/IP listening isn't on --
but then again, I don't know that, thus I have to make sure networking is up
in case it is. If there is a clean solution to this, then that would be a
good thing.

> Using `pidof postmaster` is wrong. There could be other postmasters
> running just fine. Should use the lock file, or not care because the
> postmaster or pg_ctl can handle this just fine.

In this RPM incarnation there is the assumption that only one postmaster is
running. This will change once multiple config files are easier.... :-)

> Might want to implement the "reload" option.

> * What happened to the contrib handling patch?
> * What is going to happen to NLS?

It's going to be integrated, if I can remember to apply the patch you sent.
I would appreciate patches against the current release as opposed to your
personal development spec file.
--
Lamar Owen
WGCR Internet Radio
1 Peter 4:11

Lamar Owen writes:

> Yes. I still haven't gotten a good answer as to a good development
> environment for the java build.

http://jdbc.postgresql.org/download.html might be a good place to start.

> > * You're defining PGLIB in the bash profile. PGLIB doesn't do anything.
>
> As of?

It's been unnecessary since 7.0. Beginning with 7.2 it's officially
useless.

> > Using `pidof postmaster` is wrong. There could be other postmasters
> > running just fine. Should use the lock file, or not care because the
> > postmaster or pg_ctl can handle this just fine.
>
> In this RPM incarnation there is the assumption that only one postmaster is
> running. This will change once multiple config files are easier.... :-)

But what's really wrong about it is that if the admin shuts down the
"official" postgresql service, he inadvertently kills of all users'
postmasters as well.

> > * What happened to the contrib handling patch?
> > * What is going to happen to NLS?
>
> It's going to be integrated, if I can remember to apply the patch you sent.
> I would appreciate patches against the current release as opposed to your
> personal development spec file.

Got a point there. ;-) Updated NLS patch attached. I'll see about
updating the contrib patch soon. Contrib isn't so important, but it'd be
a shame to ship without NLS.

TIA.

--
Peter Eisentraut peter_e(at)gmx(dot)net

On Tuesday 29 January 2002 12:13 am, Peter Eisentraut wrote:
> http://jdbc.postgresql.org/download.html might be a good place to start.

No, it's not a good place to start. It is an excellent place to start. :-)
Thanks for the pointer.

> > > * You're defining PGLIB in the bash profile. PGLIB doesn't do

> It's been unnecessary since 7.0. Beginning with 7.2 it's officially
> useless.

Ok.

> But what's really wrong about it is that if the admin shuts down the
> "official" postgresql service, he inadvertently kills of all users'
> postmasters as well.

True enough. I'll see if I can't work around it.

> > > * What happened to the contrib handling patch?
> > > * What is going to happen to NLS?

> > It's going to be integrated, if I can remember to apply the patch you
> > sent. I would appreciate patches against the current release as opposed
> > to your personal development spec file.

> Got a point there. ;-) Updated NLS patch attached. I'll see about
> updating the contrib patch soon. Contrib isn't so important, but it'd be
> a shame to ship without NLS.

Ok, which contrib patch was that? The only patch I hadn't yet added in was
the NLS stuff. And thanks for the patch against the current released spec
file -- it helps greatly, as I don't have to do the error-prone hand
applications that way.

I do need to survey the contrib stuff more -- but this prerelease set was
mostly to get the rc2 code out to RPM users quickly. I still have some
cleanup to do in the packaging, as you noticed. ;-)
--
Lamar Owen
WGCR Internet Radio
1 Peter 4:11

I wrote:

> * postgresql.init:

> Might want to implement the "reload" option.

Or I might. ;-)

Attached is a patch that implements "postgresql.init reload" and
"postgresql.init force-reload" as per LSB 1.1.0 Chapter 19[*].

Actually, that chapter makes for some fine reading. For instance, I quote

"""
Init files must ensure that they will behave sensibly if invoked with
start when the service is already running, or with stop when it isn't, and
that they don't kill unfortunately-named user processes.
"""
;-)

[*] http://www.linuxbase.org/spec/refspecs/LSB_1.1.0/gLSB/iniscrptact.html

--
Peter Eisentraut peter_e(at)gmx(dot)net

On Tuesday 29 January 2002 01:33 pm, Peter Eisentraut wrote:
> Attached is a patch that implements "postgresql.init reload" and
> "postgresql.init force-reload" as per LSB 1.1.0 Chapter 19[*].

> Actually, that chapter makes for some fine reading. For instance, I quote

> """
> Init files must ensure that they will behave sensibly if invoked with
> start when the service is already running, or with stop when it isn't, and
> that they don't kill unfortunately-named user processes.
> """
> ;-)

You enjoyed that, didn't you? :-)
--
Lamar Owen
WGCR Internet Radio
1 Peter 4:11