Feature freeze timezone change request

On Wed, Mar 18, 2026 at 7:07 PM Jelte Fennema-Nio <postgres@jeltef.nl> wrote:

I was notified that the tentative feature freeze timestamp is on the
wiki[1] as: April 8, 2026 0:00 UTC-12

I request that the final feature freeze timestamp be set to 0:00 UTC,
not 0:00 UTC-12.

+1.

The comitfest app currently only supports dates as
the start/end of a commitfest, not datetimes. By using 0:00 UTC I (or
anyone else with admin access) can simply fill in the last day of the
commitfest, and the commitfest will be automatically closed when that
day ends in UTC.

But not for this reason. It would be fine if the CF app closed the CF
twelve hours before or after the actual feature freeze. But using UTC
for everything is way less confusing, IMHO.

PS. I love that we're not using AoE anymore[2]

+1

--
Robert Haas
EDB: http://www.enterprisedb.com

Robert Haas <robertmhaas@gmail.com> writes:

On Wed, Mar 18, 2026 at 7:07 PM Jelte Fennema-Nio <postgres@jeltef.nl> wrote:

I was notified that the tentative feature freeze timestamp is on the
wiki[1] as: April 8, 2026 0:00 UTC-12
I request that the final feature freeze timestamp be set to 0:00 UTC,
not 0:00 UTC-12.

+1.

I think the UTC-12 business is left over from when we defined it as
AoE, but we're not doing that anymore because it's too confusing.

What we've been using lately for release freezes is noon UTC, which
personally I'd prefer on the grounds that it's not as confusing
which day is meant.

regards, tom lane

On Wed, Mar 18, 2026 at 8:40 PM Tom Lane <tgl@sss.pgh.pa.us> wrote:

What we've been using lately for release freezes is noon UTC, which
personally I'd prefer on the grounds that it's not as confusing
which day is meant.

Hmm, I kind of like that idea. I don't care that much in the end. But
less confusing is better.

--
Robert Haas
EDB: http://www.enterprisedb.com

On 19.03.26 00:07, Jelte Fennema-Nio wrote:

I was notified that the tentative feature freeze timestamp is on the
wiki[1] as: April 8, 2026 0:00 UTC-12

I request that the final feature freeze timestamp be set to 0:00 UTC,
not 0:00 UTC-12. The comitfest app currently only supports dates as
the start/end of a commitfest, not datetimes. By using 0:00 UTC I (or
anyone else with admin access) can simply fill in the last day of the
commitfest, and the commitfest will be automatically closed when that
day ends in UTC.

I don't think the end of the commitfest and feature freeze need to be
the same thing. The commitfest might as well end on 31 March or 1 April.

On Thu, 19 Mar 2026 at 10:48, Peter Eisentraut <peter@eisentraut.org> wrote:

I don't think the end of the commitfest and feature freeze need to be
the same thing. The commitfest might as well end on 31 March or 1 April.

I think it's fine for the feature freeze to end at a 12:00 UTC instead
of 0:00 UTC. But I think the commitfest should end after the feature
freeze, not before. Otherwise patches will show up as committed in
PG20-1 when in fact they are part of the PG19 release.

On 2026-03-19 Th 8:28 AM, Jelte Fennema-Nio wrote:

On Thu, 19 Mar 2026 at 10:48, Peter Eisentraut <peter@eisentraut.org> wrote:

I don't think the end of the commitfest and feature freeze need to be
the same thing. The commitfest might as well end on 31 March or 1 April.

I think it's fine for the feature freeze to end at a 12:00 UTC instead
of 0:00 UTC. But I think the commitfest should end after the feature
freeze, not before. Otherwise patches will show up as committed in
PG20-1 when in fact they are part of the PG19 release.

Sure, close it on the 9th some time.

cheers

andrew

--
Andrew Dunstan
EDB: https://www.enterprisedb.com

On Wed, Mar 18, 2026 at 10:25:33PM -0400, Robert Haas wrote:

On Wed, Mar 18, 2026 at 8:40 PM Tom Lane <tgl@sss.pgh.pa.us> wrote:

What we've been using lately for release freezes is noon UTC, which
personally I'd prefer on the grounds that it's not as confusing
which day is meant.

Hmm, I kind of like that idea. I don't care that much in the end. But
less confusing is better.

+1 for April 8th, 12:00 UTC.

--
nathan

On Wed, Mar 18, 2026 at 5:41 PM Tom Lane <tgl@sss.pgh.pa.us> wrote:

I think the UTC-12 business is left over from when we defined it as
AoE, but we're not doing that anymore because it's too confusing.

Yeah -- I swapped it recently because I noticed it was still labeled
AoE, and I didn't want anyone to have to rediscover last year's
thread. It wasn't really meant to be an endorsement for UTC-12.

What we've been using lately for release freezes is noon UTC, which
personally I'd prefer on the grounds that it's not as confusing
which day is meant.

+1 for 12:00 UTC.

--Jacob

On 3/19/26 11:37, Jacob Champion wrote:

On Wed, Mar 18, 2026 at 5:41 PM Tom Lane <tgl@sss.pgh.pa.us> wrote:

I think the UTC-12 business is left over from when we defined it as
AoE, but we're not doing that anymore because it's too confusing.

Yeah -- I swapped it recently because I noticed it was still labeled
AoE, and I didn't want anyone to have to rediscover last year's
thread. It wasn't really meant to be an endorsement for UTC-12.

What we've been using lately for release freezes is noon UTC, which
personally I'd prefer on the grounds that it's not as confusing
which day is meant.

+1 for 12:00 UTC.

+1 seems reasonable and not likely to be misunderstood

--
Joe Conway
PostgreSQL Contributors Team
Amazon Web Services: https://aws.amazon.com

On Thu, Mar 19, 2026 at 12:08:18PM -0400, Joe Conway wrote:

On 3/19/26 11:37, Jacob Champion wrote:

On Wed, Mar 18, 2026 at 5:41 PM Tom Lane <tgl@sss.pgh.pa.us> wrote:

What we've been using lately for release freezes is noon UTC, which
personally I'd prefer on the grounds that it's not as confusing
which day is meant.

+1 for 12:00 UTC.

+1 seems reasonable and not likely to be misunderstood

I updated the wiki page. Shall we also remove the "(tentative)"?

--
nathan

Nathan Bossart <nathandbossart@gmail.com> writes:

I updated the wiki page. Shall we also remove the "(tentative)"?

+1

BTW, there is some weird caching issue with that page: I did not see
the up-to-date version, even after multiple refreshes. It wasn't
till I noticed I wasn't logged into the wiki and corrected that
that I saw your edit.

regards, tom lane

On Thu, Mar 19, 2026 at 12:03 PM Tom Lane <tgl@sss.pgh.pa.us> wrote:

BTW, there is some weird caching issue with that page: I did not see
the up-to-date version, even after multiple refreshes. It wasn't
till I noticed I wasn't logged into the wiki and corrected that
that I saw your edit.

I've noticed that with other pages, but I assumed it was
intentional/robot-related.

--Jacob