On Mon, 20 Apr 2026 at 07:25, Andres Freund <andres@anarazel.de> wrote:
>
> On 2026-04-19 14:36:57 -0400, Bruce Momjian wrote:
> > That is not my recollection, and I thought I would have heard more about
> > it if that was the case.
>
> I don't know what to tell you. Just looking at emails to you with a subject
> that contains release and a body that contains performance quickly unearthed:
> David:
> https://postgr.es/m/CAApHDvrun6b+cAj6bgb6_1irnu+t7GU_uCdj1XvMQsPT0KngkQ@mail.gmail.com
This isn't a "I told you so" comment, but I do feel that it's hard to
unsee the irony in the originally omitted item being discussed there
(tidstore), which in the end was the top-listed headline feature for
v17. I feel that was just too powerful a signal to how much others do
think this stuff is worthwhile mentioning.
> > I have always hesitated to expand the list of items with concern that
> > general Postgres users will lose interest in reading it. I have in mind
> > that the release notes are not for me or hackers subscribers to read.
There is a "General Performance" section for this, so maybe people who
don't care about performance can skip these more easily.
I do agree that there is some threshold. Sometimes we do commit
patches which we know increases performance some way, but nobody
tested by how much. There are likely many of these, but one example
[1] that I don't think has any business on the release notes, but it
should help performance somehow/somewhere. It just might or might not
be measurable. Filtering out those seems good.
Maybe there's something committers can put in the commit message to
make it more obvious which commits matter by referencing some actual
performance numbers that were published that showed a definitive
speedup (not just a measurment of noise). I do expect that it's a
fairly horrible job if we're going to ask Bruce to trawl each thread
to find what performance numbers were posted. In the past I've tried
to list some example numbers in commit messages to help Bruce (e.g.
final paragraph in [2] and [3]). I'm not sure if it does help, or if
there's something better that could be done instead.
David
[1] https://git.postgresql.org/gitweb/?p=postgresql.git;a=commit;h=49ce41810faca2722424b3d8fabda79bf4902339
[2] https://git.postgresql.org/gitweb/?p=postgresql.git;a=commit;h=c456e39113809376f6604e720910ccd24e18e034
[3] https://git.postgresql.org/gitweb/?p=postgresql.git;a=commit;h=58a359e585d0281ecab4d34cab9869e7eb4e4ca3
