Thread: Re: [COMMITTERS] pgsql: Stamp releases notes for 8.2.3, 8.1.8, 8.0.12.
Andrew Dunstan wrote: > Bruce Momjian wrote: > > Andrew Dunstan wrote: > > > >> Bruce Momjian wrote: > >> > >>> Log Message: > >>> ----------- > >>> Stamp releases notes for 8.2.3, 8.1.8, 8.0.12. > >>> > >>> Tags: > >>> ---- > >>> REL8_2_STABLE > >>> > >>> > >>> > >> This missed the fix for psql "\copy (query)" which I committed the other > >> day for 8.2, and should probably have been in the release notes. What's > >> the protocol there? > >> > > > > I did not mention it because it merely fixes a rare bug. It is not > > something I would normally mention in the release notes. > > > > > > ok. If there is some value to capturing every change, but still keeping the release notes a readable length, please let me know. -- Bruce Momjian <bruce@momjian.us> http://momjian.us EnterpriseDB http://www.enterprisedb.com + If your life is a hard drive, Christ can be your backup. +
On Wed, 2007-02-07 at 13:21 -0500, Bruce Momjian wrote: > If there is some value to capturing every change, but still keeping the > release notes a readable length, please let me know. If it fixes a real, non-theoretical bug and has been backpatched to a stable release branch, I would say in most cases it is worth documenting in the release notes. Describing every change made in a new feature release (i.e. 8.3.0) would be far too much verbiage, but far fewer changes are made to stable branches. Also, documenting all the significant changes in stable branch releases is valuable to let users identify possible regressions. -Neil
Neil Conway <neilc@samurai.com> writes:
> If it fixes a real, non-theoretical bug and has been backpatched to a
> stable release branch, I would say in most cases it is worth documenting
> in the release notes. Describing every change made in a new feature
> release (i.e. 8.3.0) would be far too much verbiage, but far fewer
> changes are made to stable branches. Also, documenting all the
> significant changes in stable branch releases is valuable to let users
> identify possible regressions.
Neil has a good point: the documentation policy should be different for
updates to stable branches than it is for a new major release. I think
Bruce's "too small to bother with" policy is about right for major
releases, but if we've bothered to back-patch something then it's
usually worth documenting.
regards, tom lane
Tom Lane wrote: > Neil Conway <neilc@samurai.com> writes: > > If it fixes a real, non-theoretical bug and has been backpatched to a > > stable release branch, I would say in most cases it is worth documenting > > in the release notes. Describing every change made in a new feature > > release (i.e. 8.3.0) would be far too much verbiage, but far fewer > > changes are made to stable branches. Also, documenting all the > > significant changes in stable branch releases is valuable to let users > > identify possible regressions. > > Neil has a good point: the documentation policy should be different for > updates to stable branches than it is for a new major release. I think > Bruce's "too small to bother with" policy is about right for major > releases, but if we've bothered to back-patch something then it's > usually worth documenting. OK, I am more liberal in adding to a minor release, but I avoid cases where the bug has been around for a long time and the error case is rare. -- Bruce Momjian <bruce@momjian.us> http://momjian.us EnterpriseDB http://www.enterprisedb.com + If your life is a hard drive, Christ can be your backup. +