On 2020/02/26 15:53, torikoshia wrote: > Hi, > > I'm preparing for the coming release and attached a patch for updating documents. > > The release date isn't fixed, so the date > is temporarily '202002XX'. > > Any thoughts? Thanks for starting this discussion and making the patch! +1 to prepare the release of new minor version of pg_bigm 1.2. We should release new minor version of pg_bigm 1.1, too? The last minor version was released 2016 and there are some fixes since that. IMO probably it's time to release that. Thought? +<li><p>Fix compiler warning in bigmstrcmp() (Torikoshi Atsushi, Fujii Masao)</p></li> bigmstrcmp() is not user-visible function, so I think that it's better not to expose it even in release note. Thought? So, what about "Fix compiler warning caused when built with PostgreSQL 12 or later."? + <p>Stabilize the result of pg_gin_pending_stats() regression test (Torikoshi Atsushi, Fujii Masao)</p> + <p>Previously, this test failed on PostgreSQL built with non-default block size.</p> Like PostgreSQL's release note, isn't it better to exclude the change about regression test from the release note? Because users basically are not interested in the regression test. *If* we write down something in release note regarding this change, we should document something like "pg_bigm was confirmed to work with PostgreSQL buillt with non-default block size", instead. This is very minor information that almost all users don't care about, though. So, regarding the change of regression test, I'm inclined to add nothing into the release note. Thought? Regards, -- Fujii Masao NTT DATA CORPORATION Advanced Platform Technology Group Research and Development Headquarters
Fork