diff options
author | Tom Lane <tgl@sss.pgh.pa.us> | 2024-01-10 18:09:29 -0500 |
---|---|---|
committer | Tom Lane <tgl@sss.pgh.pa.us> | 2024-01-10 18:09:29 -0500 |
commit | 3d185cfc09c32b1f250adfbb62c0c7a3f96c5af2 (patch) | |
tree | ba1980b95f7cbe9cd89c028ee76f654035b454cf /doc/src | |
parent | 390298f0806588ceb131dd47760208bd2cd6298f (diff) | |
download | postgresql-3d185cfc09c32b1f250adfbb62c0c7a3f96c5af2.tar.gz postgresql-3d185cfc09c32b1f250adfbb62c0c7a3f96c5af2.zip |
Restore initdb's old behavior of always setting the lc_xxx GUCs.
In commit 3e51b278d I (tgl) caused initdb to leave lc_messages and
other lc_xxx GUCs commented-out in the installed postgresql.conf file
if they were going to be set to 'C'. This was a hack for cosmetic
purposes, and it was buggy because lc_messages' wired-in default is
not 'C' but '' (empty string). That led to --no-locale not having
the expected effect, since the postmaster would then obtain
lc_messages from its startup environment.
Let's just revert to the prior behavior of always de-commenting the
lc_xxx entries; the argument for changing that longstanding behavior
was weak in the first place.
Also, fix postgresql.conf.sample's erroneous claim that the default
value of lc_messages is 'C'. I suspect that was what misled me into
making this mistake in the first place.
Report and patch by Kyotaro Horiguchi. Back-patch to v16 where
the problem was introduced.
Discussion: https://postgr.es/m/20231122.162700.1995154567625541112.horikyota.ntt@gmail.com
Diffstat (limited to 'doc/src')
0 files changed, 0 insertions, 0 deletions