Re: Assorted small doc patches - Mailing list pgsql-hackers

From David G. Johnston
Subject Re: Assorted small doc patches
Date
Msg-id CAKFQuwb_wGh0j3Qm-ZVvhhLbwzVLHGyLH54csD0Q+N15MQxs7A@mail.gmail.com
Whole thread Raw
In response to Re: Assorted small doc patches  (Peter Eisentraut <peter.eisentraut@enterprisedb.com>)
List pgsql-hackers
On Wed, Jun 1, 2022 at 7:05 AM Peter Eisentraut <peter.eisentraut@enterprisedb.com> wrote:
On 31.05.22 22:12, David G. Johnston wrote:
> Anything I should be doing differently here to get a bit of
> reviewer/committer time on these?  I'll add them to the commitfest for
> next month if needed but I'm seeing quick patches going in every week
> and the batch format done at the beginning of the month got processed
> through without issue.

These patches appear to have merit, but they address various nontrivial
areas of functionality, so either a) I just pick out a few that I
understand and deal with those and leave the rest open, or b) I'm
overwhelmed and do none.  It might have been better to post these
separately.

I did for quite a few of them, per the links provided.  But I get your point, the originals weren't on -hackers for many of them and moving them over singly probably would have worked out better.


I'll start with one though: v0001-database-default-name.patch

I don't understand why you propose this change.  It appears to reduce
precision.

As the proposed commit message says we don't tend to say "database user name" elsewhere in the documentation (or the error messages shown) so the removal of the word database doesn't actually change anything.  We only need to provide a qualifier for user name when it is not the database user name that is being referred to, or basically when it is the operating system user name.

The last hunk is the actual bug - the existing wording effectively reads:

"The default database name is the operating system user name."

That is incorrect.  It happens to be the same value when no other user name is specified, otherwise whatever gets resolved is used for the database name (but it is still a "default" because the database name was not explicitly specified).

David J.

pgsql-hackers by date:

Previous
From: Tom Lane
Date:
Subject: Re: plperl tests fail with latest Perl 5.36
Next
From: Nathan Bossart
Date:
Subject: Re: silence compiler warning in brin.c