On Tue, 18 May 2021 at 14:09, Peter Smith <smithpb2250@gmail.com> wrote:
> Yeah, I would like to work my way through all of these warnings in my
> spare time and report back to this thread (after 1-2 months?) with a
> detailed analysis.
I'd recommend for any patches that they come in bite-sized chunks. A
committer is going to have to re-review each change. For me
personally, I'll probably run for the hills if I see a patch that
renames 200 variables.
I'd think about a dozen would be good. Starting with ones that are
least likely to raise objection also seems like a good idea. That way
you'll have an idea if you want to trouble yourself with the more
questionable ones when the less questionable ones raised too many
questions. Like I mentioned, start with ones like i, buf, tmp, lc. If
those are accepted then move on to the more difficult ones. Unless
you discover bugs, then there's not really any urgency to fix these.
Doing it in bite-sized chunks is less likely going to cause
frustration for you if some of the work is rejected after you've gone
to all the trouble.
Also, going by what's mentioned in [1], in particular [2], I'm not so
certain that these changes will be received well by everyone. So I
recommend just taking it slow.
David
[1] https://www.postgresql.org/message-id/flat/877k1psmpf.fsf%40mailbox.samurai.com
[2] https://www.postgresql.org/message-id/22920.1069708226%40sss.pgh.pa.us