Re: Reword docs of feature "Remove temporary files after backend crash"

On Fri, Apr 1, 2022 at 9:42 AM Daniel Gustafsson <daniel@yesql.se> wrote:
> I think "useless" is a bit too strong and subjective given that it's describing
> an unknown situation out of the ordinary.  How about "outdated" or "redundant"
> (or something else entirely which is even better)?

It's the existing wording, though, and unrelated to the changes the
patch is trying to make. It also seems accurate to me.

> > I've also made a CF entry - https://commitfest.postgresql.org/35/3356/
>
> This has been sitting the CF for quite some time, time to make a decision on
> it.  I think it makes sense, having detailed docs around debugging is rarely a
> bad thing. Does anyone else have opinions?

I don't like it. It seems to me that it will result in a lot of
duplication in the docs, because every time we talk about something
that happens in connection with a crash, we'll need to talk about this
same list of exceptions. It would be reasonable to document which
conditions trigger a crash-and-restart cycle and which do not in some
centralized place, but not in every place where we mention crashing.

-- 
Robert Haas
EDB: http://www.enterprisedb.com