| From: | Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> |
|---|---|
| To: | Peter Eisentraut <peter(at)eisentraut(dot)org> |
| Cc: | Jelte Fennema-Nio <postgres(at)jeltef(dot)nl>, Robert Haas <robertmhaas(at)gmail(dot)com>, Michael Paquier <michael(at)paquier(dot)xyz>, Andres Freund <andres(at)anarazel(dot)de>, Alvaro Herrera <alvherre(at)alvh(dot)no-ip(dot)org>, Bharath Rupireddy <bharath(dot)rupireddyforpostgres(at)gmail(dot)com>, Daniel Gustafsson <daniel(at)yesql(dot)se>, PostgreSQL Hackers <pgsql-hackers(at)postgresql(dot)org> |
| Subject: | Re: Support a wildcard in backtrace_functions |
| Date: | 2025-02-26 15:40:17 |
| Message-ID: | 1775094.1740584417@sss.pgh.pa.us |
| Views: | Raw Message | Whole Thread | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-hackers |
Peter Eisentraut <peter(at)eisentraut(dot)org> writes:
> I think the additions to typedefs.list are useless. Since nothing
> actually uses these types, the collection on the buildfarm won't find
> any uses of them and thus won't include it in its list, and then the
> next update from the buildfarm will overwrite your changes.
It's not quite that simple. Lately, at least, we haven't just blindly
replaced typedefs.list with the buildfarm's list. When I do it I do
a sanity check on the additions and removals, partly by seeing what
indentation changes result. A manually-added typedef that happened
not to be in the buildfarm's list would probably survive, if removing
it resulted in any ugly changes.
Still, it is better to make sure the typedef is actually used to
declare some variables, so that you're not so dependent on the
amount of manual curation that happens. If there's no place where
that can plausibly be done, maybe you don't need the typedef after
all?
regards, tom lane
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Tom Lane | 2025-02-26 16:13:54 | Re: Statistics Import and Export |
| Previous Message | Fujii Masao | 2025-02-26 15:36:49 | Re: Extend postgres_fdw_get_connections to return remote backend pid |