Обсуждение: Re: pgsql: Consolidate docs for vacuum-related GUCs in new subsection

On 2025-Jan-10, Melanie Plageman wrote:

> Consolidate docs for vacuum-related GUCs in new subsection

Hmm, doesn't this need a corresponding rearrangement of the
postgresql.conf.sample file and the GUC grouping in guc_tables.c/h?

-- 
Álvaro Herrera         PostgreSQL Developer  —  https://www.EnterpriseDB.com/
"Los cuentos de hadas no dan al niño su primera idea sobre los monstruos.
Lo que le dan es su primera idea de la posible derrota del monstruo."
                                                   (G. K. Chesterton)

> On 11 Jan 2025, at 10:02, Alvaro Herrera <alvherre@alvh.no-ip.org> wrote:
>
> On 2025-Jan-10, Melanie Plageman wrote:
>
>> Consolidate docs for vacuum-related GUCs in new subsection
>
> Hmm, doesn't this need a corresponding rearrangement of the
> postgresql.conf.sample file

That's a good point.

> and the GUC grouping in guc_tables.c/h?

I don't know what our policy around this is, and maybe the backpatching hazard
isn't too bad here, but it doesn't entirely seem worth the churn.

--
Daniel Gustafsson

Daniel Gustafsson <daniel@yesql.se> writes:
> On 11 Jan 2025, at 10:02, Alvaro Herrera <alvherre@alvh.no-ip.org> wrote:
>> and the GUC grouping in guc_tables.c/h?

> I don't know what our policy around this is, and maybe the backpatching hazard
> isn't too bad here, but it doesn't entirely seem worth the churn.

I think the entire point of that categorization is to line up with the
docs, so our policy should be to fix this.

            regards, tom lane

Thanks to Álvaro for pointing this out. I didn't think of it.

On Sun, Jan 12, 2025 at 2:21 PM Tom Lane <tgl@sss.pgh.pa.us> wrote:
>
> Daniel Gustafsson <daniel@yesql.se> writes:
> > On 11 Jan 2025, at 10:02, Alvaro Herrera <alvherre@alvh.no-ip.org> wrote:
> >> and the GUC grouping in guc_tables.c/h?
>
> > I don't know what our policy around this is, and maybe the backpatching hazard
> > isn't too bad here, but it doesn't entirely seem worth the churn.
>
> I think the entire point of that categorization is to line up with the
> docs, so our policy should be to fix this.

I wrote a patch to reorder postgresql.conf.sample. But when I started
looking at guc_tables.c, it doesn't seem like those are grouped
according to the current docs order. Part of this is because some of
the GUCs have different data types. But this appears to be more than
that. For example, in master guc_tables.c,
autovacuum_vacuum_cost_limit and vacuum_cost_limit are together (in
docs in master they were in different sub-sections). Is guc_tables.c
meant to be consistent with the ordering in the docs?

- Melanie