Re: Custom tuplesorts for extensions

Hi Justin!

On Mon, Jul 25, 2022 at 2:23 AM Justin Pryzby <pryzby(at)telsasoft(dot)com> wrote:
> Note that 0001+0002 (without the others) incurs warnings:
>
> $ time { make -j4 clean; make -j4; } >/dev/null
> tuplesort.c:1883:9: warning: unused variable 'i' [-Wunused-variable]
> tuplesort.c:1955:10: warning: unused variable 'i' [-Wunused-variable]
> tuplesort.c:2026:9: warning: unused variable 'i' [-Wunused-variable]
> tuplesort.c:2103:10: warning: unused variable 'i' [-Wunused-variable]
>
> (I wondered in the past if cfbot should try to test for clean builds of subsets
> of patchsets, and it came up recently with the JSON patches.)
>
> Also, this comment has some bad indentation:
>
> * Set state to be consistent with never trying abbreviation.

Thank you for caching this. Fixed in the revision attached.

Testing subsets of patchsets in cfbot looks like a good idea to me.
However, I'm not sure if we always require subsets to be consistent.

------
Regards,
Alexander Korotkov