Created attachment 228745 [details]
patch for audio/libamrwb
Switched from PORTVERSION to DISTVERSION.
PORTREVISION has already been bumped due to changes in pkg query %e results
URLs such as MASTER_SITES have been switched from http to https.
Variable DISTNAME is now close to the default value in bsd.port.mk.
Variables have been sorted to silence portclippy :)
Regenerated distinfo by running make makesum "FORCE_FETCH_ALL=yes" "FETCH_ARGS=-pm".
Created attachment 228746 [details]
A commit in branch main references this bug:
Author: Tatsuki Makino <email@example.com>
AuthorDate: 2021-10-24 20:19:31 +0000
Commit: Daniel Engberg <diizzy@FreeBSD.org>
CommitDate: 2021-10-24 20:19:48 +0000
audio/libamrwb: Makefile adjustment
Adjust Makefile to follow Porter's Handbook and framework more closely,
also pet portclippy while at it.
Approved by: arrowd (mentor)
Differential Revision: https://reviews.freebsd.org/D32622
audio/libamrwb/Makefile | 14 +++++++-------
audio/libamrwb/distinfo | 1 +
audio/libamrwb/pkg-descr | 2 +-
3 files changed, 9 insertions(+), 8 deletions(-)
Committed, thanks for your contribution!
USES=+autoreconf part was not explained, ditto for bug 259202. Attaching build logs, esp. successful, is meaningless, nobody looks at them, so all they do is clutter bugzilla and eat disk space.
Yes, I forgot to write about autoreconf.
The scripts already generated by autotools are supposed to be fine, but I added them to keep up with any changes in the new version of autotools.
The poudriere log of successful builds is used as evidence that there are no problems with the patch.
Some committers require maintainer to submit poudriere log and will not commit without it.
(In reply to Tatsuki Makino from comment #5)
> The poudriere log of successful builds is used as evidence that there are
> no problems with the patch.
Simply saying that "poudriere build ok" is more than enough. Any committer would have to rebuild the port(s) herself prior to committing anyway. Last but not least, user-provided build logs simply cannot be trusted.
> Some committers require maintainer to submit poudriere log and will not
> commit without it.
This requirement is nonsense, you can safely ignore it. Again, any committer must perform the build- and, ideally, run-test before committing anything, and may not shift this responsibility to submitter or maintainer.
(In reply to Alexey Dokuchaev from comment #6)
If my log submission is not required, I will not submit it :)
However, there is also the issue of the distribution of time resources between mine(maintainers), yours(commiters), and the build machine's.
I want to help committers save as much time as possible so that they can process as many reports as possible.
I don't know the performance of committer's build machine, but if committer has maintainer's logs, committer can find problems that maintainer missed before using build machine. Committer can just look through that log while build machine is doing other work.