Created attachment 207971 [details] libiconv-1.16.patch libiconv-1.16 has been released since April. 26, 2019. Our libiconv port has Japanese specific patch which seems discontinued updating for a long time. It's time to obsolete the discontinued patch and update libiconv. The `libiconv-1.16.patch` updates converters/libiconv port. I have fixed portlint warnings, too. For people who really want to use the Japanese specific patch, I suggest to move libiconv 1.14 to japanese/libiconv114 and switch to use japanese/libiconv114 via `portmaster -o japanese/libiconv114 converters/libiconv` and `pkg set -o converters/libiconv:japanese/libiconv114`. The `libiconv114.patch` creates japanese/libiconv114 port. Please apply two patches and announce following message in UPDATING text. ``` AFFECTS: users of converters/libiconv with PATCHES option Japanese specific patch has been obsoleted. Because it seems discontinued updating for a long time. Users who really want to use the patch, Please switch to use japanese/libiconv114 port. portmaster -o japanese/libiconv114 converters/libiconv pkg set -o converters/libiconv:japanese/libiconv114 ```
Created attachment 207972 [details] libiconv114.patch
Tijl, can you have a look at this?
This requires an exp-run. (When the patch is ready and all)
Created attachment 214301 [details] libiconv-1.16.v2.patch I added a small tweak to enable some more charsets in localcharset.c, that was previously in config.charset. Other than that, the patch looks good. I'm not sure exactly how to handle the Japanese patch, or if we should create a separate port for that. That option was not enabled by default in any case. This is ready for an exp-run to get us started working on any fallout, we can think about if we want to create a libiconv114 with the Japanese patches available there afterwards.
This is ready for an exp-run.
Exp-run looks fine
A commit references this bug: Author: zeising Date: Mon May 18 18:32:01 UTC 2020 New revision: 535769 URL: https://svnweb.freebsd.org/changeset/ports/535769 Log: converters/libiconv: Update to 1.16 Update converters/libiconv to 1.16. This update removes the extra patches for CP932 and EUCJP-MS, as they are not compatible with this release. Add an UPDATING entry for this change. ABI/API stable, no need to bump dependent ports: https://abi-laboratory.pro/index.php?view=timeline&l=libiconv PR: 240954 Submitted by: Yuichiro NAITO Exp-run by: antoine Approved by: maintainer timeout (gnome@) Changes: head/UPDATING head/converters/libiconv/Makefile head/converters/libiconv/distinfo head/converters/libiconv/files/patch-include-iconv.h.in head/converters/libiconv/files/patch-libcharset__lib__config.charset head/converters/libiconv/files/patch-libcharset_lib_localcharset.c head/converters/libiconv/files/patch-src__Makefile.in head/converters/libiconv/pkg-plist
I committed the update of libiconv. If anyone needs it (and requests it), I'll add the libiconv 1.14 back, with the japanese patches. Keep this PR open for a bit in case there's fallout.
Thanks for commiting. > Keep this PR open for a bit in case there's fallout. OK. I don't close this PR as well.
Does anybody fix UPDATING message as follows? It will be matched by `pkg updating`. Index: UPDATING =================================================================== --- UPDATING (revision 536426) +++ UPDATING (working copy) @@ -6,7 +6,7 @@ you update your ports collection, before attempting any port upgrades. 20200518: - AFFECTS: users of converers/libiconv + AFFECTS: users of converters/libiconv AUTHOR: zeising@FreeBSD.org Support for the Japanese specific patches have been discontinued.
A commit references this bug: Author: zeising Date: Mon May 25 17:06:33 UTC 2020 New revision: 536482 URL: https://svnweb.freebsd.org/changeset/ports/536482 Log: UPDATING: fix typo in converters/libiconv title. PR: 240954 Reported by: Yuichiro NAITO Changes: head/UPDATING
(In reply to commit-hook from comment #11) Thank you!
can we close this PR now?
(In reply to Matthias Andree from comment #13) > can we close this PR now? I don't have any issues for this PR. I feel it's ok to close if zeising@ does same.
No reports of fallout, it is time to close this. Thanks for the reminder mandree!