Port zar forward to 0.14.1 #56

EnronEvolved commented 2025年08月09日 19:30:24 +02:00 (Migrated from github.com)

This pull request is a minimal bodge-job that allows zar to build successfully on 0.14.1, and pass zig build test without issue.

Due to ongoing writergate changes, an update to the latest Zig will have to wait until a release is cut.

This pull request is a minimal bodge-job that allows `zar` to build successfully on 0.14.1, and pass `zig build test` without issue. Due to ongoing [writergate](https://github.com/ziglang/zig/pull/24329) changes, an update to the latest Zig will have to wait until a release is cut.

moosichu commented 2025年08月10日 15:31:31 +02:00 (Migrated from github.com)

Some of the tests fail with this PR - just taking a quick look. :) Thank you for doing this work!

Some of the tests fail with this PR - just taking a quick look. :) Thank you for doing this work!

moosichu commented 2025年08月10日 17:47:45 +02:00 (Migrated from github.com)

https://github.com/moosichu/zar/tree/EnronEvolved-port-forward

I've setup a branch for this work... just as a heads up, please do make sure that you use the zig formatter when editing code. Spotted a few todos in here as well - would be nice if we could get the tests working again and also resolve those todos. Managed to resovle the build options - but hopefully now after formatting the diff will be a bit simpler to reason about as well.

https://github.com/moosichu/zar/tree/EnronEvolved-port-forward I've setup a branch for this work... just as a heads up, please do make sure that you use the zig formatter when editing code. Spotted a few todos in here as well - would be nice if we could get the tests working again and also resolve those todos. Managed to resovle the build options - but hopefully now after formatting the diff will be a bit simpler to reason about as well.

EnronEvolved commented 2025年08月10日 21:58:01 +02:00 (Migrated from github.com)

(削除) Which tests are failing? I can't reproduce the test failures locally. I've just been running zig build test. Are there any tests left? (削除ここまで)

Edit: I found the CI results on the branch. You're testing it with the nightly version of Zig. This code specifically targets version 0.14.1, and there are breaking changes between nightly and 0.14.1. I strongly suggest sticking to 0.14.1 until the 0.15.0 release is cut, and writergate becomes a stationary target.

Probably could have just merged the pull request on to that branch. It would make this situation much less confusing.

~~Which tests are failing? I can't reproduce the test failures locally. I've just been running `zig build test`. Are there any tests left?~~ **Edit**: I found the CI results on the branch. You're testing it with the nightly version of Zig. This code specifically targets version 0.14.1, and there are breaking changes between nightly and 0.14.1. I strongly suggest sticking to 0.14.1 until the 0.15.0 release is cut, and writergate becomes a stationary target. Probably could have just merged the pull request on to that branch. It would make this situation much less confusing.

EnronEvolved commented 2025年08月10日 23:51:36 +02:00 (Migrated from github.com)

I added a null commit at the end so this pull request can be "merged" rather than closed.

I added a null commit at the end so this pull request can be "merged" rather than closed.

moosichu commented 2025年08月11日 21:18:20 +02:00 (Migrated from github.com)

Sorry - just to be clear I'm testing locally. And the issue is specific to Darwin. I wasn't even looking at the CI tests but I will make them track against 14.0.1 (and merge it all tomorrow).

Probably could have just merged the pull request on to that branch. It would make this situation much less confusing.

I couldn't push to your branch and was planning to rebase before merging - and also just figure out what was causing the test failures.

Sorry - just to be clear I'm testing locally. And the issue is specific to Darwin. I wasn't even looking at the CI tests but I will make them track against 14.0.1 (and merge it all tomorrow). > Probably could have just merged the pull request on to that branch. It would make this situation much less confusing. I couldn't push to your branch and was planning to rebase before merging - and also just figure out what was causing the test failures.

moosichu commented 2025年08月15日 10:56:55 +02:00 (Migrated from github.com)

Sorry about being late there - got it in, and thank you so much for doing the bulk of the work on this one. Merged it now as-is (haven't gotten around to sorting the tests yet) - as I thought that might unblock things and didn't want to cause frustration there.

Sorry about being late there - got it in, and thank you so much for doing the bulk of the work on this one. Merged it now as-is (haven't gotten around to sorting the tests yet) - as I thought that might unblock things and didn't want to cause frustration there.

moosichu commented 2025年08月15日 10:58:57 +02:00 (Migrated from github.com)

Also if you feel like my slowness is too much of a blocker and you want to get on with things please do let me know and I'm happy to grant commit access to the repo. I think it would be good to focus on getting the tests green again next though.

Also if you feel like my slowness is too much of a blocker and you want to get on with things please do let me know and I'm happy to grant commit access to the repo. I think it would be good to focus on getting the tests green again next though.

EnronEvolved commented 2025年08月15日 11:20:45 +02:00 (Migrated from github.com)

Not to worry: I'm quite patient. Writergate, and related stdlib refactors, are the bigger blocker.

I'll leave twiddling test setups to someone who actually knows how to use Github's CI; I might do a little light code cleanup in the meantime.

Not to worry: I'm quite patient. Writergate, and related stdlib refactors, are the bigger blocker. I'll leave twiddling test setups to someone who actually knows how to use Github's CI; I might do a little light code cleanup in the meantime.