Add support for the Friendly Forge Format (F3)

[earl-warren]

This pull request is a mirror of a corresponding pull request on gitea.com, please go there to discuss and comment.

---

Continuation of this previous pull request

[earl-warren]

How can it be that lint on windows reports an error and lint on GNU/Linux does not?

[silverwind]

How can it be that lint on windows reports an error and lint on GNU/Linux does not?

I think the only logical reason would be presence of //go:build tags that run only on Windows, but that doesn't seem to be the case here. Might just restart and see if it goes away.

[earl-warren]

How can it be that lint on windows reports an error and lint on GNU/Linux does not?

I think the only logical reason would be presence of //go:build tags that run only on Windows, but that doesn't seem to be the case here. Might just restart and see if it goes away.

But there are no //go:build tags in this PR. Is there a way to re-launch the CI and verify this is not a false negative?

[silverwind]

Just push an empty commit to restart CI:

git commit --allow-empty -m 'restart ci'

[earl-warren]

Now that's a new and interesting false negative.

[earl-warren]

The CI is broken.

[delvh]

We know. https://github.com/go-gitea/gitea/pull/21108

[earl-warren]

Now that #21108 is merged, the CI is back to another false negative: failing lint on windows but not on GNU/Linux, despite the absence of build tags.

[earl-warren]

Now that #21108 is merged, the CI is back to another false negative: failing lint on windows but not on GNU/Linux, despite the absence of build tags.

This pull request is not compiled for windows to workaround the linter false negative.

[earl-warren]

The arm64 build crashes go 1.19

Because the gitea/test-env image for arm64 has 1.19 and not 1.19.1 which is missing a critical bug fix.

The test-env image must be rebuilt for arm64 and published.

[earl-warren]

The arm64 test image was rebuilt with 1.19.1, trying again.

[silverwind]

Side note:

This pull request is a mirror of a corresponding pull request on gitea.com, please go there to discuss and comment.

gitea core development takes place on GitHub, so it should be reviewed/discussed here :)

[earl-warren]

Side note:

This pull request is a mirror of a corresponding pull request on gitea.com, please go there to discuss and comment.

gitea core development takes place on GitHub, so it should be reviewed/discussed here :)

I did not want to force people to agree to the GitHub terms of service to participate in this particular pull request. Moving the discussion to gitea.com is more inclusive in a way that is not disruptive to the Gitea development process.

[earl-warren]

Side note:

This pull request is a mirror of a corresponding pull request on gitea.com, please go there to discuss and comment.

gitea core development takes place on GitHub, so it should be reviewed/discussed here :)

I did not want to force people to agree to the GitHub terms of service to participate in this particular pull request. Moving the discussion to gitea.com is more inclusive in a way that is not disruptive to the Gitea development process.

I commented further on the issue you opened that is related to this topic.

[Gusted]

Just to mention this again, a valid point was raised on the previous PR for F3. I would like to state again that Loïc and the other contributors will change the license to MIT when this PR is approved to be merged.

[6543]

@earl-warren could you add me to your fork as contrib (write access)? so i can update that pull ... also if you would not force-push I would aprechiate!

[6543]

blocked by https://lab.forgefriends.org/friendlyforgeformat/gof3/-/issues/21

[6543]

please update branch & update f3 dep