Codewise this looks fine. But I am uncertain about the test. Can we wait with merging this until the reva part is merged and add a reva bump to this PR?

Update: Ah, it doesn't even build without the reva bump :-)

Ahh that makes sense. I converted the PR to a draft for now.

How's your reva release schedule these days? I guess it's not urgent now that v2.30.0 has just been released and this won't be part of the upcoming release anyway. So I'll just wait for the next reva release.

How's your reva release schedule these days?

You can even bump to the latest commit in main. No need to wait for a release.

You can even bump to the latest commit in main. No need to wait for a release.

Could you please give me a hint on the steps necessary here? I remember appending the commit hash to reva in the go.mod file, although there was some quirk with it being two hashes or something?

go get github.com/opencloud-eu/reva/v2@main
go mod tidy
go mod vendor

should do the trick. Though I just learned that the next reva release is just around the corner. So maybe we should just wait for that. Sorry for the confusion.

go get github.com/opencloud-eu/reva/v2@main
go mod tidy
go mod vendor

should do the trick. Though I just learned that the next reva release is just around the corner. So maybe we should just wait for that. Sorry for the confusion.

Just tried locally, that does the trick - thanks. I agree through, let's wait for the release, then I'll rebase (or bump as second commit of not done already).

@JammingBen I guess we can move forward with this now?

@JammingBen I guess we can move forward with this now?

Yeah I'm trying to find out why https://ci.opencloud.eu/repos/3/pipeline/1078/197 fails now. It checks if edition is empty and fails if so, so it actually makes sense that it fails. But how was this working before? Or did the test not run before?

Yeah I'm trying to find out why https://ci.opencloud.eu/repos/3/pipeline/1078/197 fails now. It checks if edition is empty and fails if so, so it actually makes sense that it fails. But how was this working before? Or did the test not run before?

Before 08bba95 the edition was never empty (AFAIK this was done before all tests where actually running). Then tests where enabled which caused the failure you're just seeing. Then I added the commit you're currently reverting.

I guess we need to adapt the test expectations now. (Or remove tests which no longer make sense).