Cristian Le

I can also see this fitting in the `discover` definition. I guess there would be some ambiguity when defining multiple deiscover steps. For reference, `ctest` uses language like `ctest --no-tests={ignore,error}`

> > I can also see this fitting in the `discover` definition. I guess there would be some ambiguity when defining multiple deiscover steps. > > Interesting thought, storing the...

Looking at it, seems doable. I guess there aren't unit tests around it to hook to? I'll be incapacitated due to a cold probably until Monday though

Initial unit tests added. Right now commit checkouts and detached merged commits behave the same, haven't checked how to mimic Github's `refs/remotes/pull/N/merge`. Also I haven't covered `fmf_id.url` and other `remote`...

To try to summarize: - Original `git rev-parse --abbrev-ref --symbolic-full-name @{u}` did not work on detached heads for tags and detached merges (github's action/checkout) - Instead of that, it is...

One quick note, I am working on moving this git ref resolution logic to `tmt.utils.git` and making it cached according to `tmt_root`. Edit: aand done.

Ah, no, I had an error there: ``` :: [ 12:13:25 ] :: [ BEGIN ] :: Running 'expected_ref=$(git describe --tags eae4d527)' fatal: No names found, cannot describe anything. ::...

> @LecrisUT, would mind adding a short release note about the change? I'll try to think of some words tomorrow. I guess the part that should be mentioned is when...

A bit unfortunate that PyCharm does not support `#:` docstrings [^1], but that's a topic for a different PR. @psss what do you think of the wording for the release...

The main difference is that with the draft releases is that it is only visible to GH contributors, while the PR is more interactive. `changesets/action` itself, I don't think it's...