Andreas Abel

Historically, I think AppVeyor was the tool of choice for a Windows CI due to limited availability of alternatives. A GHA CI could be similar to this: https://github.com/haskell/text-icu/blob/52822ec7684e39d3ee5afcd0d807f902971c0761/.github/workflows/cabal-mac-win.yml

So what is the proposed design? Spinning some ideas: 1. Maybe a field: ``` x-hackage-run-tests: False ``` There could be other `x-hackage` fields like this: `-build`, `-build-tests` to opt out...

> I would prefer we just add metadata state to packages which maintainers can set from the "maintainers corner". How does this metadata state work? Where/how is it stored and...

Maybe a duplicate of #768?

But I ain't working yet, the `matrix.hackage.org`. At least it does not display anything sensible... Note how it blends the package I visited first (`Agda`) with the once I am...

Well, in the long run, we are likely going this way. But maybe we should hold back a little until `matrix.hackage` is working reasonably.

Thanks for the PR! Is _Maintainers_ identical to _package maintainers_? - If yes, the same terminology should be used and the link moved from the current _package maintainers_ to the...

All I meant is: ``` MAINTAINERS (linking to the hackage maintainers pages): Alice, Bob, Carlo For MAINTAINERS and hackage trustees * edit package information ``` Here, replace MAINTAINERS by either...

Me too. Revision to `base >= 4.10` here: https://hackage.haskell.org/package/basement-0.0.13/revisions/ Ideally, one would have revised the `if impl(ghc < 8.0) buildable: False` to `...8.2...`, but Hackage doesn't allow revision of `if`-conditions.

@vincenthz : This issue is fixed now and can be closed.