Closed ptomato closed 1 month ago
cc @doehyunbaek @jhnaldo - any idea what is happening?
Thanks for letting me know the problem.
See actions/runner-images#10788 and actions/runner-images#10636
This issue seems the main reason for the problem.
Should we wait until this issue is resolved in actions
, or should we try to resolve this issue now?
What would probably be wisest is to add a step to https://github.com/tc39/test262/blob/main/.github/workflows/esmeta-test262.yml (and https://github.com/tc39/ecma262/blob/main/.github/workflows/esmeta-typecheck.yml) that installs sbt
, that's skipped if it's already present. That way, if github fixes it, the job will be faster, but it'll work regardless.
Additionally, it'd probably be good to create a single esmeta action that both of these workflows can use, so that there's only one place needed to change for stuff like this.
https://github.com/tc39/test262/pull/4270 should fix it. It doesn't skip installing SBT if already present, but I think it's pretty unlikely that GitHub would re-add it, given that people have been complaining for three weeks already.
Since a couple of days ago, the ESMeta CI job is failing because it cannot locate the
sbt
binary.