Make sure scripted jobs use the same commit like publish-local did

[mkurz]

Please see the comment I added in the test-scripted file.

Now that scala-steward runs monthly, it opens many pull requests at once, in parallel, increasing the chance that these pull requests fail, because of the problem I described in my comment. The problem existed before as well and pull requests failed every now and then, but I never really looked into it what the problem was, until now, when I realized what's going on.

Let's look at one of those failed pull requests: https://travis-ci.com/github/playframework/playframework/builds/225836519:
As you can see, the publish-local jobs used the version 2.8.1+1156-e905d7a5, so the distance from the 2.8.0 tag is 1156 and the current commit is e905d7a (being a merge commit GitHub provides as if the pull request already had been merged, which in reality didn't happen course). Now after that publish local but before the (failed) scripted job started one pull request (#10847) got merged. So based on the information in my comment, the master branch moved forward, and now the scripted jobs use a different version: 2.8.1+1158-5a053e91. Distance got increased by 2 (1 commit for the version update + 1 merge commit) and of course the hash changed.

You can see the difference between the two master branches here:

• At the time the publish-local job started: https://github.com/playframework/playframework/commits/e905d7a5
• At the time the scripted job started: https://github.com/playframework/playframework/commits/5a053e91

[mkurz]

Oh yes and this pr fixes #10300, #10244 and #10108

[mkurz]

@Mergifyio backport 2.8.x

[mergify]

Command backport 2.8.x: pending

Waiting for the pull request to get merged

[octonato]

Clever!

And thanks for the whole explanation. Makes lot of sense.

[mergify]

Command backport 2.8.x: success

Backports have been created

• #10872 Make sure scripted jobs use the same commit like publish-local did (backport #10855) has been created for branch 2.8.x