Tails

#2 Updated by bertagaz 2015-10-25 06:27:49

• Assignee changed from bertagaz to intrigeri
• % Done changed from 0 to 80
• QA Check set to Ready for QA

intrigeri wrote:
> Bumping priority: we need these tests to run so that we have enough data for our meeting on the 28th, you know the drill :)

Seen that yesterday and wanted to fix it today. I’ve removed the remaining workspaces waiting for the real fix so that the automated tests can keep on running.

I’ve pushed 2c5b033 in the jenkins-jobs Git repo. It configures the workspace-cleanup plugin the same way we did before, I don’t know why we moved to this buggy configuration.

It seems to work: meanwhile the test_Tails_ISO_test-10376-fix-startup-page-roadmap-test-is-fragile #9 has been run on isotester4, and there’s no remaining files in the job workspace.

#3 Updated by intrigeri 2015-10-27 13:29:49

• Status changed from Confirmed to Resolved
• % Done changed from 80 to 100
• QA Check changed from Ready for QA to Pass

(Lecture warning.)

> I’ve pushed 2c5b033 in the jenkins-jobs Git repo. It configures the workspace-cleanup plugin the same way we did before, I don’t know why we moved to this buggy configuration.

Thankfully we have Git. git log -p --reverse points me to:

commit 827001183b7bb2c061496a585a7b9f148a56dd79
Author: bertagaz <bertagaz@ptitcanardnoir.org>
Date:   Wed Jun 10 20:37:49 2015 +0200

    Fix the workspace cleanup config: is a publisher.

As you see, the commit message isn’t terribly helpful. I think you referred to this change in Feature #9350#note-9, which doesn’t include more details. And I suspect that the commit message for the fix you’re proposing (“Fix the workspace-cleanup publisher configuration (hopefully)”) won’t be very helpful either, next time we have to dig through the Git history.

I don’t feel I’m in a position to make an informed decision so, well… here I’ll trust your testing, cross fingers, decide that this revert is probably right, and close this ticket.

Please put a bit more time into your commit messages in the future, so we can actually understand changes N months later, and revert them if needed without having to do the same leap of faith again. Referencing, in commit messages, the ticket it is about would be a pretty good start (so next time we look into 2c5b033, we have a chance to understand what was broken), as a compromise, when you don’t feel like describing the actual problem your commit fixes. Thanks in advance!

And of course: thanks for working promptly on this problem!

#4 Updated by intrigeri 2015-10-27 14:21:29

• Assignee deleted (intrigeri)