You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
This repository was archived by the owner on Nov 6, 2023. It is now read-only.
Repository navigation
This repository was archived by the owner on Nov 6, 2023. It is now read-only.
There are multiple regressions in our testing process:
This commit contains duped targets, but Travis passed: https://travis-ci.org/EFForg/https-everywhere/builds/154001093
Here were also two identical hosts merged, but I'm not sure if Travis was able to detect that, beacause both lived in a separate PR: #6489
This commit contains a test url that doesn't match any target or exclusion: https://travis-ci.org/EFForg/https-everywhere/builds/154047712
On #6426 we have the problem again, that a slow host breaks the fetch test. Referring to the log, the fetch was tried only once.
I think we had all these tests before, so this is likely a regression of the last few changes that were made. I already had a look at the recent changes, but the test code itself didn't change much. Maybe some tests are now running in a wrong context or not at all?
cc @Hainish@cooperq
Activity
Hainish commentedon Aug 22, 2016
I see this and will investigate today
Hainish commentedon Aug 22, 2016
Separating these into separate issues
Hainish commentedon Aug 23, 2016
The fetch is run twice, it just isn't printed twice. It only prints when it is finished comparing, even if it takes two tries.
You can verify this by applying the diff in the attached patch and then running the fetch test.
diff.patch.txt
Hainish commentedon Aug 23, 2016
I don't think we can do anything about two PRs conflicting with each other, but neither conflicting with
master.