It failed its Travis tests, but only on mysql with PHP 7.1.
This is not likely to be due to the content of the PR, which has nothing to do with the DB and contains no PHP subtleties.
The failure pattern looks much like that create by the trivial CSS(!) change of @asmecher yesterday (based on stable-3_1_2 as well): Plenty of “broken pipe” in various tests.
What should I do?
The PR is not urgent, but I would like to get it done and off my memory.
The Travis tests occasionally hit some kind of race condition and fail; when we see this we can restart it and generally it’ll pass. (This is worse with the 3.1.2 branch than the master branch.) It’s annoying and difficult to debug and is on my list!
In other words, please don’t worry about it. If we see a few branches passing, the others usually just need to be kicked over.
Public Knowledge Project Team