Sounds good to me.
Sun, Jun 17
Thanks @MarcoAurelio !
Sat, Jun 16
Indeed. It is a purely textual convention with no run-time impact on any of the tooling involved. Within the objectcache and rdbms libs, @aaron and I have been (loosely) following the convention of using [optional] Description. Default: "x".
@Dispenser Please specify the wiki+page and/or full url at which you can consistently see this, and whether it applies to logged-in, logged-out, or both. If logged-in only, does this require specific user preferences to be set?
Fri, Jun 15
@Alexia Reminder about the CR at https://gerrit.wikimedia.org/r/#/c/mediawiki/core/+/425439/3/includes/OutputPage.php
Measured the overhead of xhprof's sample mode.
@Jdlrobson It does indeed appear to be a bug somewhere relating to cache pollution, in particular it seems a cache key somewhere is being shared between mobile and desktop that shouldn't be.
Is this resolved now?
Closing report of exception that has been inactive for several months. Feel free to re-open once you have more information. Alternatively, you can also look for support on https://discourse-mediawiki.wmflabs.org/.
Ah, I assumed it was a base class test that subclasses extend and override a utility method controlling the subject class. But I see it's actually all in a single test. Yeah, that's a tough one.
Would it work to add a @covers MyClassName to the doc block for the test case subclass?
Confirmed. Thanks :)
I believe we either need to write a PHPCS sniff that is able to fully understand even edge-cases like the one above