[08:41:52] <dcausse>	 just realized that 'trigger' in CirrusSearchUserTesting is not a thing... somehow got confused quite a bit not being able to activate a bucket using it (I made trigger != bucket name)
[08:55:22] <pfischer>	 o/ dcausse: Just looked at the docs… What are you testing/comparing?
[08:55:51] <dcausse>	 pfischer: it's related to T404858
[08:55:51] <stashbot>	 T404858: A/B test using defaultsort with the completion suggester - https://phabricator.wikimedia.org/T404858
[08:58:27] <dcausse>	 mainly helping users searching for lastname first and also skip some popular prefixes like 'List of', e.g. typing 'trigonometric iden' would suggest 'List of trigonometric identities'
[09:09:29] <pfischer>	 Thanks! Should that be covered by the auto-assigned buckets (index.php case described in the docs)? Or do you need `wgCirrusSearchActiveUserTest` as a trigger to instruct cirrus to consider the A/B test at all?
[09:16:20] <dcausse>	 pfischer: yes, it works well in the end but was trying to trigger a bucket manually passing the URI param cirrusUserTesting=$testName:$bucketName, this is where I got confused because I thought that 'trigger' in https://gerrit.wikimedia.org/r/c/operations/mediawiki-config/+/1198295/2/wmf-config/ext-CirrusSearch.php#664 was still a thing
[09:17:00] <dcausse>	 so was expecting T404858_default_sort:dsort to do what I want, turns out you need T404858_default_sort:default_sort
[09:23:43] <dcausse>	 I think this "trigger" thing got removed in https://gerrit.wikimedia.org/r/c/mediawiki/extensions/CirrusSearch/+/685621 
[10:46:37] <dcausse>	 lunch
[13:23:16] <inflatador>	 <o/
[14:01:24] <dcausse>	 o/
[14:18:32] <ebernhardson>	 \o
[14:19:37] <inflatador>	 .o/
[14:20:34] <dcausse>	 o/
[14:31:16] <gehel>	 Trey314159: I'll be 5' late
[14:34:27] <Trey314159>	 ok
[15:02:42] <inflatador>	 workout, back in ~40
[16:02:08] <pfischer>	 gehel: are you around for triage?
[17:27:08] <dcausse>	 dinner
[18:27:18] * ebernhardson wonders why we had truncate_keyword on near_match_asciifolding, but not near_match.
[18:30:45] <ebernhardson>	 in the patch that added the truncation it was applied to keyword, lowercase_keyword, and near_match_asciifolding. i guess i simply forgot near_match
[18:32:57] <ebernhardson>	 btw david was right, T408399 was a 1 :)
[18:32:57] <stashbot>	 T408399: Truncate labels.*.near_match fields - https://phabricator.wikimedia.org/T408399
[19:00:07] <inflatador>	 been back, but now time for the doctor!