[08:21:58] <zpapierski>	 relocating
[09:20:49] <dcausse>	 time to break the k8s cluster
[09:20:58] <zpapierski>	 yay :)
[09:21:05] <zpapierski>	 I mean, we shouldn't probably break it :)
[09:21:21] <zpapierski>	 but it's staging, they'll forgive is if we do
[09:21:53] <dcausse>	 AFAIK it'll be the first time it's infected by a java app :)
[09:22:16] <zpapierski>	 "infected" - great choice of words :)
[09:30:38] <dcausse>	 ok it did not went too far: err: no releases found that matches specified selector() and environment(staging-eqiad), in any helmfile
[09:36:13] <zpapierski>	 huh, I don't remember an enviroment of that name
[09:36:29] <dcausse>	 yes it's me, should be simply "staging"
[09:55:51] <zpapierski>	 where are mediawiki logs on cindy?
[10:01:06] <dcausse>	 zpapierski: /src/mediawiki-vagrant/logs/
[10:01:13] <zpapierski>	 thx
[10:01:19] <dcausse>	 since there are multiple wikis you need to pick the right one
[10:01:30] <zpapierski>	 that I remember
[10:01:32] <dcausse>	 generally cirrustestwiki is the one used for the tests
[10:05:46] <zpapierski>	 anyway - how's your testing?
[10:09:18] <dcausse>	 got a small error that was fixed by Janis and now I'm waiting for the deploy command to return :)
[10:18:55] <dcausse>	 failing to contact swift now
[10:21:47] <zpapierski>	 can't contact as in rejected? or can't authorize?
[10:25:52] <dcausse>	 connection refused, it going through envoy so there are quite some moving pieces that could fail
[10:26:10] <dcausse>	 will try again after lunch
[10:34:39] <zpapierski>	 maybe incorrect port is provided on that environment
[10:34:56] <zpapierski>	 or proxy might not even be set up for staging
[10:35:14] <zpapierski>	 ah, I still get cindy failures, even though title completion works :)
[10:35:30] <zpapierski>	 s/ :)/ :(
[12:29:34] <dcausse>	 zpapierski: feel free to ship a noop patch maybe it's just master that's broken
[13:46:07] <mpham>	 does T284721 have any affect on users? I'm having a hard time following what/who is affected by this issue. I also was confused because it mentions OpenSearch and PrefixSearch, and I didn't think we were using the former anywhere
[13:46:11] <stashbot>	 T284721: PrefixSearch/OpenSearch API behaves inconsistent depending on capitalization - https://phabricator.wikimedia.org/T284721
[13:54:09] <zpapierski>	 it is confusing - OpenSearch is a technology (ironically, also created by Amazon) that handles search bars auto complete
[13:54:32] <zpapierski>	 why they decided to call their fork of ES that is beyond me
[13:55:09] <zpapierski>	 in any case - it this situation, both are using the same backend, so it's basically about prefixsearch
[13:55:39] <zpapierski>	 this isn't really a bug - they don't like how the documents are scored now
[13:56:02] <zpapierski>	 I can be wrong (dcausse ?) but a profile dedicated for them could solve their issue
[13:56:33] <zpapierski>	 in any case - they are not actively pursuing this now, but they asked us to explain the difference - which I think we can do
[13:58:28] <zpapierski>	 I'll answer their question (though probably not today)
[13:58:37] <dcausse>	 we could fix the case mismatch by tweaking this particular query, not sure that the old prefixsearch is tunable with profiles like other queries
[13:59:56] <zpapierski>	 right
[14:00:24] <zpapierski>	 I'll verify that, too (since I finally understand profiles, unless I already forgot)
[14:00:51] <dcausse>	 as to who are affected by this: mostly expert users that uses shortcut: accessing Village Pump by typing WP:VP, but I think generally they press enter to enter the page
[14:01:35] <dcausse>	 but here since these are templates selected from the visual editor they must select it from the list of suggestions so perhaps a bit annoying
[14:02:13] <zpapierski>	 yeah - the second use case is what matters to them
[14:02:17] <zpapierski>	 (I assume)
[14:04:00] <zpapierski>	 here's dcausse 's comment from CR that adds some clarity, I think - https://gerrit.wikimedia.org/r/c/mediawiki/core/+/700172/1#message-f2928be5a29a7635a84c4e4cbdbcb7726006d840
[14:10:11] <zpapierski>	 relocating
[15:00:11] <mpham>	 thanks for the explanation. the OpenSearch name is really confusing. jeez. Anyway, doesn't seem like a super high priority issue to me since it seems to be at most a little annoying to just expert users 
[15:41:51] <zpapierski>	 Thiemo's team is working off of technical wishes list - I'm guessing idea here came from that list
[15:42:03] <zpapierski>	 but it speaks nothing about priority itself, of course
[16:18:57] <volans>	 ryankemper: FYI I've merged the patch for the search cookbooks gerrit/702884, I've tested a couple of them in dry-run mode and all seemed ok.
[16:19:10] <volans>	 but lmk if there is any issue when you use tehm for real
[17:07:22] <dcausse>	 dinner