[00:01:50] <James_F>	 legoktm: Need me to go upgrade a bunch of C+1s to +2s now? :-)
[00:07:16] <legoktm>	 James_F: oops, I just did :p
[00:07:34] <James_F>	 Ha. :-)
[00:07:35] <legoktm>	 I'm writing up the MWTidy::checkErrors() email right now
[00:07:55] <James_F>	 Nice.
[00:08:11] <legoktm>	 hmm
[00:08:14] <James_F>	 Also K.rinkle merged your PlatformSettings commit.
[00:08:15] <legoktm>	 Thanks one is failing
[00:08:19] <James_F>	 Eurgh.
[00:08:26] <legoktm>	 00:06:21 RecursiveDirectoryIterator::__construct(./../../extensions/Flow): failed to open dir: No such file or directory
[00:08:26] <legoktm>	 00:06:21 RecursiveDirectoryIterator::__construct(./../../extensions/MobileFrontend): failed to open dir: No such file or directory
[00:08:36] <James_F>	 What lovely code.
[00:08:45] <legoktm>	 it's just missing the dependencies
[00:08:57] <James_F>	 So does Thanks not have those dependencies defined?
[00:35:16] <James_F>	 legoktm: Also you should reply to a.nomie on https://gerrit.wikimedia.org/r/#/c/424973/ so he can merge or you can update the patch. :-)
[00:38:33] <legoktm>	 James_F: https://etherpad.wikimedia.org/p/mwtidy-checkerrors
[00:40:48] <James_F>	 legoktm: LGTM.
[00:44:08] <legoktm>	 sent
[00:45:01] <James_F>	 Thanks.
[00:47:30] <James_F>	 legoktm: So... Looks like PhpTagsFunctions doesn't work with PHP7; I've made https://gerrit.wikimedia.org/r/#/c/425449/ which fixes the parser test failure (so making it pass in PHP5 and HHVM), so if you merge that then we can add codesniffer to the repo and later fix those (if we care)?
[00:52:07] <legoktm>	 force merge then?
[00:53:16] <James_F>	 legoktm: We don't test the repo against PHP7 yet, just HHVM, so it's "fine".
[00:53:25] <legoktm>	 I think php7 will run in gate
[00:53:33] <James_F>	 legoktm: But you need to merge https://gerrit.wikimedia.org/r/#/c/425449/ first. :-)
[00:53:39] <legoktm>	 we'll see :p
[00:53:40] <James_F>	 Oh. Hmm.
[00:54:05] <James_F>	 Ah, you did. Thanks.
[00:59:07] <James_F>	 Meh. You're right.
[00:59:10] <James_F>	 OK, force merge time it is.
[01:58:55] <legoktm>	 I tried to test installing OATHAuth with the web installer and it's just flat out broken
[01:59:07] <legoktm>	 Does no one actually use the web installer??
[02:00:50] <legoktm>	 (Also I'm most likely the person who broke it? >.<)
[02:19:18] <paladox>	 I use the web installer
[02:19:38] <paladox>	 Also the web installer seems broken when proc_open is disabled
[02:19:41] <paladox>	 legoktm: ^^
[02:20:12] <legoktm>	 Well that's unrelated to my breakage :p
[02:23:55] <legoktm>	 paladox: have you filed a bug about the proc_open breakage?
[02:24:12] <paladox>	 Nope not yet
[02:27:17] <paladox>	 legoktm done https://phabricator.wikimedia.org/T191947
[02:28:12] <MaxSem>	 Krinkle: is there a Python port of Intuition?
[02:28:34] <Krinkle>	 MaxSem: there is not
[02:29:03] <MaxSem>	 whyyyyyy :P
[02:29:07] <Krinkle>	 https://github.com/Krinkle/intuition/issues/18
[02:29:19] <Krinkle>	 If there is a use case and resourcing, I'd be happy to help formulate a plan.
[02:30:34] <Krinkle>	 Depending on which features you need, it will be easy or hard. E.g. if you need only the messsages, a port of the js-env would suffice. If you need plural/gender, then it's gonna be harder, as it will require a shift from using PHP files from MW to generating code based on CLDR/JSON data, once that is done, it'd be easier to generate both PHP and Python interfaces.
[02:31:05] <legoktm>	 MaxSem: you could look at pywikibot.i18n, which TWN already supports
[02:31:18] <legoktm>	 not sure how easy it is to use it outside of pwb
[02:31:50] <MaxSem>	 my use case is bot. but since it only needs to post wikitext, a very simple KV store would suffice
[02:32:11] <legoktm>	 pywikibot!
[02:32:42] <MaxSem>	 wheee
[02:32:44] <MaxSem>	 thanks
[03:17:02] <legoktm>	 https://gerrit.wikimedia.org/r/#/c/425454/ - Unbreak creating extension tables from the web installer
[03:37:06] <no_justification>	 I hate the web installer
[03:37:10] <no_justification>	 I wish we only had a CLI installer
[03:42:26] <legoktm>	 > is_bool( $wgCategoryTreeDisableCache ) && $wgCategoryTreeDisableCache === true
[03:42:33] <legoktm>	 checks is_bool and strict equality??
[03:57:45] <Krinkle>	 Just to be safe, right?
[06:46:58] <legoktm>	 thanks Krinkle
[07:00:17] <legoktm>	 Krinkle: do you know what the point of https://github.com/oyejorge/less.php/blob/master/lib/Less/Less.php.combine is?
[17:36:57] <Krinkle>	 legoktm: btw, status on https://phabricator.wikimedia.org/T179055 ? (RemexHtml job) - assuming it's been on php7 for a while now, I suppose we can just commit that live change, right?
[17:38:08] <Krinkle>	 Regarding less.combine, I would imagine it is/was used for some ancient hack that encourages single-file php distribution, e.g. from the time people would download a php.zip file from a download.html page and add/include as a single file.
[17:38:19] <Krinkle>	 Possibly minified because... it feels faster.
[17:38:31] <Krinkle>	 I've had to deal with minified PHP code during my internship.
[17:38:34] <Krinkle>	 Highly furstrating.
[17:41:24] <Krinkle>	 legoktm: Woops, looks like they still use it for tarball releases indeed. One of the three install methods in the README is to download zip from https://github.com/oyejorge/less.php/releases and include [path]/Less.php which is a file that only exists as result of concatenating the other files based on that .combine file.
[17:45:35] <AryehGregor>	 anomie: Don't additional tests for different usergroup adds/removes belong as tests for SpecialUserrights or such, not the API module?  It doesn't do any of the work of computing which groups to add/remove etc.
[17:46:32] <anomie>	 AryehGregor: Meh. It doesn't hurt to check that logic again as an integration test, and it does verify that the 'remove' parameter is getting passed to the backend properly.
[18:41:04] <James_F>	 legoktm: Question from me to you on https://phabricator.wikimedia.org/T182020#4120341
[19:58:36] <James_F>	 legoktm: Did you want to land https://gerrit.wikimedia.org/r/#/c/313689/ for MW1.31? If so, who's going to merge it?
[19:59:20] <legoktm>	 I don't think there's any urgency on that, but I haven't touched it since 2016
[19:59:49] <James_F>	 Fair point. :-)
[20:02:35] * bd808 promises cookies for folks like legoktm who are still working on library extraction
[20:03:27] <legoktm>	 :) it's a bit on the back burner right now :(
[22:00:22] <Krinkle>	 RoanKattouw: with T172477, we could enhance todays proposal to have the client crypto token be a nested hmac of the requestor's IP as well.
[22:00:22] <stashbot>	 T172477: Attribute anonymous contributions to the first IP address used in a session - https://phabricator.wikimedia.org/T172477
[22:00:39] <Krinkle>	 And thus still attribute it the first IP, even without a session.
[22:01:00] <Krinkle>	 Of course, this is just for the edge case of total compromise of attackers being able to set cookies on our domains.
[22:02:40] <Platonides>	 too… many… channels…
[22:03:38] <Krinkle>	 my join-channel count is at the lowest it's been in over 2 years.
[22:03:50] <Krinkle>	 which is... 19
[22:04:01] * Platonides currently has 105 windows
[22:05:39] <paladox>	 lol
[22:06:14] <Platonides>	 maybe a few are pms, but still