[00:35:32] <Kronf>	 Hi
[00:35:41] <Kronf>	 Has anyone seen JeLuF recently?
[00:36:09] <Susan>	 Not on IRC, nope.
[00:36:15] <Susan>	 He's probably reachable via e-mail.
[00:36:52] <Kronf>	 Ok thanks. I tried it via the Meta mail-special page
[00:37:10] <Kronf>	 a few days ago
[00:37:24] <Susan>	 Is it urgent?
[00:37:35] <Susan>	 I imgaine a few people have Jens' other contact info...
[00:37:38] <Susan>	 imagine
[00:40:41] <Kronf>	 It concerns another project running on his server. It's not too urgent, but the problem is we haven't had contact to him for a few months
[00:45:51] <Kronf>	 Do you think there's someone I could ask for making contact?
[00:49:08] <DaBPunkt>	 gn8 folks
[03:26:42] <Riley|Cloud>	 Any users here fimilar with all the new account creation updates and such?
[04:45:35] <superm401>	 Riley|Cloud, I worked on Special:GettingStarted.
[04:45:36] <superm401>	 What's up?
[04:47:27] <Riley|Cloud>	 Just curious about which users see the old create an account page
[04:49:34] <superm401>	 Oh, the form (blue button and such)?  No one currently.  It was experimental at first, but it's currently everyone.
[04:50:57] <Riley|Cloud>	 Isn't it still being used?
[04:51:58] <Riley|Cloud>	 superm401: Wait. No one can currently see which one?
[04:52:39] <superm401>	 Riley|Cloud, no one sees the old one.
[04:52:51] <Riley|Cloud>	 Account creators do
[04:53:05] <Riley|Cloud>	 I know that for sure (/me is one)
[04:56:24] <superm401>	 Riley|Cloud, alright, I'll be conservative and say everyone who goes to the signup page logged out.
[04:56:31] <superm401>	 The relevant code is at https://gerrit.wikimedia.org/r/gitweb?p=mediawiki/extensions/E3Experiments.git;a=blob;f=experiments/acux/accountCreationUX.js;hb=HEAD
[04:56:52] <superm401>	 spagewmf wrote it, so he's the guy to ask for details.
[04:57:27] <Riley|Cloud>	 Nope..
[04:57:28] <spagewmf>	 what up folks?
[04:57:54] <spagewmf>	 If you create an account while logged out you get the new ACUX experience.
[04:57:56] <superm401>	 Riley|Cloud, you see it logged out?
[04:58:14] <Riley|Cloud>	 You don't get logged out when you go to the signup page
[04:58:22] <superm401>	 Riley|Cloud, I know that.
[04:58:29] <superm401>	 I said "everyone who goes to the signup page logged out"
[04:58:45] <superm401>	 I.E. you were logged out/have never ever logged in *before* going to the signup page.
[04:58:51] <Riley|Cloud>	 oh
[04:59:26] <Riley|Cloud>	 So all logged in users who go to signup page get the new view then?
[04:59:37] <Riley|Cloud>	 old*
[05:00:45] <superm401>	 Yes.
[05:00:48] <spagewmf>	 Riley, yes. There were enough variations of account creation (reason, e-mail an account to someone, extra details) that we punted on them for logged-in users.
[05:01:58] <superm401>	 For the record, the check is https://gerrit.wikimedia.org/r/gitweb?p=mediawiki/extensions/E3Experiments.git;a=blob;f=experiments/acux/accountCreationUX.js;hb=HEAD#l871
[05:03:01] <Riley|Cloud>	 oh okay thanks
[05:03:03] <superm401>	 spagewmf, although that's in a debug block.  Is there another check, or is it always in debug?
[05:03:11] <superm401>	 Riley|Cloud, no problem
[05:04:07] <Riley|Cloud>	 I was just originally wondering if account creators got a different view :)
[05:04:49] <superm401>	 I see, isUserInExperiment() is called even without DEBUG.
[05:05:00] <superm401>	 Yeah, they may eventually, but it's not as high a priority.
[05:05:14] <superm401>	 I do get that account creators are important (e.g. for people who can't read the CAPTCHA).
[05:05:24] <superm401>	 But they aren't as likely to get discouraged by a bland form.
[05:20:14] <Riley|Cloud>	 superm401: I would prefer an even blander form
[05:20:26] <superm401>	 Riley|Cloud, :)
[05:22:03] <Riley|Cloud>	 What would be neat for AC's is to remove the password fields and the blah about the emails "* You do not have to provide an e-mail address blah blah."
[09:16:06] <Joy>	 hi there, is someone working on the recent job queue explosion?
[09:16:48] <Nemo_bis>	 Joy: it was reported about ten hours ago on this channel, try and check the logs
[09:17:19] <Joy>	 Nemo_bis: all i see is your entry in the log, but no apparent action :) it's been going on for a while now
[09:17:36] <Nemo_bis>	 AaronSchulz> Reedy: meh, that's expected
[09:17:40] <Nemo_bis>	 AaronSchulz> it's ok as long as no problems are caused
[09:17:51] <Nemo_bis>	 I meant the logs of this channel
[09:18:05] <Joy>	 ah
[09:18:08] <Nemo_bis>	 I've no idea what Aaron meant, but in short the answer to you question is: No.
[09:19:26] <Nemo_bis>	 It's going down again now, apparently some weirdness in the new deduplication system: it needs to duplicate things first to then show off how good it is at deduplicating, I guess.
[09:19:30] <Nemo_bis>	 :p
[09:19:48] <Joy>	 sadly the rate it's going down is much more steady than the rate it occasionally goes up
[09:20:07] <Nemo_bis>	 I don't see how that would be sad
[09:20:27] <Joy>	 i've been waiting for a special:whatlinkshere update for a day and a half now, and it saddens me :)
[09:21:14] <Joy>	 it's annoying to have to go through hundreds of items manually to verify i didn't miss anything
[09:21:30] <Nemo_bis>	 you'd better wait 48 h
[09:22:10] <Joy>	 ?
[09:22:27] <Nemo_bis>	 http://ganglia.wikimedia.org/latest/graph_all_periods.php?c=Miscellaneous%20pmtpa&h=spence.wikimedia.org&v=506&m=enwiki_JobQueue_length
[09:22:46] <Nemo_bis>	 at this rate things should be back to normality in 56 h
[09:23:08] <Nemo_bis>	 or 44 h to get to pre-wmg6 situation (1 million)
[09:23:12] <Nemo_bis>	 *wmf6
[09:23:34] <Joy>	 ah... well, the graph doesn't carry a promise that what happened yesterday won't repeat, so i wanted to check :)
[09:24:33] <Joy>	 ta
[10:22:24] <Nemo_bis>	 andre__: are you sure benny can/will work on https://bugzilla.wikimedia.org/show_bug.cgi?id=42752 ?
[10:22:32] <Reedy>	 Nemo_bis: It's dropped by around 100,000 in an our or so
[10:22:39] <andre__>	 Nemo_bis, not totally, no
[10:22:49] <Nemo_bis>	 Reedy: yes, going about 350k every 4 h AFAICS
[10:22:58] <Nemo_bis>	 andre__: he asked help to Aaron
[10:23:08] <Nemo_bis>	 anyway that bug will soon be obsoleted
[10:23:20] <andre__>	 I hope Benny will reassign it to somebody if it's impossible for him in the next days :)
[10:23:29] <Nemo_bis>	 tomorrow wmf6 will be deployed
[10:23:36] <Nemo_bis>	 too late to fix it, damage done
[10:32:45] <legoktm>	 Reedy: do you know about https://bugzilla.wikimedia.org/show_bug.cgi?id=36780#c9 ?
[10:33:09] <legoktm>	 I haven't heard anything about it since I commented...
[11:20:01] <legoktm>	 Nemo_bis: just so i dont have to check the history of a bunch of bugs, you're just adding see also links right?
[11:23:35] <Nemo_bis>	 legoktm: where?
[11:23:49] <Nemo_bis>	 and see also links are really meant to be "see also"
[11:23:52] <legoktm>	 to all the email bugs
[11:24:08] <Nemo_bis>	 no, there's a component for that, I'm listing only relevant bugs of course
[11:24:52] <legoktm>	 gotcha
[11:24:54] <Nemo_bis>	 every now and than some user from en.wiki comes uo with yet another proposal of a new system for blocking emails, so the situation is confusing
[11:25:02] <Nemo_bis>	 *comes up
[11:25:22] <legoktm>	 yeah, i didn't realize how many bugs there were about email
[15:20:52] <Seddon>	 what on earth happened to the job queue yesterday
[15:36:28] <apergos>	 is it still a few million or whatever?
[15:40:44] <Nemo_bis>	 4,6 millions and no longer decreasing as quickly as this morning
[15:42:41] <apergos>	 do we know what was making it drop?  or increase even?
[15:44:00] <Nemo_bis>	 Aaron only said that it's expected
[15:44:25] <Platonides>	 if many of those were repeated there would be quickly drops, as repeated ones got removed
[15:44:46] <Nemo_bis>	 "as long as no problems are caused" – which I've no idea what could mean, given that usually a high job queue is considered a problem in itself
[15:56:06] <MaxSem>	 is there a InitializeSettings group that contains precisely the same wikis as wikipedia.dblist? cause 'wiki' also contains stuff like Commons and Meta
[16:16:49] <apergos>	 I didn't even know there was a wikipedia.dblist,, never used it
[16:18:50] <siebrand>	 identi.ca to be phased out eventually. http://status.net/2012/12/18/upcoming-changes-in-the-status-net-service
[16:23:20] <Platonides>	 it's confusing what will they be migrating to
[16:23:29] <Platonides>	 "thanks to Brion Vibber for making the whole multi-homing daemon system actually work (kind of) efficiently" :)
[16:24:19] <apergos>	 maybe you should be looking for a group that contains the stuff in special.dblist and use that to exclude form the larger list
[16:24:38] <apergos>	 oh noes, (about identi.ca)
[16:25:05] <Platonides>	 it's some kind of migration
[16:25:17] <Platonides>	 although I don't know why it's said that way
[16:25:35] <apergos>	 hm
[16:25:48] <apergos>	 well we'll see, he didn't even give a time frame for that bit so hopefully it's a ways off
[16:26:53] <Platonides>	 in case it would just be "we will be doing an account registration freeze before deploying the new version (and dropping unused accounts)"
[16:31:08] <Nemo_bis>	 Platonides: do you understand what pump.io actually is?
[16:31:46] <Platonides>	 I saw http://pump.io/, but it looks more like a streaming protocol
[16:32:02] <Platonides>	 (with a server implementation)
[16:32:17] <Nemo_bis>	 He says identi.ca will be transitioned when everything will be able to be migrated to pump.io accounts and so
[16:33:55] <Platonides>	 then why the pesimistic tone?
[16:45:04] <Nemo_bis>	 Platonides: whose pessimistic tone?
[16:47:09] <bianjiang>	 did anyone know this bug https://bugzilla.wikimedia.org/show_bug.cgi?id=43221 ?
[16:49:45] <bianjiang>	 whoami
[17:00:37] <hoo>	 bianjiang: Taking a look
[17:01:32] <bianjiang>	 hoo: Thanks
[17:01:40] <hoo>	 Does this always happen with new renameuser logs?
[17:03:04] <bianjiang>	 not sure what kind of rc will trigger that...
[17:03:28] <bianjiang>	 it looks similar to an old bug: https://bugzilla.wikimedia.org/show_bug.cgi?id=40299
[17:03:35] <hoo>	 bianjiang: In that case it's a renameuser log ... and the renameuser log has recently been altered, so this could be the caus
[17:03:35] <hoo>	 e
[17:03:38] * hoo  clicks

[17:07:34] <bianjiang>	 hoo: i'm guessing some log lib is not up-to-date, https://www.mediawiki.org/wiki/Logging_to_Special:Log
[17:07:54] <hoo>	 bianjiang: Nah, the cause is much more trivial, hang on a second
[17:09:16] <hoo>	 ok, I'm wrong
[17:13:53] <hoo>	 I've just installed the extension to my test wiki and going to (try) to reproduce this
[17:14:47] <hoo>	 yep, works out :)
[17:16:01] <bianjiang>	 cool!
[17:16:57] <hoo>	 Not my patch is wokring, but I could reproduce it :P
[17:36:08] <bianjiang>	 hoo: sorry for pushing, do you think the bug is easy to fix or not?
[17:36:42] <hoo>	 bianjiang: I'm trying atm, but I'm not sure whether core or the extension is to blame
[17:39:38] <bianjiang>	 hoo: i c.  just saw the bug is updated (you did that?)
[17:40:23] <hoo>	 nah
[18:41:31] <Nemo_bis>	 str4nd: so what about your gentle pseudo-squatter?
[20:27:28] <superm401>	 ^demon, can you setup a repo for Guiders (used by GuidedTour)? https://www.mediawiki.org/wiki/Git/New_repositories/Requests/Entries
[20:27:31] <superm401>	 Thanks.
[20:28:13] <^demon>	 Ok, we need to bikeshed over the name.
[20:28:24] <^demon>	 integration/* is for CI-stuff (eg: jenkins, zuul, testswarm)
[20:30:13] <^demon>	 superm401: If it's a separate library for the extension, we could make it like mediawiki/extensions/GuidedTour/guiders (it won't conflict or anything).
[20:30:19] <^demon>	 We do the same thing with Parsoid for some of its libraries.
[20:30:36] <superm401>	 That's fine, as long as it's a separate repo.
[20:30:53] <^demon>	 Yeah. It'll inherit the permissions, which will make it easy to manage.
[20:31:26] <^demon>	 Ok, I'll do this now.
[20:32:45] <superm401>	 Cool, thanks.
[20:32:57] <^demon>	 Done: https://gerrit.wikimedia.org/r/#/admin/projects/mediawiki/extensions/GuidedTour/guiders
[20:33:37] <superm401>	 Wow, thanks.
[20:34:34] <^demon>	 You're welcome. Repo creation is quick & easy. The thing that usually takes me longer is setting up acls.
[20:34:37] <^demon>	 That's more manual.
[23:39:52] <AaronSchulz>	 CategoryTree::renderParents     10.0.6.49       1054    Unknown column 'page_namespace' in 'field list' (10.0.6.49)
[23:39:55] <AaronSchulz>	 ...