[02:51:19] marktraceur: hi [02:59:24] drecodeam: Hey! [03:01:26] some of the whitespace changes which I had removed came back in the new change which kaldari had compiled [03:01:43] anyways I will fix them, but I had to ask you a few things [03:02:35] marktraceur: If there are only whitespace changes in a file, then what should I do ? [03:21:30] marktraceur: if you are busy with something, we can after some time ?? [03:23:46] drecodeam: My instinct is to revert it [03:23:57] Not include the file in that commit [03:24:24] marktraceur: then when should I include those files, when I have other changes in the file aswel ? [03:24:28] Either fix the whitespace troubles later, next time you touch the file, or make a commit for whitespace changes [03:24:46] Ok, I ll keep that in mind from now on. [03:24:52] (my opinion, I'm not sure there's a convention) [03:25:30] marktraceur: Ya, thats the general practice followed in other places also, I could see that when I googled for this :P [03:27:14] Another thing, in this change ID, the most recent one in flickr branch - https://gerrit.wikimedia.org/r/#/c/12269/ . [03:27:57] in mw.FlickrChecker.js (113), you have asked me to replace 'JPEG' to mime - type [03:28:20] could you explain that a little more ? [03:32:58] drecodeam: If you could explore it, only [03:33:04] It depends on what the flickr api accepts [03:33:48] marktraceur: for all sizes other than the original file, it only gives JPEG image format [03:34:23] for pro users when we have a permission to use the original file, I need to add changing that format to the one that Flickr provides [03:44:39] !performance [03:44:39] [03:47:53] drecodeam: OK, then I rescind that comment [03:49:44] marktraceur: thanks a lot, sorry if some of my questions are pretty naive. I don't have much idea about a lot of things [03:52:43] Don't worry, drecodeam: we all have areas of expertise and ignorance. I was pretty shocked by a display of non-knowledge earlier today by a team of people, including one describing xyrself as a methodologist. [03:54:25] Amgine: hehe. thanks [05:16:03] Amgine: That sounds like a good story! [05:16:05] :) [05:16:24] It's logged. [05:16:52] Oh, then it's in my scrollback probably [05:16:55] * marktraceur scrolls [05:17:14] https://meta.wikimedia.org/wiki/Research_talk:Wikipedia_Editor_Survey_August_2012/Questions [05:19:33] Amgine: Which part of that page? Several appear to be today [05:20:09] The section where they admit they are reinventing every damn wheel in demographic survey questionaires. [05:22:35] (shortly after 10:30ish [05:44:11] Amgine: Maybe I'm just not seeing it. *shrug* [05:46:33] No worries. Just that every demographic question in the draft survey, which they already used pretty much unchanged in 2011, is very wrong, and many of their questions do not conclusively collect one type of answer, so cannot be usefully used. [05:56:02] Ah [05:56:31] Amgine: I would not have noticed that, even staring at the page for hours :P have you helped to fix this? [05:57:04] I did sort of try. They presented me with their methodologist. [05:57:35] [06:06:53] RoanKattouw_away: ping [06:09:42] Amgine: I'm sorry to hear that :/ [06:24:12] preilly: Pong [06:25:01] RoanKattouw: can I use your login to watch the nbc feed? [06:25:27] Does it take AT&T credentials? [06:25:32] RoanKattouw: yes [06:25:36] Alright [06:25:40] Lemme change my password real quick [06:25:41] RoanKattouw: PM [06:31:18] RoanKattouw: are you going to reply? [06:31:30] Yeah hang on [06:31:36] ATT is being annoying [06:31:50] RoanKattouw: is it UVERSE [06:32:00] Yes and no [06:32:11] It's marketed as UVerse but it's really just 6 Mbps DSL [06:32:43] I think 10 or 12 is the highest they offer in my neighborhood [06:39:07] RoanKattouw: https://bugzilla.wikimedia.org/show_bug.cgi?id=38125 [07:54:24] ! [07:54:25] ok, you just typed an exclamation mark with no meaning in the channel, good job. If you want to see a list of all keys, check !botbrain [07:54:26] !g 12345 [07:54:26] https://gerrit.wikimedia.org/r/#q,12345,n,z [07:54:28] * RoanKattouw pokes wm-bot [07:54:34] Meh apparently my conn sucks [08:15:24] RoanKattouw that ! prevents spam, you don't want to autocomplete all keys [08:15:36] !d [08:15:45] fine, autocompletion is off here [08:15:49] but it's for #mediawiki [08:16:05] @configure autocomplete=true [08:16:05] This value can't be stored [08:16:13] Yeah sorry about that [08:16:25] I was testing a friend's IRC bouncer [08:16:35] it's ok :D [08:27:49] irc bouncer? [08:44:32] It's like an IRC relay that keeps him on IRC when his laptop is turned off [12:51:18] <^demon> hashar: What's up with https://gerrit.wikimedia.org/r/#/c/12598/? I can't seem to review it? [12:51:35] ^demon: that is a draft :-) [12:51:45] ^demon: though I have sent patchset 2 as a non draft [12:52:01] <^demon> Ugh, that's kinda annoying [12:52:03] ^demon: so it shows up :-( you can safely ignore it, I might eventually have it usable in december :) [12:52:09] <^demon> I mostly hate the drafts feature, the more I use it. [12:52:31] ^demon: I have added you as a reviewer :-) [12:52:44] ^demon: also you have some korean / chinese alter ego using the mail chad@wikimedia.org :) [12:52:47] <^demon> Anyway, was going to -1 it. The reason we didn't do eval.php as a class was so you don't have to "global $foo" everything. [12:52:48] in Gerrit I mean [12:52:54] <^demon> So you're already in the global scope. [12:53:09] ahhhh [12:53:20] <^demon> Yeah, I was gonna do it before but Tim said no, it was by design. [12:53:20] I haven't thought about that [12:53:36] anyway I have renamed it to cli [12:53:49] fighting with reflection class right now :) [13:01:33] wikitrunk> help [13:01:34] string(25) "/** the help command!! */" [13:01:35] yeahhh [13:01:49] so the cli could be self documented :) [13:02:15] <^demon> Oh upstream, they're writing a "prolog cookbook" to document some cool things you can do with it. [13:02:31] ^demon: yup you shared the change earlier [13:02:42] I have no idea what prolog is used for though [13:02:53] is that to edit / change the workflow? [13:02:57] <^demon> Yeah. [13:03:22] <^demon> Apparently, you can also have it detect what sorts of files are being changed--and require a certain reviewer as a result. [13:03:34] should we use that if wanted to add a new "deployed" state to our changes? [13:03:41] ahhh [13:04:04] I had a bug about adding Asher as a reviewer whenever a sql file got changed [13:04:09] I think Rob reassigned to you [13:04:25] I did some ant magic to have it detect whereas a change introduce any .sql change [13:04:40] (by using git diff HEAD^) [13:04:55] but it remains unfinished [13:06:28] <^demon> Yeah, there's not an easy way to really do that in gerrit that I know of yet. [13:06:38] <^demon> Like I said, apparently it can be done in prolog but I want to see the examples. [13:07:02] I would guard you against prolog, mainly cause that is not a well spread language [13:07:13] but probably not worth than writing the logic using ant :-D [13:07:37] make sure to use plunit : http://www.swi-prolog.org/pldoc/package/plunit.html ;) [13:36:15] I thought drafts were private, and you (annoyingly) could view them [13:36:21] was that fixed? [13:38:18] <^demon> No, it's not fixed. [13:38:34] <^demon> You can still always view them from gitweb if you know where to look [13:38:46] <^demon> hashar's draft is visible because he did something that messed it up. [13:39:05] patchset 2 was submitted to refs/for/master [13:39:08] so that made the change public [13:40:00] <^demon> But for some reason, until you added me as a reviewer it was still not letting me review. [13:40:23] <^demon> Even though I could see it. [13:40:36] <^demon> There's some bug here, methinks. [13:50:39] ^demon: maybe because the first patchset was a draft [13:50:58] <^demon> Yeah. If you'd pressed "Publish" it would've opened up for review. [13:51:01] then a non draft patchset was submitted, might have made the change public [13:51:17] <^demon> Some reason pushing it to refs/for/master instead put it in between. [14:10:50] @infobot-share-on [14:10:51] You can't configure this channel to share local db, because this channel is using shared db with another channel, thus the local db is locked [14:10:51] petan: Invalid arguments [14:10:58] hashar ^ [14:11:52] !google <3 hshs [14:11:52] http://google.com/search?q=+%3c3+hshs [14:13:21] !google <3 hshs [14:13:21] http://google.com/search?q=%3c3+hshs [14:13:27] hashar fixed [14:13:37] well done :) [14:13:41] :D [14:14:06] you can also use $url_encoded_1 - 9 as one word's or * instead of number for all parameters [14:14:19] like !test3 [14:14:21] !test3 [14:14:24] !test2 [14:14:24] $* [14:14:27] test2 [14:14:29] not 3 [14:14:39] !test2 a b cd gfsd gdsf ghsh [14:14:39] a b cd gfsd gdsf ghsh [14:16:30] hashar I will talk here because ^demon hates me [14:16:47] hashar I chose $* over that what you suggested because it's used in bash as well [14:17:00] and I don't like to introduce new syntax [14:17:04] like lua did [14:17:06] <^demon> I don't hate you petan <3 [14:17:36] I don't know why lua has ~= instead of c-ish one [14:17:51] I had a head ache from lua's syntax [14:19:25] this is for docs... [14:19:26] !google some special symbols % 6 7#%#$%@ [14:19:26] http://google.com/search?q=some+special+symbols+%25+6+7%23%25%23%24%25%40 [14:20:12] <^demon> hashar: I was thinking, rather than trying to setup the database *on* the gerrit host, what if we just made a new small instance called gerrit-db that attempts to do a misc. db server install? [14:20:26] <^demon> Then we can strip all that junk out of gerrit.pp [14:20:39] http://meta.wikimedia.org/wiki/Wm-bot#Infobot [14:38:09] ^demon: the db stuff is mostly fine [14:38:31] <^demon> Well, it couldn't hurt to segregate it. And it'd make it more like production. [14:38:33] that is just lacking a way to get the root account [14:38:39] though Debian has a maintenance user I think [14:38:42] maybe we could use that [14:38:58] <^demon> I'm going to make a new instance and we'll try to get mysql setup for it. [14:38:58] there must be some deb helper to get a sql connection [14:39:18] well [14:39:22] just setup the db on the same instance [14:39:38] the issue would still be to pass the password to gerrit [14:40:04] ideally we would have a private puppet repository on a per project basis :/ [14:40:22] or a way to generate the password on first time and store them centrally in puppetmaster on a per node / per project basis [14:40:29] but my puppet knowledge is realllllyyy limited [14:41:11] ^demon: debian definitely have a system user [14:41:17] [client] [14:41:17] host = localhost [14:41:18] user = debian-sys-maint [14:41:19] password = fooobarsecret [14:41:29] I am not sure what its privileges are though [14:42:12] <^demon> I'm going to try the "database on a second host" idea. I really think it'll be easier. [14:42:18] <^demon> And again, it'll be more like production. [14:42:41] well if you want that DB to be installed by puppet you face the same issue [14:42:50] if you set it up manually, you can as well have it setup on localhost :-) [14:43:14] anyway, the database is just an IP address / hostname in the Gerrit class [14:43:27] <^demon> I know. [14:43:32] which can be passed to the role class via a global variable filled in open stack manager [14:43:54] <^demon> I'm just going to install it manually on the second host. I'd rather keep the main host like production and not self-host. [14:44:17] <^demon> Ideally we'll use a database service, but that's not available yet. [14:51:27] ^demon: makes sense :-D [14:51:45] if you keep it similar to production, might as well remove the $self_db parameter that is floating around [14:54:20] <^demon> I am in my current patchset. [14:57:23] \O/ [14:57:33] ^demon: was [14:57:34] [14:57:42] ^demon: was the gerrit/role split of any help ? [14:57:51] <^demon> Yes, a lot [14:58:47] <^demon> I was also thinking of making the $name parameter to gerrit::instance() the hostname. [14:58:54] you can have a look at the varnish class for inspiration and refers to its author : mark :) [14:58:57] he is very helpful [14:59:21] and Faidon gave me tonnnns of advices :-D [14:59:54] <^demon> Uh oh, syntax error :( [15:00:07] <^demon> What was the trick to check those locally? [15:00:15] <^demon> (As in on my laptop) [15:00:20] ^demon: just pass the hostname to the gerrit_config hash maybe ? [15:00:31] ^demon: there is a rake script at the root [15:00:35] rake validate [15:00:41] <^demon> I'm killing gerrit_config and making them parameters to the instance. [15:00:42] rake validate[somefile,someotherfile] [15:00:46] or just: puppet parser validate [15:00:54] I should add them as a pre commit hook in my repo [15:01:23] ^demon: make sure to check with mark / paravoid . Cause I am pretty sure they will prefer a config hash, that is what is done on varnish [15:01:32] <^demon> I don't have puppet :( [15:01:36] ohhh [15:01:40] <^demon> hashar: I already talked with Ryan, this is what he wanted. [15:01:48] sudo gem install puppet && sudo gem install puppet-lint [15:01:56] <^demon> Ugh, gems. [15:01:58] ^demon: ahhhazehahahaha [15:02:11] ^demon: different ops have different styles I guess :-D [15:02:41] since Ryan is most probably going to be the final reviewer, you definitely want to adopt whatever style he prefers indeed [15:03:10] ^demon: also sudo gem install rake [15:03:14] ^demon: if you don't have it yet [15:03:18] <^demon> Well doing them as parameters means we can set the variables from within labs. [15:03:28] <^demon> And not have to add a new role for each instance in labs we want to spin up. [15:03:32] ohhhh that is true [15:03:43] our openstack manager does not support hashes / array yet ;-/ [15:04:10] for varnish I have just added a new role but that is far from being ideal [15:04:18] cause it is all hardcoded for beta [15:04:30] so you end up having a role class setting up a bits.beta.wmflabs.org domain [15:04:46] it is very unlikely any other project we will want to set up a varnish bits cache though [15:05:40] chrismcmahon: beta is a bit broken at the momment [15:06:16] chrismcmahon: the shared filesystem prevents git from completing its action [15:12:32] <^demon> hashar: Also, once you pass the config to gerrit::instance, it can store it somewhere so you can reference it in all config files without having to know the role. [15:13:02] <^demon> Like gerrit::instance::foobar or something. [15:14:52] <^demon> We should also move the apache config to templates, so we can replace the hostname in. [15:24:06] ^demon: notpeter is rewriting the apache classes [15:24:21] I think he is going to make them more modular / configurable [15:35:38] <^demon> Ok, created gerrit-db, installed mysql-server and created reviewdb with a user "gerrit". [15:35:46] <^demon> Password info is in /root/mysql [16:13:07] <^demon> hashar: The generic::mysql::root change should probably be spun off on its own now. [16:32:53] ^demon: let me split it out :) [16:33:12] <^demon> Ok, great :) [16:33:38] though I might as well delete it [16:33:56] <^demon> If you're not going to resubmit it separately, can just leave it out. [16:34:41] ^demon: well just remove it [16:34:49] there is no $mysql_root_password anyway [16:35:02] my original intent was to have the root pass to be generated and stored by puppet [16:35:05] but that is not possible [16:35:09] anyone knows whose live hack this is? Aaron's? [16:35:09] if ( $host === 'test.wikipedia.org' && !$secure [16:35:09] - && !preg_match( '!thumb\.php!', $_SERVER['REQUEST_URI'] ) ) [16:35:09] + && !preg_match( '!thumb(_handler)?\.php!', $_SERVER['REQUEST_URI'] ) ) [16:35:09] so all the code can be trashed out [16:35:23] MaxSem: most probably [16:35:43] <^demon> Done. [16:35:48] ^demon: thanks :) [16:40:51] ^demon: out for now :) [16:40:53] see you later! [17:47:23] <^demon> JeroenDeDauw: Somebody took the time to write examples of how to use prolog for things like "no self review" :) https://gerrit-review.googlesource.com/#/c/37102/ [17:48:12] ^demon: cool [18:31:07] when is the echo deployment window? [18:32:43] jorm: 1-3, quoth werdna [18:32:56] awesomesauce. [19:02:01] kaldari: http://www.mediawiki.org/wiki/Extension:Echo/Review [19:02:25] bsitu: http://www.mediawiki.org/wiki/Extension:Echo/Review [19:02:57] AaronSchulz: I have resubmitted the memcached debug log stuff that cause some notice in production. https://gerrit.wikimedia.org/r/#/c/16668/ [19:03:21] AaronSchulz: the reason was trying to get socket information out of a non socket resource (our memcached lib uses a file pointer, not a socket :p ) [19:04:18] werdna: got time for an AbuseFilter query? is there a performance cost to having many filters? especially those for especially rare events with many lines in the filter? [19:04:40] sort of. it depends on the way the query is written. [19:06:06] werdna: How would you like review on these things? [19:06:21] Thanks Werdna. [19:06:25] marktraceur: maybe put under the dot point that you've looked at it? [19:09:35] AaronSchulz: you are too fast :) danke! [20:12:24] werdna: This stuff is hard to review without any comments [20:15:30] kaldari: sorry :(. Which parts do you think need more comments? [20:15:40] heh, most of it :) [20:15:44] :D [20:16:13] it's understandable without it, but takes longer to figure out [20:17:42] kaldari: nod. [20:21:37] werdna: what are email subject params for? [20:21:58] parameters for the message being used for the email subject line [20:22:13] what would be an example? [20:22:41] array( 'user', 'commentText' ) [20:23:04] Then a message with the text "You have a new message from user $1. Its content is $3" would expand out correctly [20:23:11] ah [20:23:22] ('user' expands into a formatted and an unformatted parameter. That's why it's $1 and $3) [20:49:58] werdna: What is this about: $title = $event->getTitle() ? $event->getTitle() : Title::newMainPage(); ? [20:50:03] specifically the Title::newMainPage() part [20:50:09] I haven't seen that before [20:50:15] it's existed for yonks [20:50:36] so it just makes sure there is some title [20:50:47] where is that? [20:51:01] NotificationController.php line 74 [20:51:26] yeah [20:51:32] job queue expects a title [20:51:38] it just adds a bogus one if one isn't specified [20:51:43] it's never used [20:51:50] OK, makes sense [20:52:26] good to add a comment if you're ever putting something unintuitive in [20:52:43] yeah [20:52:54] I am going to be in comment training after this aren't I? :)) [20:53:24] hehe, time for the whips and gummy bears! [20:53:40] what do the gummy bears have to do with anything? [20:53:49] positive reinforcement [20:54:01] everyone loves gummy bears [20:54:40] :D [20:55:03] kaldari: Good call. Gummy bears are much better than a carrot, and whips are much scarier than a stick. [20:55:31] yeah, sticks and carrots are very ineffective :) [20:55:41] unless you're a rabbit [20:58:53] I like carrots [20:58:55] big fan [21:13:31] werdna: OK, I think I'm done with the first pass [21:13:38] woohoooo [21:14:05] let me fix up those last few issues [21:18:19] kaldari: bsitu gerrit-wm: New patchset: Werdna; "Echo: A few adjustments based on feedback from Ryan Kaldari and Benny Situ." [mediawiki/extensions/Echo] (master) - https://gerrit.wikimedia.org/r/17226 [21:25:49] werdna: should I +1 or +2 (are other people also reviewing it in gerrit)? [21:25:57] +2 and I can get going :) [21:26:02] ok [21:26:05] you raised the issues so I figure you should review [21:26:42] I have to actually test it if I'm +2ing though :) [21:27:15] might be an idea [21:28:45] Image shrinking! https://gerrit.wikimedia.org/r/17341 [21:32:05] "Honey, I shrunk the PNGs!" <-- far less dramatic movie [21:32:44] werdna: found a small problem [21:32:51] https://gerrit.wikimedia.org/r/#/c/17226/ [21:33:18] kaldari: aha [21:34:30] I haz error [21:34:30] probably the same for line 109 [21:34:33] kaldari: new patchset [21:34:56] DiscussionParser.php EchoDiscussionParser 'title' => $article->getTitle(), [21:35:01] $article is undefined [21:35:14] Reedy: which line? [21:35:32] 62 [21:35:40] just use $title [21:36:14] yeah, fixed it [21:36:57] overlay seems to work now [21:39:14] werdna: yep [21:39:26] werdna: how do I give myself a notification anyway? [21:39:44] post anonymously on your user talk page [21:41:01] werdna: no worky! [21:41:20] do I need to configure it? [21:41:25] shouldn't [21:41:34] make sure you run your jobs [21:41:39] if you have jobs running separately [21:41:58] I don't (that I know of) [21:43:44] you probably won't then [21:43:52] php maintenance/runJobs.php [21:43:55] just to make sure it's clear [21:45:35] werdna: why does it require the post to be signed? [21:46:10] kaldari: probably because you hadn't pulled in patchset 5/6, which had the bugfix that Reedy mentioned a few seconds ago [21:46:13] let me test that it works unsigned too [21:46:18] ah [21:49:25] werdna: while you're in the css... [21:49:53] hmm? [21:50:30] I won't make you change all the font sizes to ems right now, but at least bump the mw-echo-timestamp font-size from 8px to 9px. I can barely read it. [21:50:54] kk [21:51:40] you gotta keep up old folks in mind! :) [21:51:46] up=us [21:52:09] * marktraceur gets out some popcorn GIFs for the last 8 minutes of the deployment window [21:52:17] heh [21:54:46] kaldari: done, and I've also fixed the unsigned issue [21:54:51] which had a few more causes [21:54:53] cool [21:55:02] are we good? :p [21:55:08] retesting... [21:56:36] I don't see the new patch [21:56:50] werdna: ^ [21:57:18] sry [21:57:29] there we go [21:59:19] werdna: seems to work. Still need the font-size bump though [21:59:22] * werdna sneakily extends deployment window by a half hour [21:59:25] kaldari: should be in that patch? [21:59:34] nope [21:59:51] https://gerrit.wikimedia.org/r/#/c/17226/8/modules/base/ext.echo.base.css [21:59:59] oh sorry [22:00:06] I was looking at the other css file :P [22:00:29] OK, I like it now :) [22:00:56] alright, wanna get this show on the road then? [22:01:48] one sec... [22:02:24] werdna: Deprecated: Call-time pass-by-reference has been deprecated in /var/www/wikimedia-git-repos/core/extensions/Echo/includes/DiscussionParser.php on line 292 [22:02:40] tried to write on the talk page [22:03:09] Note there are two people have have a 3pm deploy window. https://wikitech.wikimedia.org/view/Software_deployments [22:03:51] we might need to reschedule this one then [22:03:51] Ryans doesn't affect this [22:04:05] Multiwrite will be MW stuff [22:04:31] tewwy: where abouts? [22:04:42] who's in charge of the MultiWrite deploy? [22:04:52] kaldari: that's wednesday [22:04:54] oh today is wednesday [22:04:56] never mind [22:04:59] lols [22:05:07] Asher/Aaron at a guess [22:05:20] AaronSchulz: prod? [22:05:51] werdna: moo? [22:06:02] AaronSchulz: "MultiWrite" – is that you? [22:06:08] yes, why? [22:06:27] apparently you have a deploy window now – is it going to cause trouble if our Echo deployment goes overtime? [22:06:45] my deploy is to test2?wiki, [22:06:55] so no? [22:07:10] I don't think so [22:07:15] <3 [22:07:36] bsitu: strict warnings should be gone with latest patchset [22:07:42] Great! :-) [22:07:46] * werdna changes his PHP error settings to include strict warnings [22:07:48] so excite. [22:07:50] werdna: mw-echo-overlay needs a z-index in order to work in monobook [22:08:01] fucking monobook. [22:08:05] werdna: must be at least 3 [22:08:10] I think [22:08:12] echo's overlay should be top-level. [22:08:19] well, -1 under modal windows. [22:08:35] wernda: okay, running test [22:08:44] fucking monobook [22:08:49] i should probably install mediawiki on this laptop at some point. [22:08:57] fucking skins/* [22:09:32] * werdna tries with monobook [22:09:37] did wikipedia seriously look like that once? [22:09:38] holy hell [22:10:32] unfortunately [22:11:56] working on it. [22:12:45] fucking monobook [22:12:52] I don't understand why some people still like that skin [22:13:03] hey it has a place in our hearts :) [22:13:10] it was written in like 2004 :p [22:13:43] monobook was pretty excellent for its time. to be honest, i don't really think anyone knew exactly what we were doing until very recently. [22:17:11] monobook is like the King James Bible - it has lots of problems, but some people are strangely attached to it [22:17:18] [22:17:58] i'm not sure anyone knows what we're doing. [22:19:53] kaldari: https://gerrit.wikimedia.org/r/#/c/17351/ [22:20:11] kaldari: https://gerrit.wikimedia.org/r/#/c/15830/ [22:20:18] jorm: improving a CMS. [22:24:51] jorm: we need a z-index guide for the MediaWiki stylebook [22:24:56] er styleguide [22:25:09] so people don't just pick numbers randomly [22:25:40] they always will [22:25:44] New patchset: Werdna; "Add Echo" [mediawiki/core] (wmf/1.20wmf8) - https://gerrit.wikimedia.org/r/17352 [22:25:45] alolita: hello [22:25:46] kaldari: [22:26:14] good idea. [22:26:21] "What does the device say about its z-index level?" "It's over 9000!!!!" [22:28:32] when I worked at viacom, we had to occassionally use z-indexes in the billions in order to place things over 3rd party iframed ads (which were always in z-index war with the websites. [22:28:39] Danny_B: hello [22:29:36] z-index:∞ + 1; [22:29:41] https://gerrit.wikimedia.org/r/#/c/17352/ [22:34:04] are we deploying? [22:34:23] yep [22:34:26] waiting for Aaron's ok to run scap [22:34:29] since he has a concurrent window [22:34:32] but only to testwiki [22:34:36] then we'll bang on it for 5 minutes [22:34:38] and then to mediawiki [22:34:46] rock on. [22:35:18] once it's done we can run a battery of talk page edits and the like. [22:35:22] reversions, etc. [22:35:32]