[00:03:55] wouldn't that defeat Cave Mode? [00:16:49] only in so far as sometimes it would help to get feedback before i head too deeply down rabbit holes. [00:17:06] also: i've been disturbingly productive the past hour or so. [00:37:48] *robla was going to cat herd for a 1.18 call, but wants to figure out the pdf server thing first [17:16:31] JeroenDeDauw: ping [17:21:04] Krinkle: pong [17:29:59] JeroenDeDauw: never mind, I thought you were going to ask about Macs and VPN in general, which would be better suited here than on the etherpad chat [17:34:21] RoanKattouw: Do you recommend using instanceof, is_null or is_object for the check from makeTitleSafe > [17:34:45] I usually just use ! [17:34:52] k [17:35:10] *RoanKattouw is working on fixing https://www.mediawiki.org/wiki/Special:Code/MediaWiki/96837#c23717 , thinks to have fixed it in the code but the bug won't go away [17:40:22] RAAAAAAAAAAHHH [17:40:25] FUCK YOU PHP [17:40:31] if ( $this->idsLoaded && !isset( $this->data[$id] ) ) { [17:40:38] And guess what $this->data[$id] is null [17:40:42] And isset() returns true for null [17:42:56] hehe [17:43:00] *RoanKattouw resigns to using false instead [17:43:08] or array_key_exists [17:43:14] Yeah [17:44:16] In JS != null is like isset() in PHP. Returns true for both undefined and null [17:44:52] although in JS null can in fact equal null, which is interesting :D [17:45:11] oh, of course in php too. It's MySQL where null isn't null [17:45:36] Yeah in MySQL null is like NaN [17:46:47] IIRC, C++ does not have an isNaN() function. So what I found on stackoverflow was essentially to check for if ( !( var >= minusInfinity ) ) [17:47:49] The nl i18n has "Speciaal:Extensies"? Really? That's some nice jargon confusion right there [17:48:09] I thought it was translated as "Uitbreidingen"? [17:48:29] Translating "Gadgets" to something that translates back to "Extensions" seems like tempting fate to me [17:48:48] I mean I understand if it doesn't round-trip to the same word, but it should at least not round-trip to a word with a /different meaning/ [17:51:31] well let's be honest [17:51:39] they only have different meaning because we've chosen to chop them up that way [17:52:04] Well sure [17:52:10] But we can avoid asking for trouble [17:52:36] The case we have here is that translateToEnglish( $specialPageNames['gadgets']['nl'] ) == 'Extensions' [17:54:32] wait til we have plugins too. that will be fun :-P [17:55:01] what the fuck? http://lists.wikimedia.org/pipermail/foundation-l/2011-October/069201.html [17:55:01] what word do they use in the interface for extension? [17:55:24] yeah, I heard a rumor [17:56:05] omfg [17:56:14] what I didn't hear was the decision to lock down the site [17:56:24] there's NO WAY that law stays. [17:56:59] never understimate government stupidity [17:57:38] I would ask them to wait til it's actually passed before taking that action [17:57:41] it looks like this is basically burlusconi trying to avoid getting further embarressed. [17:59:23] Don't underestimate how crazy Berlusconi is and what he gets away with [17:59:56] damn it, time to read the whole thread I guess [18:00:27] Yup [18:00:41] I actually got a query about the technical feasibility of their pplans [18:01:21] i hadn't heard anything about this until just now. [18:02:07] seriously? they are going to lock the wiki? [18:02:17] if they are worried about liability [18:02:23] They can't hard-lock it [18:02:25] why don't they even fucking bother to talk to our lawyers? [18:02:28] another option would be mass resignation as sysops/crats [18:02:31] They say they did ask [18:02:44] I asked whether they'd responded but that question wasn't answered [18:02:58] I was like "are you sure this applies since we're in the US" but they seemed quite certain [18:03:06] it doesn't [18:03:11] i don't know, reading on this - i mean, it may be that if they edit on the English Wikipedia, that the law also applies. [18:03:22] Italian law doesn't mean shit to us [18:03:23] I think this needs a lawyers opinion. [18:03:28] this is why we host in the US [18:03:31] Good luck enforcing that against us, yeah [18:03:38] right, no. [18:03:45] the editor is subject to italian law as an italian citizen... that's the problem. [18:03:55] i mean, "if italian citizen edits english wikipedia, then that citizen is subject to the law" [18:03:59] but if they aren't op on en pedia, the ball will land in [18:04:00] Yes and? [18:04:02] they'd have to post the retraction. [18:04:04] our court (so to speak :-P) [18:04:05] That citizen can't make us do shit [18:04:05] ugh [18:04:18] https://bugzilla.wikimedia.org/showdependencytree.cgi?id=29068&hide_resolved=1 [18:04:18] or get fined or whatever. [18:04:33] They can't actually lock the wiki completely [18:04:54] they can protect every page, right? [18:05:03] yes. [18:05:07] we should de-sysop anywho who tries that [18:05:12] er no [18:05:30] we should try to mdiate, not add fuel to the fire [18:05:34] *mediate [18:05:35] true [18:05:45] Their plan, IIRC, was to add .* to the title blacklist [18:05:46] are the only editors of that wiki people from italy? [18:06:00] if not, they have no right to block the wiki [18:06:20] it's not a matter of rights ... well it is, but not in the sense you mean [18:06:37] They also want to use JS to prevent reading but acknowledge that's circumventable [18:06:52] preventing reading is a bfd [18:06:56] that's for damn sure [18:07:53] I think it's very wrong, and a very incorrect action to effectively shut down a language wikipedia because a country decides to enact a stupid law [18:08:12] why don't they shut down every wikipedia then? [18:08:18] there's a change someone from Italy will edit it [18:08:27] I bet a lot of people from italy edit enwiki [18:08:28] be realistic [18:08:51] we're talking about a wiki where probably 99.9% of the edits are from folks that will be affected by the law [18:08:52] it's a terrible reaction [18:09:02] suggest an alternative [18:09:13] wilful violation of the law? [18:09:26] can we offer anything in the way of legal protection? [18:09:44] why should we? people have the ability to choose whether to follow the law or not [18:10:05] we shouldn't force people to follow the law [18:10:07] if we suggest that people violate the law [18:10:22] then we should be thinking about ways to help people with the consequences of that [18:13:18] http://lists.wikimedia.org/pipermail/foundation-l/2011-October/069221.html this sort of thing is a problem. however bear in mind that the law hasn't passed yet, we're supposed to be in the protesst phase [18:15:15] Oh, the Italians went ahead and did it [18:15:19] https://it.wikipedia.org [18:16:08] so there it is then [18:16:20] I had a feeling when we got the report in the other channel [18:32:27] Reedy: Has robla talked to you about Translate CR? [18:38:22] I did talk to Reedy about it [18:41:04] OK [18:41:07] And? [18:49:37] so, what is confusing about the design for the admin stuff? [18:49:49] RoanKattouw: ^ [18:49:54] So, on https://www.mediawiki.org/wiki/File:Feedback-Dashboard-Phase1-AdminActions.png [18:50:09] After a while we assumed that comment #2 is what it looks like to an unprivileged user [18:50:17] did you read: http://www.mediawiki.org/wiki/Feedback_Dashboard/design#Administrative_Behaviors [18:50:20] And that #3 is what it looks like for an admin, right? [18:50:33] it's explained in the document. [18:50:38] Ooooh right [18:50:41] non-admins NEVER see hidden comments. [18:50:41] NOW I do get it [18:50:56] heh. [18:51:02] #2 and #3 are both for admins, and #2 turns into #3 when the link is clicked [18:51:04] Thanks for the cluebat [18:51:09] yeah, i always explain stuff in the document. [18:51:14] right. exactly that. [18:51:34] *RoanKattouw sends an e-mail to Andrew with the transcript [18:53:11] word. [18:53:52] jorm: So when logged-in as a user the comment is not visible at all (not greyed out either), and as an admin it is greyed out until the admin unhides it temporarily to view it [18:54:05] right. [18:54:07] k [20:26:17] i am merging some changes for CentralNotice fixes that i am about to deploy. CN (afaik) only runs on meta, which is running 1.18, so i will merge into the 1.18wmf1 branch, but should i also merge into 1.17wmf1? [20:26:44] No [20:26:52] 1.17wmf1 is going to die in a matter of hours [20:27:03] orly [20:27:13] We're deploying 1.18 to all wikis later today [20:27:16] glad i asked :) [20:27:18] So alternatively, if you're unsure [20:27:26] You can wait a day, and everything will be on 1.18 [20:27:40] cool thanks roan [20:57:30] RoanKattouw: awesome [20:57:39] ? [20:57:44] (1.18 today) [20:57:56] Oh, right [20:57:58] I can't wait to nuke the 1-line gadgets that set window.mygadget_disable=true; [20:57:59] *RoanKattouw will be asleep for that [21:00:07] Hm.. I see http://nl.wikipedia.org/wiki/Blankenbach is still doomed [21:27:05] RoanKattouw: do you have a second? [21:27:43] Sure [21:28:08] so for the central notice fixes i need to push out, i should be svn up'ing on fenari in /home/wikipedia/common/php-1.18, yes? [21:28:40] Yes [21:28:43] ok [21:28:46] But are you sure you want to be doing that /now/ ? [21:28:53] As opposed to tomorrow [21:29:05] *cough* 118 deployment *cough* [21:29:36] for timing reasons yes - the chnges only need to happen on Meta, which is already running 1.18 and the rest of the deployments, as i understand it are not set to happen for another hour and a half [21:29:41] but im willing to be talkd out of it [21:30:00] particularly because i need to run scap, but there i see a lo tof things that make me nervous when i run svn status: [21:30:02] This is something that you're absolutely sure only affects meta (e.g. because it changes the CN UI or something)? [21:30:13] http://pastie.org/2640264 [21:30:29] C extensions/LiquidThreads/classes/ParserFunctions.php [21:30:31] whoops [21:30:48] heh yeah [21:31:04] *RoanKattouw had fixed that file but forgot to run svn resolved, done now [21:31:08] i'll dbl check but afaik CN is only running on meta and test wiki and all of the changes i want to push are within the CN extension [21:31:37] Doesn't CN run on all wikis for notice delivery? [21:33:20] RoanKattouw: er well yeah you're right [21:33:27] 'default' => true [21:33:42] okok so we should hold off [23:08:29] TimStarling: AaronSchulz: ready to start? [23:09:59] Alright, good luck folks [23:10:00] I'm signing off [23:10:10] I guess so, as long as we can skip dewiki for now [23:10:29] If something terrible happens and you need me, call me, my cell is on officewiki [23:10:41] Otherwise I'll be back here around 07:30-08:00 ish UTC [23:10:46] I posted a tentative plan to blog.wikimedia.org [23:11:06] "Our tentative plan for today is deploying first to fr.wikipedia.org, then pl.wikipedia.org, then en.wikipedia.org, and then probably a few more sequentially before deploying to the rest in bulk. ??At the end of our window, we will be stopping deployment, even if we???re not done (scheduling a followup window if needed)." [23:11:35] is asher on holiday or something? [23:11:38] TimStarling: does that look reasonable? [23:11:47] yes, sounds good [23:12:55] Asher is, but he's back tomorrow [23:13:01] and he was online earlier [23:13:22] he's officially back tomorrow, but he said he'd be around in case of emergency tonight [23:13:35] maplebed_ is our ops support for this deploy [23:13:40] ok, well there's no emergency so far, as long as we can delay dewiki [23:14:23] TimStarling: what would you think of pulling db44 out of rotation, and then deploying to dewiki? [23:15:06] (delaying dewiki altogether is another option...it wouldn't kill us to wait 24 hrs on that) [23:15:35] TimStarling: are you editing db.php? [23:16:26] not anymore [23:16:33] thnaks. [23:17:13] I think it would be safer to delay deployment on it [23:17:31] the only difference is three missing indexes; the rest of the schema changes are there. [23:18:11] we can wait....I can schedule a separate window for dewiki [23:21:33] TimStarling: if you think we should delay it and robla says ok, that's what'll happen, but it does seem like an easy change to bring db44 back into line so the deploy could continue. [23:22:00] if you want to do the schema change on it then go ahead [23:22:11] if it's done before the end of the deployment window then we can deploy to dewiki [23:22:23] but it's a million rows, we're not sure how long it will take [23:22:41] that sounds like a good plan to me. we have to bite the bullet one way or another to get the index built [23:22:51] ok, I'll pull it out and start the change. [23:23:08] TimStarling: do you want to wait to get the itwiki stuff figured out, or should we just start? [23:23:24] *robla assumes AaronSchulz will be the one to pull the trigger [23:23:40] I can do it, since Aaron is not talking to us right now [23:23:55] db44 is now out. [23:24:16] TimStarling: you mean AaronSchulz? [23:25:06] yes [23:25:10] um..hey [23:25:31] yes, no one is doing dewiki now :) [23:25:39] AaronSchulz: ready to deploy to frwiki? [23:25:54] if so...fire when ready [23:25:58] yeah, I had to fix some stupid bug [23:31:47] guillom: when you have a sec, could you switch http://meta.wikimedia.org/wiki/Wikimedia_maintenance_notice over to point to the latest blog entry? [23:32:50] Anyone interested in a sml chat ?? [23:34:32] done [23:34:51] mcbaine, please stop asking in these channels. We're in the middle of a major operation. [23:37:13] ok [23:43:05] PHP Warning: Invalid argument supplied for foreach() in /usr/local/apache/common-local/php-1.18/includes/specials/SpecialRecentchanges.php on line 808 [23:45:03] TimStarling, AaronSchulz, ^demon, shall we postpone the commit access queue review meeting to Thursday since you are in the midst of deploying? [23:45:31] Invalid argument supplied for foreach() in /usr/local/apache/common-local/php-1.18/skins/Standard.php on line 113 [23:46:01] is ^demon online? [23:47:17] The standard one would seemingly exist in trunk also [23:47:58] yes, ^demon just got online. [23:48:25] so, TimStarling, AaronSchulz, ^demon, what would you prefer? ten-min. meeting now + the Thursday meeting, or a somewhat longer meeting Thursday? [23:48:57] maybe it's better to wait until thursday, since there's already a few things going on at once here [23:49:49] robla: TimStarling the schema change to db44 is finished. I have not yet put hte host back in rotation. [23:49:59] ok, TimStarling, we'll wait, then. thanks [23:50:30] TimStarling: if you have a chance later in your working day to put notes in OTRS regarding the two applicants I emailed you about, please do. I can carry on from there. [23:50:46] ok [23:50:51] Reedy: RC->setup() must not get called before the line 808 stuff [23:51:16] *AaronSchulz doesn't get those in trunk [23:51:53] is it replicable onwiki? [23:53:42] ok, ^demon, TimStarling, AaronSchulz, we'll review the commit access queue on Thursday instead of today, happy deploy day :) [23:54:08] Reedy: I tried hitting RC with no result rows, that didn't trigger it [23:54:10] maybe it is a feed [23:55:47] probably [23:56:05] So much for a stable connection