[00:21:21] ^demon|away: http://lists.wikimedia.org/pipermail/pywikipedia-l/2012-February/thread.html#7236 The Pywikipediabot community seems to be consensing on "no" re moving to git. [00:21:26] Lcawte|Away: ^^^ do you have a preference? [00:34:02] <^demon|away> sumanah: If they're wanting to stick with SVN (or basically not use git) that's up to them. In the long run they'll need to find a new home though (eg: google code or sf.net) since the svn box won't be up *forever* [00:34:31] <^demon|away> s/the svn box/svn/ [00:35:06] <^demon|away> But long-run. There's no immediate rush. [00:35:07] ^demon|away: this would be on the scale of more than 2+ years from now? [00:35:22] ^demon|away: maybe stick something about that in https://www.mediawiki.org/wiki/Git/Conversion#Affected_development_projects ? [00:35:52] <^demon|away> I'd like to have svn completely read only within the next year. [00:36:03] ^demon|away: that is news to me [00:36:20] ^demon|away: that you're thinking on the scale of 1 year from now [00:36:42] <^demon|away> I just made up 1 year now when you asked me a timeframe :) [00:37:58] <^demon|away> It's not set in stone. It's all part of that fuzzy "migrate remaining stuff as time permits and people are ready for it" [00:41:41] TimStarling, ^demon|away, AaronSchulz, as you probably know, we do not need a commit access queue review meeting today [00:41:48] they're all either contractors or we're waiting on more info [00:41:53] I replied to the request for info [00:42:06] ok, thanks sumanah [00:42:11] <^demon|away> I hadn't even looked, been swamped today. Thanks. [00:42:19] yay, meeting cancellation. :) [00:42:22] happy hacking. [00:42:25] wee [01:53:40] San Francisco people: http://www.sfphp.org/events/49609532/?eventId=49609532&action=detail "PEAR and PECL have evolved a lot over the years and to this day, their intricacies still puzzle many developers. This talk aims at providing more insight into the projects, their evolution, when to use them and when not to." [01:54:36] so, Inez, jpostlethwaite, raindrift, robla ^^^ that's tonight at 6:30pm and there are still spots available in case you wanna go [01:56:57] shoot...I'm probably a little too tired to go to this [14:56:31] aude, I've been reminding technical contributors to apply for Wikimania scholarships [14:56:48] aude: do you know whether you're happy with the # of applications for scholarships you've been getting? [15:36:24] sumanah: I'm with the community, I can't figure out Git... and the "new features" (I may not be informed on all of them, but I mean branches and that stuff) Git offers seem pointless [15:36:59] Lcawte: ok, if they don't appeal to you then I'm not going to try to talk about it and explain :) [15:37:29] Lcawte: applying for a Wikimania scholarship? [15:37:43] also, Lcawte, looks like the WMDE hackathon date will probably be June 1-3 [15:38:23] As I've said, unfortunatly not... Wikimania 2012 is in the last week of school for me... but if you want to get Wikimania moved forward a week, I will :P [15:38:38] Ok, cool, is that channel chatter or is it set in stone/official? [15:39:06] (Moving Wikimania forward a week would also allow Wikimedians to attend Comic Con!) [15:39:25] <^demon|away> Yeah, we're not moving Wikimania ;-) [15:39:40] Tss. Killjoy. [15:40:50] Oh, WMDE event is in June now? [15:40:53] <^demon|away> There should be a Comic Sans Con. [15:41:06] :D [15:42:13] Lcawte: There hasn't been an official announcement. And earlier, the rumor mill said May 18-20 so who knows [15:42:29] RoanKattouw: Daniel and I are like 90% sure it's June 1-3 now. [15:42:31] sumanah: I may have been missinformed about Git, so if you want to explain or send me a simpleish page link then yeah... I'll rethink it [15:42:53] Lcawte: ^^^ [15:43:04] Hmm, even longer until I can look at travel costs :/ [15:43:44] Lcawte: let's see. ^demon|away, do you know any "here are the benefits of git" guides that don't act like the reader is an idiot for needing to be persuaded? [15:43:46] Lcawte: It is more technically complex, yeah, but that's because of the additional features and the different model. But if you just want to submit a change, that's still easy [15:44:24] ( git clone $url to check out a repo, git checkout -b myfunproject to start a branch, and 'git review' to submit a change into Gerrit if you have git-review installed) [15:44:25] <^demon|away> sumanah: No, not really. [15:44:37] (oh and git commit to commit your changes locally, duh) [15:44:43] <^demon|away> I've got a fantastic blog post on how git-repack and deltas work. [15:44:52] EOL uses Git (its a project translated on twn) and I found it hard just to check the right thing out with that... [15:45:01] <^demon|away> But hardly entry-level git material :) [15:45:12] I love git just because it allows me to commit locally (and offline). [15:46:22] Lcawte: In brief, you can create commits locally and push them to the server later (great for working without wifi), you can tell it 'save my work so I can go do something else now' in one command, and it'll allow us to review changes before they go into "trunk" (master [15:47:08] now technically the latter would be possible with SVN if we reengineered our code review process [15:47:11] ^demon|away: Is that blog post on line yet? [15:47:25] Yes, but not without human intervention in merging things into trunk [15:47:38] Gerrit automates this process [15:47:53] <^demon|away> RoanKattouw: http://metalinguist.wordpress.com/2007/12/06/the-woes-of-git-gc-aggressive-and-how-git-deltas-work/ is the blog post I'm referring to [15:48:38] Lcawte: so, do those help you a bit? [15:48:47] <^demon|away> !git-repack is http://metalinguist.wordpress.com/2007/12/06/the-woes-of-git-gc-aggressive-and-how-git-deltas-work/ [15:48:47] --elephant-- You don't have permission to do that. [15:48:48] Oh that one [15:48:52] <^demon|away> Oh shut up elephant. [15:48:56] !git-repack is http://metalinguist.wordpress.com/2007/12/06/the-woes-of-git-gc-aggressive-and-how-git-deltas-work/ [15:48:58] !git-repack is http://metalinguist.wordpress.com/2007/12/06/the-woes-of-git-gc-aggressive-and-how-git-deltas-work/ [15:48:58] --elephant-- Successfully added keyword: git-repack [15:49:16] <^demon|away> Roan's in the cabal. [15:50:10] RoanKattouw: 20% today? [15:50:14] Aye [15:50:21] *hexmode is getting nervous about the deadline [15:50:51] <^demon|away> hexmode: Just remember Adam's words on the subject :) [15:51:17] RoanKattouw: If you have a chance after you do what you talked about yesterday, could you look at the un-assigned bugs or krinkle's? [15:51:22] Hmm, MW is moving to Git right? [15:51:27] ^demon|away: adam said what? [15:51:31] <^demon|away> RoanKattouw: My favorite tidbit from that past "Finding/verifying an optimal (space-minimizing) delta-derivation graph feels NP-hard. I now wave my hands furiously." [15:51:34] hexmode: Sure [15:51:39] Lcawte: Yes [15:51:51] <^demon|away> hexmode: "I love deadlines. I like the whooshing sound they make as they fly by." [15:51:56] heh [15:51:56] Lcawte: see https://www.mediawiki.org/wiki/Git/Conversion [15:52:01] *Lcawte goes and searches for a new project as a backup [15:52:04] yes, that sounds about right [15:53:35] <^demon|away> Lcawte: The git conversion is coming like a moving train. You can try to stand in the way, but you'll probably end up looking very foolish, very dead, or both :) [15:54:02] ^demon|away: um, maybe we can find a way to phrase that more pleasantly in the blog post and the wikitech-l mail [15:54:06] hey! I'm here now! [15:54:16] I never stand on train lines though... I stand to the side and watch it go by... [15:54:51] Like the drawing btw... https://www.mediawiki.org/wiki/File:Ideal_git_workflow.jpg [15:55:11] <^demon|away> I should've drawn faces on the people. [15:55:26] <^demon|away> Would've made the diagram more realistic :) [15:55:47] or, ^demon|away, we can ask Heather to actually make it nice-looking [15:56:18] :) [15:56:31] guillom, RoanKattouw, I've added quotes from you in https://www.mediawiki.org/wiki/Git/Conversion#Rationale [15:57:14] <^demon|away> "The delta derivations don???t have to obey causality: a commit made chronologically later can be used to derive one made earlier. It???s just a bunch of blobs in a graph, there isn???t even a strictly necessary notion of time attached to each blob at all to begin with! That data is maintained at a higher level." [15:57:31] <^demon|away> That's a benefit too, but will probably scare people more than encourage them :) [15:58:11] That's one of the scary but awesome parts, yeah [15:58:22] Lcawte: if you could write another message to http://lists.wikimedia.org/pipermail/pywikipedia-l/2012-February/thread.html#7237 drawing together the consensus, that would be great [15:58:32] <^demon|away> Simply put, "find the best commit to diff to, which isn't necessarily the previous commit" [15:58:35] sumanah, ^demon|away, just let me know :) [15:59:45] Hmm, I need to put the heating on, its freezing up here... [16:00:09] <^demon|away> RoanKattouw: Wanna solve the NP-hard problem of finding and verifying an optimal delta-derivation graph? :p [16:00:15] <^demon|away> Sounds like a fun summer project. [16:00:39] heatherw: I think it would be great if you could make https://www.mediawiki.org/wiki/File:Ideal_git_workflow.jpg nicer to look at! But I don't have a spec to give you other than that! [16:00:51] ^demon|away: You mean just proving that it's NP-hard? ;) [16:00:58] okay [16:01:17] <^demon|away> RoanKattouw: No, I think solving the underlying problem. Proving its NP-hard is probably easier. [16:01:35] RoanKattouw: also, could you reply on http://www.mediawiki.org/wiki/Special:Code/MediaWiki/98500#c30435 [16:02:03] hexmode: I did, see https://www.mediawiki.org/wiki/Special:Code/MediaWiki/98500#c30470 [16:03:01] Reedy: johnduhart: you guys said you were looking for roan's reply ^^^ [16:03:02] <^demon|away> RoanKattouw: graph problems are fun :) [16:03:32] hexmode: Reedy didn't get an e-mail for some reason. johnduhart has already fixed it, the fix is on my to-CR list [16:03:39] RoanKattouw: They said they were waiting for your reply yesterday, now I'm confused [16:03:42] ah [16:03:46] We spoke today, it's all good [16:03:50] :) [16:04:04] *hexmode crosses that off his list [16:04:59] <^demon|away> RoanKattouw: By the way I forked svn2git into our operations/software repo yesterday. [16:05:36] <^demon|away> (the kde version in C++ we're using, not the ruby gem version) [16:09:47] <^demon|away> Platonides: Also thought you'd like to know ^ [16:10:17] <^demon|away> I've already applied your change to use CR paths instead of useless svn paths. [16:11:26] *RoanKattouw wonders what the difference between "CR paths" and "SVN paths" is here [16:12:25] <^demon|away> The commit summaries were including "svn path=/trunk/phase3; revision=123" [16:12:36] <^demon|away> Changed that to include a link to the current path in CR. [16:12:50] <^demon|away> https://gerrit.wikimedia.org/r/#change,2332 [16:12:54] Aah [16:13:00] !g [16:13:00] --elephant-- I don't know anything about "g". You might try: !1.17wmf1 !b !class !despair !e !git-repack !hook !log !pad !rev !vectorspans [16:13:15] !g is https://gerrit.wikimedia.org/r/`e1 [16:13:15] --elephant-- Successfully added keyword: g [16:13:18] !g 2332 [16:13:18] --elephant-- https://gerrit.wikimedia.org/r/2332 [16:49:50] hexmode: How would I reproduce those JS errors on commons labs? [16:50:13] hi all [16:50:46] hi Nurbek [16:50:48] RoanKattouw: I just enabled gadgets and saw them [16:51:03] let me see if I can get you a simple test case [16:51:10] Enabled which Gadgets? Visited what page? Got what errors? [16:51:16] I need to launch some action, after some time... in mediawiki... any ideas how to realize it? [16:51:36] Nurbek: what is the action, and why does it need to be on a delay? [16:52:35] I need to send emails to existing users, after some time. [16:52:48] to notify them. [16:54:25] sumanah: can you help me? [16:54:29] any ideas? [16:54:44] Nurbek: I am fairly unskilled here and the people of #mediawiki as a whole are better suited to answer you [16:55:02] RoanKattouw: with only "Add an [edit] link for the lead section of a page." -- first gadget under "Interface: Editing and uploads" -- I get "addOnloadHook is not defined" on visiting http://commons.wikimedia.beta.wmflabs.org/wiki/Main_Page [16:55:07] ok, [16:55:32] [16:56:01] boom [16:56:08] Both addOnloadHook and importScript [16:56:23] It's dying in GalleryDetails for me [16:56:29] because I have that enabled I guess [17:00:00] hexmode: Hmm, do you have shell access to the beta labs system? [17:01:28] I want to try something [17:01:49] RoanKattouw: I do, but can I give it to you? [17:02:02] That would be nice too [17:04:26] OK, get this, it works in debug mode [17:04:34] k, you should be able to log in to "deployment-{nfs-memc,squid,web,...}" [17:04:39] This is probably a bug in ResourceLoader [17:04:42] OK, thanks [17:04:48] deployment-web is where I'd find the config files? [17:04:57] /usr/local/apache [17:05:00] yes [17:05:43] *RoanKattouw lols @ ASCII art [17:05:47] common/live under there [17:05:57] petan|wk is cute that way ;) [17:06:45] CommonSettingsDeploy.php , found it [17:07:52] hrmph, -rw-rw-r-- 1 petrb depops 377 2012-01-29 20:31 CommonSettingsDeployment.php [17:08:08] Could I be put in the depops group please? [17:08:21] oh lol, never mind, I have sudo [18:08:59] is this the right channel to ask about extending the AuthPlugin? [18:09:17] #mediawiki is probably better [18:09:42] RoanKattouw: ok. thanks [18:14:08] Nikerabbit: robla: 2min till 20% [18:14:20] make that 1min [18:14:34] *robla isn't seeing raindrift in the office [18:15:03] *hexmode wasn't getting tab completion, either ;) [18:15:12] uga [18:15:18] hi Nikerabbit [18:15:56] well....let's get started without him. it'll be the Nikerabbit show! :) [18:16:14] Nikerabbit: have you had a chance to look at 33167? [18:16:38] Nikerabbit: thank you for working on documentation during the code slush! [18:16:38] hexmode: no, I got bavk home about 4h ago and just got out of Sauna :) [18:16:44] AaronSchulz started looking at it [18:17:21] *robla pings AaronSchulz IRL [18:17:23] :) [18:17:26] yea I heard that [18:17:33] Nikerabbit: ah, didn't know how long you were traveling [18:17:50] now that I'm back, I can help him with if needed [18:17:59] *AaronSchulz is fixing CU [18:18:13] wonderful! I think that's one of the hairier blockers [18:18:15] Nikerabbit: 34246 looks simple [18:18:30] (CU=CheckUser) [18:18:41] about to ask, yeah [18:18:47] hexmode: it does, not sure why it is high priority though [18:18:57] 'Lightning Angel a modifi?? le niveau de protection de ?? Communication/pt ?? ?[move=sysop] ' ... where doth that '?' cometh from :) [18:19:09] Nikerabbit: because it is a regression for 1.19 [18:19:18] Nikerabbit: it's high priority because a lot of logs will be broken if we deploy [18:19:24] arbcom would kill us [18:19:27] hexmode: that is not mentioned inthe bug [18:19:39] I'll note it :) [18:19:43] robla: was talking about the latter bug [18:19:51] oh, sorry [18:19:53] the first bug is of course high prio [18:19:59] missed that... [18:20:09] !bug 34246 [18:20:09] --elephant-- https://bugzilla.wikimedia.org/show_bug.cgi?id=34246 [18:21:39] not high priority, not a blocker [18:21:58] and what comes to the other logging issue, Tim's hack probably is enough for the log (but then we shouldn't announce genderized logs for the 1.19 release) until I can make a better solution [18:22:08] enough for 1.19* [18:22:31] !bug 34198 [18:22:31] --elephant-- https://bugzilla.wikimedia.org/show_bug.cgi?id=34198 [18:22:39] robla: what about that one? [18:22:52] I think if it went live we would get a ton of complains [18:23:31] anything else wou would like to direct my effors this week? [18:23:52] hexmode: could you link to a repro on betalabs on the bug? [18:24:12] sure, I'll try to add that, 1s [18:26:23] hrm.... having trouble coming up with it [18:26:34] will ping the bug and lower/remove it [18:27:47] Nikerabbit: could you ask RoanKattouw what he hasn't got time to work on today? [18:28:05] I'm investigating JS breakage today [18:28:08] Let's see [18:28:09] he was doing some great work, but I'm not sure he can fix EVERYTHING [18:28:35] Nikerabbit: you've got a couple of fixmes to your name. [18:28:48] I don't have anything on my list that could be reassigned to someone who doesn't have a deep understanding of ResourceLoader [18:28:48] one of them is related to the bug that AaronSchulz is working on [18:29:06] but the other: http://www.mediawiki.org/wiki/Special:Code/MediaWiki/109144 [18:30:17] robla: I thnk Krinkle fixed another one of them [18:30:22] robla: and Tim worked around the other one [18:30:44] ok...perhaps mark them as "new" to reflect that? [18:30:50] can do [18:30:59] cool, thanks [18:31:13] alright...I'm timing out now. I'll catch raindrift later today [18:31:48] Nikerabbit: could you poke at johnduhart's fixmes and see if you can help there? [18:32:08] actually looks like Krinkle hasn't committed his fix yet [18:32:39] johnduhart's fixmes: http://www.mediawiki.org/wiki/Special:Code/MediaWiki/98500 http://www.mediawiki.org/wiki/Special:Code/MediaWiki/110772 [18:33:21] robla: but I know he found the issue, he showed it to me in Brussels [18:59:32] $protectDescription .= $wgContLang->getDirMark() . "[$action=$restrictions] ("; [18:59:48] hmm, those dirmarks end up as ? in cuc_actiontext [19:09:20] Nikerabbit: how come LogFormatter::makeUserLink doesn't use [[ ]] for plainlinks but the makePageLink function does? [19:09:51] AaronSchulz: hmm [19:10:06] is cuc doing something special with unicode? [19:10:19] not that I know of :) [19:10:40] if I var dump the value before it's inserted into the db, it's ok [19:10:49] when it's read out it's broken [19:11:25] weird [19:11:25] what comes to [[ I don't mind either way [19:11:39] they weren't there before [19:11:52] so I didn't add them now [19:11:55] hey jorm, you about? [19:12:02] i am. [19:12:06] if( $skin ) { [19:12:08] $details .= $wgLang->getDirMark() . htmlspecialchars( " {$params[1]}" ); [19:12:08] chatting with heather about mobile stuff. [19:12:09] } [19:12:10] what's up? [19:12:11] in LogPage [19:12:37] Nikerabbit: maybe dirmarks shouldn't be used in such cases [19:12:51] haha [19:13:06] unfortunetly, the offending code is in WikiPage::doUpdateRestrictions [19:13:13] werdna: knows that, i told him. [19:13:29] heatherw still needs to bug me about teahouse things too :) [19:13:44] ruh-roh. [19:13:45] jorm: so I was working on using the images for our buttons instead of CSS3 gradients [19:13:46] Nikerabbit: stuffing the localized protection descriptions in log_params is pretty suck [19:13:50] http://www.mediawiki.org/wiki/Teahouse/LayoutDevelopment [19:13:53] right. [19:13:57] AaronSchulz: indeed [19:14:00] same with cuc_actiontext [19:14:02] rob moen had a way to do that. [19:14:06] or so he said. [19:14:07] jorm: problem is, it only works if the button is exactly the right size [19:14:09] AaronSchulz: one of the reasons I haven't converted the remaining logs [19:14:11] i have to highlight the specific features [19:14:40] the problem i have with the css gradient is that it's too heavy right now. [19:14:44] when I redid CU to use a table it was unfortunately before I redid the RC table to not suck for logs [19:14:44] nod [19:14:45] AaronSchulz: another reason is that for many logs what is stored has been changed many times in the history and investigating how to parse those takes lot of time [19:14:49] there seems to be a strong drop off point. [19:14:56] and the CU table was backed on the suck RC table [19:14:57] and we want to go "flatter", as it were. [19:15:03] I noticed that the gradient you supplied was pretty light on [19:15:11] right. [19:15:11] Nikerabbit: it wouldn't be so bad if they were always serialized... [19:15:12] I had trouble telling whether or not it was working at all :) [19:15:18] thought it was an issue with my monitor [19:15:22] heh. [19:15:23] but because it was just ordered lists it's hard [19:15:32] AaronSchulz: one could get simpler code if we made conversion scripts to reformat old entries to match currect format as closely as possible [19:15:43] at some point that will be needed [19:15:51] i can always help with color choices ??? jsyk [19:16:04] I don't know anything about CSS3 gradients, but I can learn :) [19:16:43] sec. i have a site that will generate them for you. [19:17:47] Nikerabbit: much of the old log code is possibly the worst code in MW [19:17:49] http://www.colorzilla.com/gradient-editor/ [19:18:01] AaronSchulz: I wouldn't be surprised, it's very awful [19:18:33] I still remember when we just appended lines to wikipages [19:19:42] jorm: I just found it :) [19:19:57] I'm figuring out what sort of gradient we actually want [19:20:28] AaronSchulz: trying to get it work without breaking anything was much harder than I expected [19:20:52] very light. [19:21:06] the colors of those buttons - i can give you the colors. [19:21:16] both the "dark" and the "light" [19:22:01] if you could? [19:22:03] that would be awesome [19:22:12] Nikerabbit: seems more fun than WikiPage was :) [19:22:23] sure. gimme 5 minutes. [19:29:07] rebooting. [19:30:25] jorm: request accepted [19:31:16] wait. nvm. [19:31:42] i thought shit had frozen but it was just software update being stupid and preventing my trackpad from doing any gestures. [19:33:26] jorm: thanks. I think that's what Rob's implemented [19:33:32] so it still comes across a bit dramatic [19:33:39] do you think the transition is too quick, maybe? [19:34:43] Nikerabbit: I going to change LogPage::addEntry() to just wrap ManualLogEntry [19:34:47] *I'm [19:35:18] hmm, well somewhat [19:35:37] some members vars still have to be set [19:35:39] *AaronSchulz sighs [19:36:06] AaronSchulz: hmm? [19:36:31] huh, ManualLogEntry doesn't truncate the comment [19:36:45] AaronSchulz: aww crap, that was the third bug I was supposed to fix [19:36:56] it's in bugzilla somewhere [19:37:22] 33374 [19:37:40] can you add some code for that? [19:39:12] one call to $wgContLang->truncate() should fix that? [19:39:19] I can do that tomorrow mowning [19:40:29] yeah, i think the issue is that the transition is too fast. [19:41:53] ok, this is much better [19:42:06] screen shot? [19:43:11] http://werdn.us/~andrew/gradients.png [19:45:14] much better. [19:45:21] though oddly, i think the dark is too dark. [19:46:55] hmm, for the blue? [19:47:34] do you want to try something lighter? [19:54:10] jorm: also, where are we using the lightblue? [19:54:19] that's the hover state. [19:54:38] dark blue normal; light blue hover; green selected. [19:54:41] red escape hatch. [19:57:39] Nikerabbit: is r110909 ok? [19:58:41] AaronSchulz: if it is the revision I think it is, it is ok [19:59:14] AaronSchulz: see https://www.mediawiki.org/wiki/Special:Code/MediaWiki/96546#c30601 [20:00:31] We need a good query batch structure, for the renameuser changes as well [20:00:50] (e.g. using query batches rather than denormalized user name values) [20:07:19] AaronSchulz: I was thinkin, that since the parameters should have #:format:name keys... we should scan the format for 'user' and 'title' for linkbatches [20:08:48] yep [20:10:02] AaronSchulz: although that would not fix it for existing log entries [20:10:23] migration script :) [20:10:45] AaronSchulz: are you feeling eager to make one? [20:10:54] I was hoping you were [20:13:17] AaronSchulz: I'm not sure I feel confident doing that... it has to get things right the first time [20:25:14] Reedy: did you get much in the way of responses on your mail to engineering@ about extension versions? [20:25:49] Only Roan and Niklas [20:25:59] Who updated the etherpad for a couple of the extensions [20:27:08] ok...so, by default, everything should be trunk, so I'll make that change now [20:27:52] I think I already did [20:28:06] updated the script that is [20:28:26] I've left fundraising to take from 1.18wmf1 [20:29:29] What's ArticleFeedback set to? [20:29:36] from 1.18wmf1 for both [20:29:40] OK [20:29:50] I'm working on v4 so it can be pulled from trunk [20:30:24] K4-713: Where do you want the fundraising extensions to be branched from? Trunk or 1.18wmf1? (To go into 1.19wmf1) [20:31:03] Hm... good question. [20:31:39] We're still using a slightly patched version of 1.17 on the payments cluster, so I'm not sure how much it matters. [20:31:52] Don't you run a seperate branch for that? [20:33:17] Yeah, but we're still merging our extensions straight from trunk. [20:33:21] heh [20:33:26] Hmm [20:33:50] I wonder how up to date what's in 1.18wmf1 is in relation to trunk [20:34:03] I honestly have no idea what's in there. [20:34:06] heh [20:34:15] I know Kaldari did some updates to the ContributionReporting/Tracking extensions but AFAIK they're not live [20:34:23] we should just use trunk for the fundraising stuff, I'm guessing [20:34:37] Yeah, that's my feeling. [20:35:01] It's not difficult to change them for 1.18wmf1 if the need arises [20:35:04] ...wait, actually, I take it back. [20:35:23] K4-713: is there any fundraiser code that gets deployed to the main cluster? [20:35:52] I know DonationInterface does not, and it's probably less brain damage for everybody if we keep it out of the equation. [20:36:07] CentralNotice is everywhere [20:36:18] ContributionReporting is on test and foundation [20:36:27] ContributionTracking is on donate, foundation and test [20:36:49] ok...so the answer on those matters a great deal then [20:36:59] LandingCheck is on donate, test and foundation [20:37:13] DonationInterface is on test, foundation and donate [20:37:39] VariablePage was removed as being unused :) [20:37:58] Let me run a thing, and get a sense of the diff between 1.18 and trunk. [20:38:17] For those... four that aren't DonationInterface. [20:39:11] ...starting with CentralNotice. [20:41:35] Hmmm... Do we not use the MWSearch extension? :/ [20:42:45] We do... But it's not in commonsettings? :/ [20:43:07] oh, ahh [20:49:24] Reedy: Wow, I wish the people who generally do our actual cluster deploys were reachable. Everybody who knows off the top of their head what would be safe, is unavailable for comment. [20:49:45] They're all in India? [20:50:13] I actually don't know where they've gone. [20:50:26] :| [20:51:05] I didn't think the office was that big ;) [20:51:39] Plenty of hiding places. [20:53:14] Well... probably the safest thing to do, would be to take whatever we've M'd FT to the current deploy branch. [20:54:20] We have this nasty habit of cherry-picking changes to deploy. I have no idea how long it's been on a lot of these since we've done a clean cut-over. [20:55:02] jesus [20:55:16] 1.18wmf1 to trunk of centralnotice makes a 2.49MB patch [20:55:24] o_O [20:55:33] <^demon|away> Reedy: w/ i18n? [20:55:44] yeah [20:55:49] There is a lot of that, yes. [20:56:01] probably simpler to remove them [20:58:05] 1 meg smaller [20:59:23] ContributionReporting should be fine to take from trunk. [21:07:03] Oh, svn files [21:07:06] how we hate you [21:07:55] heh. :) [21:08:05] it's actually not too bad [21:08:23] So, Reedy: You know how I said we don't deploy anything from DonationInterface to the cluster? [21:08:25] Lies. [21:08:33] haha [21:08:43] We need the donationinterface_langonly.php, and all the i18n. [21:09:48] https://www.strongspace.com/reedy/public/fr.7z [21:10:04] contribution tracking has almost nothing changed [21:10:14] Addition of a title setting and an i18n alias file update [21:10:26] AaronSchulz: in https://www.mediawiki.org/wiki/Special:Code/MediaWiki/110897 ... don't you think it's stupid to convert it to array first, when logentry converts it back to object immediately? [21:10:27] LandingCheck very similarily [21:10:41] Yeah, that's about what I expected. Ish. [21:10:57] ContributionReporting has nothing scary too [21:11:19] hmm, the centralnotice diff looks useless [21:11:36] it seems to suggest lots of things have been deleted :/ [21:12:32] Yeah, I saw a couple revs that seemed to be killing deprecated code. [21:12:55] So that might be correct then [21:12:59] ie nearly at trunk anyway [21:13:34] I'll go with that [21:13:55] brb, drink [21:16:22] donation interface has a 10 meg diff file :D [21:18:55] K4-713: if that's the case with DI, it would seem weird that we can't just pull it from trunk, cause we're still essentially only going to be using all the language files, and trunk should be more upto date [21:19:24] Yeah, absolutely. [21:19:59] Which means it's just features being arkward now ;)( [21:30:32] *werdna waves [21:30:38] how goes, Reedy? [21:46:12] robla: only two blockers left for the triage \o/ [21:46:45] hey jorm, are we still going to send people to Requested Articles as an option on the ACW page? [21:46:58] *hexmode goes to look for bugs against AFTv5 [21:47:05] that's a howie call. [21:47:13] Our current buttons are wizard, draft and create [21:47:18] for anons, it should be in there. [21:47:22] but not for logged ins. [21:47:56] hexmode: Which two? [21:48:19] RoanKattouw: 33989 and 34057 [21:48:33] !b 33989 [21:48:33] --elephant-- https://bugzilla.wikimedia.org/show_bug.cgi?id=33989 [21:48:33] I think 34057 is just a placeholder [21:48:35] !b 34057 [21:48:35] --elephant-- https://bugzilla.wikimedia.org/show_bug.cgi?id=34057 [21:55:42] Hey Reedy: I just finished triple-checking: Everything in trunk, is good to go in 1.19, for CentralNotice, ContributionTracking, ContributionReporting, and LandingCheck. [21:55:42] hashar: around? [21:56:10] K4-713: Great, thanks! :) [21:56:11] hexmode: more or less 11pm and heading bed soon [21:56:20] hexmode: but go ahead ) [21:56:29] hashar: what is the status of https://bugzilla.wikimedia.org/show_bug.cgi?id=34057 [21:56:36] you have it marked as a blocker [21:57:18] Reedy: np. Thanks for asking. [21:58:41] hashar: ?? [21:59:14] hexmode: still pending review [21:59:40]