[06:13:33] any devs around? https://bugzilla.wikimedia.org/show_bug.cgi?id=32077 [07:56:11] omg chunked uploading is as slow as death [11:47:07] <^demon> MaxSem: I was going to create your PageImages repo. Remind me...did you have existing history you wanted to import? [11:47:48] ^demon, https://github.com/MaxSem/PageImages [11:47:59] <^demon> Ah ok. I'll do that now [11:52:40] <^demon> https://gerrit.wikimedia.org/r/gitweb?p=mediawiki/extensions/PageImages.git;a=shortlog :) [11:52:40] <^demon> Done [11:53:45] woohoo [11:53:51] thanks, ^demon [11:53:57] <^demon> You're welcome. [11:59:17] <^demon> JeroenDeDauw: I created your DidYouKnow extension and imported the existing history. [12:00:39] ^demon: awesome, thnx! :) [12:00:50] ^demon: any progress on fixing the Maps repo? [12:02:05] <^demon> I think I figured out why at least :) [12:03:09] <^demon> Ah, that's why http://www.mediawiki.org/wiki/Special:Code/MediaWiki/84104 [12:03:34] <^demon> Ok, let's see what we can do here. [14:21:06] ^demon: yeah... I figured it was because of that [14:21:13] ^demon: so is it fixed now? [14:22:17] ^demon: any particular reason not to prefer !$array over empty( $array )? They are pretty equivalent as far as I can tell (the later just does not throw a warning when the var is undefined) [14:23:23] <^demon> Not yet. [14:24:04] <^demon> Per the coding conventions: "empty() is inverted conversion to boolean with error suppression. Only use it when you really want to suppress errors. Otherwise just use !. Do not use it to test if an array is empty, unless you simultaneously want to check if the variable is unset." [14:25:04] <^demon> If you know the array exists, you might as well just use $arr or !$arr. [14:25:26] <^demon> And then you don't have to worry about mistakes if you make a typo in the variable name :) [16:39:45] ^demon|away: ok that makes sense. I'm rather sad that there is no decent (non-type conversion magic) way to check if an array is empty [16:39:53] Sort of tempts me to do $array === array() [16:40:02] No magic there and it's obvious what it does [16:40:07] But is a bit ugly :/ [16:42:28] JeroenDeDauw: if it is obvious and valid, then uses it :-] [16:42:49] $array === array() <--- that one looks fine to me [16:43:05] better than ( is_array( $array ) && is_empty( $array ) ) [16:43:29] or forging something like wfIsEmptyArray(); [16:44:00] I think === array() works [16:44:04] * RoanKattouw tests [16:44:43] Yup works fine [16:45:51] RoanKattouw: how much of a hassle would it be to get that giant aft refactoring commit reviewed and merged? i've got my work on the central feedback log ready to go, but it depends on two different outstanding commits on gerrit (one of mike's and one of matthias's), and i'm beginning to think going off the master merge would be way, way cleaner [16:46:13] and by "master merge" i mean "master branch" [16:46:18] typing fail [16:46:33] rsterbin: Link? [16:46:59] https://gerrit.wikimedia.org/r/#/q/status:open+project:mediawiki/extensions/ArticleFeedbackv5,n,z -- full list [16:47:15] https://gerrit.wikimedia.org/r/#/c/6319/ -- refactoring monster [16:47:43] hashar: wfIsEmptyArray seems more evil then === array() [16:47:52] totally [16:48:46] https://gerrit.wikimedia.org/r/#/c/6532/ and https://gerrit.wikimedia.org/r/#/c/6422/ -- also needed for central feedback log [17:18:01] hey hashar, you're just starting your day, right? [17:18:14] sumanah: kind of :) [17:18:22] hashar: was your trip OK? [17:18:40] meeting right now sorry :( [17:18:52] tty in half an hour or so ? [17:19:06] no prob [18:16:59] rsterbin: Re gerrit 6319, looks like Mathias wiped out some of your work in patchset 4, is that right? [18:17:08] oy. [18:17:13] i have no idea what he did in there. [18:17:42] hm, is there a way to compare one patchset to another? [18:18:56] sort of, rsterbin [18:19:18] rsterbin: Tim wrote about it in wikitech-l a few months back [18:20:00] how could i find it? [18:20:02] rsterbin: subj: [Wikitech-l] Changeset differ (was Re: Committing followups: please no --amend) [18:20:07] http://tstarling.com/gerrit-differ.php [18:21:32] "Error: Sorry, only Wikimedia changes are supported at the moment" [18:22:17] <^demon> Probably because urls changed a bunch in 2.3 [18:22:21] rsterbin: You can do it from the interface [18:22:25] provided he didn't rebase things [18:22:28] <^demon> If Tim's doing any kind of regex on that url it will have broken. [18:22:35] Find the dropdown that says "Base" and set it to "Patchset 1" or something [18:24:11] RoanKattouw: one of the changes he made is correct, the other is bad [18:24:18] blargh. [18:24:37] OK [18:24:43] Didn't he also wipe out your changes? [18:24:51] If I diff all against patchset 1, that's what it looks like [18:25:29] <^demon> Heh, http://25.media.tumblr.com/tumblr_m301gyi8gm1rqvy12o1_500.jpg [18:25:47] * AaronSchulz is always afraid to click obscure image urls [18:26:13] <^demon> It's safe :) [18:26:24] AaronSchulz are you afraid to click obscure bugzilla urls? [18:26:56] Oooooh [18:26:59] * RoanKattouw hugs gerrit 2.3 [18:26:59] * AaronSchulz wonders whether he can trust chad's words [18:27:08] Look at the dependencies section in https://gerrit.wikimedia.org/r/#/c/6422/ [18:27:12] https://bugzilla.wikimedia.org/show_bug.cgi?id=34041 [18:27:15] It marks the dependency as red and [OUTDATED] [18:27:16] especially with that smiley face [18:27:29] * AaronSchulz clicked [18:27:33] RoanKattouw: the dependency bits i don't know about [18:27:56] rsterbin: It's OK, I know how to deal with that one [18:28:18] But 6422 depends on 6319 which is the one that looks like Mathias messed it up, so you'll need to fix 6319 before I can merge 6422 [18:28:42] 6532 also depends on 6319 but at least it depends on the correct version :) [18:28:46] ok, let me check in with him and see what he was trying to do [18:28:49] AaronSchulz the bug or image [18:29:03] * ToAruShiroiNeko depends on devs [18:29:13] does that make me a bug? [18:29:46] rsterbin: I have a meeting in 30 mins so if you don't mind I'm gonna postpone reviewing these revisions until you've disentangled this [18:29:53] ok [18:30:38] <^demon> RoanKattouw: +10 for actually showing that to the user. -1000 for the glaring red which hurts mah eyes. [18:30:58] ^demon: It SHOULD hurt your eyes ;) [18:31:06] Broken dependencies are bad things [18:31:37] <^demon> There's a difference between "This is an error, highlight it" and "Color it like a 2-year-old would" [18:31:58] haha [18:32:02] fair enough [18:32:08] Get Timo to fix it [18:32:31] <^demon> Anyone can reskin gerrit and I'll be thrilled to test/review it. [18:32:32] <^demon> :) [18:32:38] <^demon> I showed you how with gerrit.config [20:53:44] RoanKattouw: hey. do i need to have you around to supervise this deployment? [20:53:57] I'm gonna run up to R63 (?) in a bit [20:54:05] sounds like a plan. i'll move in there. [20:54:19] I think it'd be good for me to supervise you all the way through actually getting stuff in production [20:54:31] yeah, that seems prudent [20:54:59] Given that this is like your 3rd deployment already, you can do it yourself from there on out [21:10:38] New patchset: Diederik; "Better debian packaging support, documentation:" [analytics/udp-filters] (master) - https://gerrit.wikimedia.org/r/6871 [21:12:16] New review: Diederik; "Ok." [analytics/udp-filters] (master); V: 0 C: 2; - https://gerrit.wikimedia.org/r/6871 [21:12:35] New review: Diederik; "Ok." [analytics/udp-filters] (master); V: 1 C: 2; - https://gerrit.wikimedia.org/r/6871 [21:12:37] Change merged: Diederik; [analytics/udp-filters] (master) - https://gerrit.wikimedia.org/r/6871 [21:21:27] raindrift1: http://test.wikipedia.org/w/api.php?action=query&meta=userinfo&uiprop=groups|rights [21:25:02] RoanKattouw: I'm flattered ;p [21:25:19] wait, nevermind [21:25:25] of course it's showing my account info, I'm logged in. guh. [21:29:27] what's the name for the special page this week? [21:32:13] Special:NewPagesFeed, I think [21:34:45] That's right [21:34:49] Currently only on testwiki [21:34:56] Ian is looking at a strange autopatrol-related bug now [21:38:46] neat [21:39:03] gerrit question: is there a way to drop a patchset? [21:43:45] <^demon> rsterbin: Click "Abandon" [21:44:07] ^demon: that abandons the whole changeset, or just the patch set? [21:44:22] <^demon> The whole change. There's no way to just remove "patch N" from a change. [21:44:34] hm [21:44:48] ok, i guess i'll have to undo it line by line. awesome. [21:50:06] rsterbin: No you don't [21:50:23] rsterbin: You can fetch the older change from Gerrit into your git clone, and then you can use it for ... things [21:50:42] If you would like me to restore change 6319 to an older patchset I can totally do that [21:50:52] patch set 3 is the right one [21:51:00] When I said "use it for things" I guess I meant resubmit it [21:51:06] OK, sweet, so patchset 3 without any modifications? [21:51:09] !g 6319 [21:51:11] i confirmed with matthias that 4 was a mistake [21:51:17] * RoanKattouw waits for bot [21:51:18] !gerrit 6319 [21:51:24] !change 6319 [21:51:42] !b [21:51:46] Come on bot [21:51:57] New patchset: Diederik; "Improved debian packaging, documentation" [analytics/udp-filters] (master) - https://gerrit.wikimedia.org/r/6878 [21:52:55] New review: Diederik; "Ok." [analytics/udp-filters] (master); V: 1 C: 2; - https://gerrit.wikimedia.org/r/6878 [21:52:56] Change merged: Diederik; [analytics/udp-filters] (master) - https://gerrit.wikimedia.org/r/6878 [21:53:45] RoanKattouw -- is this what you need? https://gerrit.wikimedia.org/r/#/c/6319 [21:53:55] I already got it [21:53:58] I was just being lazy [21:54:03] :) [21:54:04] And wondering why the bot doesn't seem to be working [21:55:07] rsterbin: OK patchset 5 restores patchset 4 [21:55:09] *3 [21:55:35] sweet [21:57:49] <^demon> RoanKattouw: https://gerrit.wikimedia.org/r/#/c/6880/ :D [21:58:17] <^demon> Look at the status and lack of Verified/Reviewed fields :) [21:58:45] <^demon> Oh whoops, I pressed publish. [21:58:46] <^demon> Crap. [21:59:19] ^demon: ?!? [21:59:30] <^demon> I had published a draft and I wanted to share it :p [21:59:33] OK [21:59:36] <^demon> But I pressed publish by accident. [21:59:37] Try again maybe? [21:59:39] <^demon> So it became a normal commit. [22:00:55] <^demon> RoanKattouw: https://gerrit.wikimedia.org/r/#/c/6881/ [22:01:07] <^demon> They're hidden unless you add someone as a reviewer :) [22:01:27] ha! [22:01:34] So private drafts even [22:01:42] <^demon> Yep :) [22:02:16] <^demon> I think that's awesome for security fixes. [22:02:22] It doesn't show any of the CR/VR stuff [22:02:31] <^demon> Nope, not until you press "Publish" and it goes live. [22:02:36] <^demon> When it's in draft, you just leave comments. [22:02:37] https://s3.amazonaws.com/mrdoob/lab/javascript/effects/solitaire/index.html [22:20:40] RoanKattouw: raindrift1 PHP Warning: require_once(/usr/local/apache/common-local/php-1.20wmf2/extensions/PageTriage/PageTriage.php) [function.require-o [22:20:40] nce]: failed to open stream: Permission denied in /usr/local/apache/common-local/wmf-config/CommonSettings.php on line 2488 [22:20:49] PHP Fatal error: require_once() [function.require]: Failed opening required '/usr/local/apache/common-local/php-1.20wmf2/extensi [22:20:49] ons/PageTriage/PageTriage.php' (include_path='/usr/local/apache/common-local/php-1.20wmf2/extensions/OggHandler/PEAR/File_Ogg:/usr/local/apache/common-local/php-1.20wmf [22:20:49] 2:/usr/local/apache/common-local/php-1.20wmf2/lib:/usr/local/lib/php:/usr/share/php') in /usr/local/apache/common-local/wmf-config/CommonSettings.php on line 2488 [22:20:49] 2 Segmentation fault (11) [22:22:08] Yeah looking [22:22:41] It's synced wrong [22:22:49] Its PageTriage directory is chmod 700 but that's not what it's supposed to be [22:22:52] Lemme resync that machine [22:23:19] Crap, lots of them are reporting this [22:23:31] Fixing it with dsh [22:24:01] mm [22:24:28] OK looks like it's fixed now [22:24:32] This may be a bug in sync-dir, maybe? [22:24:40] Or the sync was half-done and got interrupted, or ... whatever [22:24:44] mmm [22:25:49] Hrmph, so tail -F doesn't actually work [22:26:23] tail -f works fine for me.. [22:26:38] Well the file is dropped and recreated by logrotate [22:27:29] RoanKattouw: got all the issues sorted? is it safe for me to start mucking around on fenari again? [22:27:44] awjr: All yours, go ahead [22:27:54] thanks! [22:28:04] My apologies for all the confusing [22:28:11] *confusion [22:32:21] heh likewise :p [22:38:21] Deploy all the codes! [22:59:46] raindrift1: http://pastebin.com/5ZByA9J4 [22:59:53] raindrift1: PageTriage is breaking things ---^^ [23:00:13] Apparently occurs when trying to block an IP [23:04:04] Looks like it is still inserting if count( $pageIds ) === 0 [23:05:37] hrm. this isn't working for me at all. [23:05:48] filters aren't setting, nor is the scrollbar. [23:05:59] jorm: between sessions, or at all? [23:06:05] AT ALL. [23:06:08] woah [23:06:09] urp caps. [23:06:15] okay, they set for me but don't save between sessions [23:06:44] * Ironholds is busy a-filin' bugs [23:07:59] https://gerrit.wikimedia.org/r/6886 [23:09:54] RoanKattouw: Even though I got that error, the block went through - but it isn't in my logs, or their block log [23:10:23] Yeah that sounds likely [23:10:33] ok [23:10:38] hrm. maybe it's only showing articles that got added after deployment. [23:10:51] Reedy: Did you deploy that fix yet? [23:11:03] my machine is being slow at merging [23:14:31] OK no rush [23:14:33] Was just wondering [23:15:03] just pushing to site now [23:15:37] OK [23:19:59] It's seemingly still broken [23:22:40] Replicated on test [23:24:14] RoanKattouw: are there usually this many bugs? [23:24:25] The last attempted database query was: [23:24:25] "INSERT INTO pagetriage_tags (ptrt_tag_name, ptrt_tag_desc) [23:24:32] from within function "DatabaseBase::sourceFile( /var/www/wiki/mediawiki/extensions/PageTriage/sql/PageTriageTags.sql )". [23:24:32] Database returned error "1146: Table 'wikidb.pagetriage_tags' doesn't exist (192.168.0.212)" [23:24:32] I thought we were going to allow this to be tested before showing it to people, Ironholds? [23:24:40] this type of stuff being why. [23:24:53] jorm: the idea is many eyes. I'm showing it to a very small number of people (well, 3) [23:25:04] so far they've identified five obvious enhancements and three cringe-inducing bugs [23:25:12] yes. well. [23:25:13] Ewww [23:25:21] Something did happen while adding the tables [23:25:23] DB was funky [23:25:27] Let me add that table [23:25:30] jorm: are they breaking it by looking at it? [23:25:32] Reedy: Wait, wikidb? [23:25:36] Locally [23:25:39] O [23:25:40] No index on the table [23:25:40] h [23:25:43] the point was to avoid making a bad first impression. [23:25:43] or, no prefix [23:25:44] They may not have added updaters [23:25:56] Oh, no prefix [23:25:58] Ok [23:26:00] nm then [23:26:02] jorm: it's three people, all of whom are close friends o'mine [23:26:09] well, I say friends. One is an ex. But we are on speaking terms. [23:28:45] Can I disable pagetriage on enwiki for the time being? [23:37:46] Reedy: Yeah go ahead [23:37:53] It's fixed now [23:38:07] you fixed the issue? whee! You I owe booze [23:38:11] I didn't update fenari properly [23:38:14] The blocking issue, yeah [23:38:24] in the meantime my impromptu tiger team of wastrels (tm) has identified lots of glorious work for raindrift1 and bsitu :P [23:49:07] remind me, what was decided on committing release notes? [23:51:56] nothing about it in [[Git/Workflow]] [23:56:00] <^demon> TimStarling: I don't think we ever really decided. We've just been dealing with conflicts as they arise. [23:56:35] so commit release notes normally then? [23:56:47] <^demon> Yep.