[20:30:10] hexmode: is there a bug triage happening at in 30 min? [20:42:59] robla: no.... I should've announced, that, though. [20:43:40] k....also, should probably delete the invite in calendar. [20:43:58] will do [20:44:27] I'm going to be cleaning up the stuff I said I would do on the 1st of August [20:45:54] robla: maintainer of proofread extension (which is basically core to WikiSource) has said he doesn't have time to maintain it [20:46:10] I'm putting out a call for devs on wikitech-l [20:46:15] k [20:46:19] anything else I should do? [20:46:50] what's the state of 1.18 blockers? [20:47:02] 1s [20:47:27] first: something to be aware of https://bugzilla.wikimedia.org/show_bug.cgi?id=29246&list_id=20333 [20:48:00] since the image thumbnail problem is apparently gone according to reports from #wikimedia-commons [20:48:53] blockers: 3 left for deployment [20:49:28] I think one of them, thumbnails, is fixed. [20:49:44] evening [20:49:56] is there triage tonight? [20:50:20] Bryan: no [20:50:45] Bryan: but, do you have any interest in mobile or revdel? [20:51:03] *hexmode tries to lure a developer [20:51:08] in neither [20:51:28] but I don't have time for things that require commitment anyway [20:51:31] :P [20:51:37] sure... k [20:54:59] robla: Bryan: ^demon: Nikerabbit: werdna: TimStarling: got time to go over 1.18 tarball blockers? [20:55:23] since robla reminded me of them :P [20:55:37] tarball blockers? [20:55:43] y [20:55:50] not deploy blockers? [20:55:52] we did the deployment ones [20:56:02] I'm just looking at these tarball ones [20:56:14] and thinking maybe some shouldn't be there [20:56:26] when was the deploy blocker review? [20:56:44] 3-4 weeks ago? I'd have to look. [20:56:49] ummmmm [20:58:33] there are only 3 blocking right now, and I think they should be... [20:58:39] others are fixed. [20:58:49] and of the 3, I think one is fixed [20:59:33] let's talk about the individual issues [21:00:03] robla: http://etherpad.wikimedia.org/BugTriage-2011-06 has the deployment triage [21:00:20] robla: for deployment or tarball? [21:00:33] or both? [21:00:34] deployment please [21:00:35] :) [21:00:57] bug: https://bugzilla.wikimedia.org/show_bug.cgi?id=29068 [21:01:14] 28613 is the thumbnail problem [21:01:23] which I *think* is resolved [21:01:55] https://bugzilla.wikimedia.org/show_bug.cgi?id=29170 - "[[MediaWiki:Enotif body]] needs GENDER support" [21:02:20] I'll talk to Nikerabbit about that one. [21:03:09] the other: https://bugzilla.wikimedia.org/show_bug.cgi?id=21279 -- "Regular deletion of revisions deleted with rev_deleted breaks links in log entries" [21:03:22] I think I need to talk to brion on that one [21:03:33] I doubt Brion is the right assignee for that one [21:03:43] who should be? [21:03:51] I don't know anything about revdel, really [21:04:29] AaronSchulz may be a good choice. I just asked him to join us [21:04:57] k [21:05:37] robla: https://bugzilla.wikimedia.org/show_bug.cgi?id=28613#c69 comment may be something to look at if the thumbnail problem comes back [21:05:58] but it mentions https://bugzilla.wikimedia.org/show_bug.cgi?id=30192 [21:06:13] which I don't think should be a blocker [21:06:23] but let me ask in -commons [21:07:28] *hexmode asks [21:08:20] saibo, who updated the thumbnail bug for me last night is looking at it right now [21:08:25] AaronSchulz: we were discussing bug 21279 [21:08:30] !b 21279 [21:08:30] --elephant-- https://bugzilla.wikimedia.org/show_bug.cgi?id=21279 [21:08:54] <^demon> https://bugzilla.wikimedia.org/show_bug.cgi?id=29170 shouldn't be a blocker IMO. [21:09:42] so...what's needed for 21279? a change to deletedcontribs? [21:09:50] <^demon> That whole enotif thing needs refactoring anyway [21:09:55] ^demon: k. I'll remove it if Nikerabbit agrees and thinks it would take too long [21:10:06] ^demon: file a bug! [21:10:18] <^demon> There is a bug [21:10:22] <^demon> And it's marked as a dependency [21:10:31] :P [21:10:41] <^demon> 22769 is the blocker to it. [21:10:48] <^demon> Which I fixed years ago, but Brion reverted me [21:12:29] ^demon, revert brion, it's not 3RR yet [21:12:41] :) [21:13:10] so...which one are we discussing here? [21:13:30] hexmode: ^ [21:14:03] robla: ^demon was talking about the enotify one [21:14:39] I don't even know what's needed for bug 21279 [21:15:01] AaronSchulz: could you look? do you know who could look? [21:15:29] AaronSchulz: perhaps what we should do is for you and Brion to talk about 21279 when he's back from Haifa, and figure out who is going to take it [21:16:54] sounds good to me [21:17:16] and I think ^demon should revert brion :) [21:17:26] for the Enotif bug, that one is only a blocker if Nikerabbit agrees, right? [21:17:38] Yeah, I think so [21:17:51] but if he says no, then it isn't [21:17:56] <^demon> brion only reverted me because I didn't retain back-compat. [21:18:14] and I wish if we said Nikerabbit's nick often enough he would show up :) [21:18:20] hi [21:18:43] sputnick: hey [21:19:16] is there a good place to ask for dev extensions questions ? [21:19:19] hexmode: you believe bug 28613 is fixed? [21:19:23] *is it* [21:19:32] sputnick: #mediawiki is better [21:19:36] sputnick: #mediawiki would be better [21:19:37] :) [21:19:40] ok [21:20:07] robla: after talking to apergos and saibo in commons, yes [21:20:29] excellent. please mark it resolved then [21:21:00] k... was leaving it open to get more feedback but closing it works for that, too. [21:21:07] :) [21:21:08] hexmode: for the other two issues, can you write comments reflecting the result of this converation? [21:21:22] yep [21:21:47] do you have time to talk about tarball blockers? [21:24:19] a little bit [21:24:23] :) [21:24:57] tarball list: https://bugzilla.wikimedia.org/show_bug.cgi?id=28425 [21:25:38] ok, show/hide in hist is totally broken [21:25:39] k [21:25:55] You have either not specified a target revision(s) to perform this function, the specified revision does not exist, or you are attempting to hide the current revision. [21:26:01] wtf [21:26:22] ugh [21:26:34] so that def seems like a blocker, no? [21:26:44] robla: https://bugzilla.wikimedia.org/show_bug.cgi?id=15472: Feed diff-content should be hidden for users without the "read" user right [21:28:06] seems like that one is going to rot, shouldn't block till it is fixed [21:28:28] that's marked as low priority. I'm betting it's only a blocker because it used to be "highest" for some weird reason [21:28:59] 25985 - When $wgSpamRegex is triggered be sure to return to editing and not throw the users work away [21:29:25] not sure it is a regression [21:29:32] !b 25985 [21:29:32] --elephant-- https://bugzilla.wikimedia.org/show_bug.cgi?id=25985 [21:30:54] not a blocker [21:31:10] 27132 - Uncomment movefile right in default settings [21:31:17] !b 27132 [21:31:17] --elephant-- https://bugzilla.wikimedia.org/show_bug.cgi?id=27132 [21:31:34] ugh [21:31:37] not a blocker [21:31:47] That's also stupidly easy [21:31:51] It's more a case of when, than if [21:32:03] unless Bryan hasa time [21:32:10] which he doesn't [21:32:45] but you do have my blessings [21:33:07] how long ago was it commented out? [21:33:13] When it was added, probably [21:33:21] running on the cluster? [21:33:32] for sysops on commons etc [21:33:34] <^demon> For sysops only, iirc. [21:33:39] Which is patently stupid. [21:33:40] Bryan: blessing to uncomment? [21:33:40] It seems to be "stable" enough [21:33:44] But that's a separate isue. [21:33:45] issue [21:33:46] yeah [21:34:13] It was enabled by Brion for admins only, if that gives you an idea of when it was. [21:34:29] it was originally not very stable, so we limited it to sysops because we thought that they would be better able to report problems [21:34:42] might as well make the stupidly simple change so that we don't have to discuss it again [21:34:45] Now an entire bureaucracy has built up around the user right. [21:34:53] Separate user groups, etc. [21:35:03] Because moving a file is vastly different than moving a wiki page... [21:35:31] robla++ [21:35:38] 28083 - Load MediaWiki:Filepage.css only when it exists [21:35:44] !b 28083 [21:35:44] --elephant-- https://bugzilla.wikimedia.org/show_bug.cgi?id=28083 [21:35:56] though I just realized the bureaucracy will probably mobilize if we change it at this point, so it's probably not going to save us time, but whatevs [21:36:17] Just override it in CommonSettings.php [21:36:26] Easily sorted for WMF till we get lolconensus [21:36:36] <^demon> s/lol/local/ ? [21:36:48] ^demon: he probably meant lol [21:36:51] No, I meant lol [21:37:53] any thoughts on 28083? [21:39:18] should it block 1.18, that is? [21:40:00] I don't think so. [21:40:11] Ditto [21:40:27] It should be fixed, though. [21:40:40] Myra, bugs have a habit of being like that [21:40:55] lolconsensus reached [21:40:58] 28863 - Remove references to code deprecated before 1.17 in 1.18 [21:41:05] seems like it should be [21:41:11] !b 28863 [21:41:11] --elephant-- https://bugzilla.wikimedia.org/show_bug.cgi?id=28863 [21:41:13] wontfix [21:41:35] It's not a blocker, but I'm not quite sure it's a wontfix etiher [21:41:52] later maybe [21:41:59] <^demon> It's a "as we get to it." [21:42:01] k [21:42:03] <^demon> There's no rush. [21:42:07] 29102 - Upgrade fails "Unknown character set: 'mysql4" [21:42:17] !b 29102 [21:42:17] --elephant-- https://bugzilla.wikimedia.org/show_bug.cgi?id=29102 [21:42:22] <^demon> I thought we discussed that one weeks ago. [21:42:27] only on upgrading old bugs [21:42:29] <^demon> Yes it's a bug, but somewhat of an edge case. [21:42:33] hexmode: please mark 28863 as normal (or low) priority [21:42:40] will do [21:42:45] when done here [21:43:11] ^demon: right we talked about it and right, not a blocker [21:43:48] workaround is to upgrade to 1.16 + MySQL5 first? [21:43:49] 29232 - Let Http / MWHttpRequest handle redirects safely on any CURL version [21:43:53] https://bugzilla.wikimedia.org/show_bug.cgi?id=29232 [21:43:59] robla: yeah, think so [21:44:41] should really be fixed, but won't hurt to stay unfixed for another release [21:44:50] I think I could fix 29232, but, then, I'm not sure how much we care about older curl versions like brion mentioned [21:45:24] How much older? [21:45:39] 2009 [21:45:49] quite recent imho [21:45:50] Can we use a sane levelling point like our min php version? [21:45:59] 2 years is ancient history [21:46:07] Reedy: agreed [21:46:23] http://svn.wikimedia.org/viewvc/mediawiki/trunk/phase3/includes/Wiki.php?r1=91941&r2=91942& [21:46:26] ie is it fixed in curl from the time when 5.2.3 (?) was being pushed out [21:46:28] I think that totally broke revdelete [21:46:32] anyone want to check that right off? [21:46:43] at least from history buttons [21:46:52] AaronSchulz: can you fix it, then? [21:47:02] *hexmode is glad AaronSchulz found it [21:47:50] 29246 - API errors occasionally with unknown error 231 [21:47:56] !b 29246 [21:47:56] --elephant-- https://bugzilla.wikimedia.org/show_bug.cgi?id=29246 [21:48:02] related one today [21:48:07] let me get the num [21:48:24] what was the branch point? [21:48:30] !1.18 [21:48:30] --elephant-- I don't know anything about "1.18". [21:48:39] !branchpoints [21:48:39] --elephant-- I don't know anything about "branchpoints". [21:48:39] !b 30201 [21:48:39] --elephant-- https://bugzilla.wikimedia.org/show_bug.cgi?id=30201 [21:48:41] !branchpoint [21:48:41] --elephant-- I don't know anything about "branchpoint". [21:48:45] heh [21:48:55] *Bryan kicks elephant2 [21:49:02] anyway [21:49:10] i'll leave it at this for today [21:49:10] related in that it was api and errors [21:49:12] good night [21:49:18] Bryan: tyvm [21:49:47] I've been getting a lot of HTTP 504 errors with the API. [21:49:48] Reedy: can you tackle 29246? [21:49:53] hexmode: might be a dupe of 30086 btw [21:49:58] Myra: recently? [21:50:02] sounds like an ops problem [21:50:05] anyway [21:50:07] night [21:50:13] night! [21:50:17] Yeah, I doubt it's a MW problem. [21:50:21] hexmode: Yes, recently. [21:50:32] I get e-mails from cron. [21:51:02] Myra: any chance you could find when the first one showed up? [21:51:19] robla: what's the branch point? [21:51:47] AaronSchulz: http://www.mediawiki.org/wiki/Special:Code/MediaWiki/92475 [21:52:06] sigh [21:52:29] sigh? [21:52:44] hexmode: Looks like I've been getting them since sometime in 2010? [21:52:51] :P [21:52:59] Earliest e-mail I have is March 2010. [21:53:13] ok, so not a *recent* issue [21:53:18] at least yours isn't [21:53:29] maybe it is happening even more often now [21:53:31] It's hard to say whehther they've picked up in frequency lately. [21:53:43] I dunno. It's been going on for a while now. I've mostly been ignoring it. [21:53:49] It'd be great if it stopped happening, though! [21:53:50] I'm guessing that people at Wikimania are learning about all sorts of cool new toys, so I wouldn't be surprised if it's getting hit a little harder [21:54:10] (new to them, at least) [21:54:12] I may just need to revert some crap [21:54:21] http://www.mediawiki.org/wiki/Special:Code/MediaWiki/91949 [21:54:22] robla: so today that might be an issue [21:54:40] but 30083 was last month [21:55:19] ^demon: 29246 you have time for that? [21:55:23] !b 30083 [21:55:23] --elephant-- https://bugzilla.wikimedia.org/show_bug.cgi?id=30083 [21:55:26] er [21:55:36] roan, not ^demon [21:55:44] <^demon> I was going to say... [21:55:44] or maybe reedy [21:55:46] <^demon> Ermmm, API? [21:55:47] <^demon> :p [21:55:58] If the 504 errors are increasing in frequency.. it's something we need to find out where shizz is breaking [21:56:15] oh....the new uploader hits the api a lot harder, I imagine [21:56:31] Hello. Does !b 26170 belong to the component/extension indicated? - I'm modifying it and in passing I could correct that. [21:56:41] !b 26170 [21:56:41] --elephant-- https://bugzilla.wikimedia.org/show_bug.cgi?id=26170 [21:56:53] <^demon> MarcoAurelio: Yes, it's correct. [21:56:59] ktnx [21:57:05] !b 30086 [21:57:05] --elephant-- https://bugzilla.wikimedia.org/show_bug.cgi?id=30086 [21:57:10] why toes tsvn reverse merge do nothing? [21:57:19] that one, robla, not *83 [21:58:44] oh, nv [21:59:08] 29408 - Upload fails and unable to move existing .edm files without extension: "Key 'something.' is not in a proper format" [21:59:13] !b 29408 [21:59:13] --elephant-- https://bugzilla.wikimedia.org/show_bug.cgi?id=29408 [21:59:14] yay, conflict [21:59:37] hexmode: what did you decide on 30086? [21:59:45] Can someone assign me the the api 504/whatever bugs, and i'll have a look tomorrow [22:00:18] Reedy: sure thing [22:00:29] is there 2 bugs for api related errors? [22:00:37] ie the more constant ones [22:01:00] "more constant"? [22:01:08] happening more often [22:01:22] you mean not the intermittant one w/ json reports? [22:01:34] I think there may be [22:02:27] if so, I'll give 'em both to you , Reedy [22:02:51] Yeah, if they're more site/ops esk issues, they are probably a bit more high priority [22:03:01] Reedy: is 30086 api stuff? [22:03:23] It's possible it is, yeah [22:03:27] or just upload/infrastructure? [22:03:52] It could be both [22:04:26] robla: I decided to let Reedy take point... I was just pointing it out b/c someone mentioned it as an api error [22:04:43] it wouldn't harm to see if there are any useful errors about [22:04:59] I can ask on -commons [22:05:09] they're my new buddies :) [22:05:38] hexmode: it would probably be a good idea to make neilk_ aware of the upload errors, even if Reedy is the one investigating [22:05:58] yes, I think he reads all the bugs though [22:06:05] *hexmode makes sure component is right [22:06:17] i did log a bug saying uploadwizard was slow, so he should be in the loop [22:06:28] cc him on the 30086 probably anyway [22:06:36] :( [22:06:44] not right component [22:07:17] whwat sort of errors should I ask about in -commons? [22:07:36] just any errors? [22:09:48] robla, hexmode -- I'm aware that people are complaining a lot :) [22:10:11] :) [22:11:01] ready for the next one... [22:11:08] only a couple more [22:11:13] 29408 - Upload fails and unable to move existing .edm files without extension: "Key 'something.' is not in a proper format" [22:11:18] !b 29408 [22:11:18] --elephant-- https://bugzilla.wikimedia.org/show_bug.cgi?id=29408 [22:12:23] last comment on bug: "Attempting to move already-uploaded .dem files, without changing the .dem extension, causes the error "The new file extension does not match its type" to appear (and the file doesn't move)." [22:13:50] seems like it shouldn't block tarball if the other part is already solved [22:14:17] yup [22:14:58] well...hm [22:15:54] yeah, not a blocker [22:15:55] hm? [22:16:01] k :) [22:16:07] 29847 - Regression: &action=unwatch is broken with JS off [22:16:13] !b 29847 [22:16:13] --elephant-- https://bugzilla.wikimedia.org/show_bug.cgi?id=29847 [22:17:36] seems like it is only an issue if enotifwatchlist is on [22:18:49] robla: ^demon: any opinion if that should block tarball? [22:19:19] <^demon> If it's actually a regression and we can repro it. [22:19:44] AaronSchulz: were you able to repro 29847? [22:20:19] not with the watch/unwatch star [22:20:30] the only link I can think of that would break was in emails [22:20:37] but I changed the code to show a form in that case [22:20:42] :) [22:20:44] (the case of no valid token) [22:20:50] by my reading, both Aaron and Brion tried it, so hexmode, you may want to take a run at it before marking "worksforme" [22:20:51] ok, so it is fixed? [22:21:23] I'll verify the emailed link works, right? [22:21:52] r92012 [22:21:56] it should be fine now [22:22:09] !r 92012 [22:22:09] --elephant-- http://www.mediawiki.org/wiki/Special:Code/MediaWiki/92012 [22:22:38] oh! and pre-branch! [22:22:42] :) [22:23:12] btw, while we were doing this [22:23:35] Saibo verified that we never delete thumbnails on commons: https://bugzilla.wikimedia.org/show_bug.cgi?id=30192 [22:24:16] robla: that's a minor privacy/copyright concern -- you have to know the url of the old thumbnail [22:24:27] robla: should it block deploy? [22:26:36] yeah, let's put it there [22:27:44] k [22:28:00] I'll make these updates later [22:28:07] *hexmode saves the log [22:28:13] tyvm everyone [22:30:31] $wgRemoveGroups['bureaucrat'] = array( 'bot', 'accountcreator', 'ipblock-exempt', 'reviewer', 'sysop') for enwiki [22:55:59] <^demon> robla: Are we having our team meeting with people in Haifa? [22:56:38] hi....I'd moved that meeting earlier this afternoon (sorry for waiting until last minute). [22:56:58] ^demon: I'm guessing you haven't synced your tablet this afternoon :) [22:57:16] <^demon> Phone still buzzed me just now. [22:57:28] <^demon> Switching timezones seems to have confused the heck out of my calendar. [22:57:46] I postponed until Thursday of next week [22:57:51] <^demon> Gotcha. [23:00:04] oh.....there's a duplicate left over from the GCal migration [23:00:06] drat