[05:51:48] https://www.mediawiki.org/wiki/Project:Current_issues#Forbidden_error_when_making_donation_36362 [06:03:17] p858snake|l: Replied. [06:30:30] PHP fatal error in /usr/local/apache/common-local/php-1.23wmf4/extensions/EducationProgram/includes/Events/Event.php line 53: [06:30:30] Argument 6 passed to EducationProgram\Events\Event::__construct() must be an array, boolean given, called in /usr/local/apache/common-local/php-1.23wmf4/extensions/EducationProgram/includes/Events/EventStore.php on line 147 and defined [06:30:36] Whoops... [06:31:10] [[Special:MyCourses]] on en.wp gives me error message... [06:33:10] JeroenDeDauw: ^ [06:39:56] Could anyone review this change: https://gerrit.wikimedia.org/r/#/c/99042/2 it backport bugs fixed in ProofreadPage [06:40:37] Reedy? [06:43:20] Tpt_: it's just a submodule commit update, so there's not much to review; it just needs to be deployed [06:48:31] ori-I: Yes, I have used the wrong word, sorry [09:38:29] ori-l: I've not worked on this thing for quite a while, not sure what is happening there [14:45:23] Hi, Does someone knows why we don't support opendocument files (like ODP for exemple) on commons? [14:50:55] Kelson: I reply myself: https://bugzilla.wikimedia.org/show_bug.cgi?id=2089 [15:11:27] :D [15:11:35] the answer is "at this point, nobody knows" [15:24:00] Kelson: ther is no need to support ODP on commons. [15:24:16] Steinsplitter: I need it [15:24:44] for wath? pleas upload a .pdf [15:24:45] Steinsplitter: if you look the number of people who follow this bug, you will see that I'm not alone [15:24:50] + i see no communety consensus [15:24:58] Steinsplitter: PDF can not be reused/modified. [15:26:44] Steinsplitter: do you see any downside of allowing this? [15:26:55] yes. spam. [15:27:12] Steinsplitter: what do yo mean exactly with spam? [15:27:29] too many "text" formats ar bad. [15:28:16] Steinsplitter: what do you mean with text? SVG is a text format for you? [15:28:31] Steinsplitter: ODP is text format? [16:48:06] perhaps someone could solve this bug some day? :) https://bugzilla.wikimedia.org/show_bug.cgi?id=27807 [16:48:10] it's open for almost three years now [16:48:37] it shouldn't be too hard to do as I read it correctly [16:50:08] Steinsplitter: ODP is very useful for Open/LibreOffice Impress presentation documents [16:50:43] to share the original instead of only the PDF representation [16:54:44] is gitblit's replica of core broken again or just slow? I don't yet see the last patch of https://gerrit.wikimedia.org/r/#/c/99000/ (5 min ago only though) [17:03:56] this one can be closed as not resolved/won't fix since there is no consensus for this change [17:03:57] https://bugzilla.wikimedia.org/show_bug.cgi?id=37626 [17:04:01] but I don't think I can do this [17:04:53] Trijn|away: There are plenty of ops who've commented on it - so I wouldn't touch it. You can always ask Andre though. [17:05:01] It's his bugzilla. ;-) [17:06:04] oh [17:10:16] maybe I can do it [17:10:17] * Trijn|away checks [17:11:21] yes, I can :D [17:13:19] closed another one too :) [17:52:49] Krinkle: I've updated https://gerrit.wikimedia.org/r/#/c/98869/ to remove loading mediawiki.ui.button by default. Could you rereview it? The downside is that this will probably result in a fair bit of double-loading depending on how wikis choose to load it by default. [18:28:56] ori-l: btw I did go to ganglia to check if wdsearch had had any effect on it.wiki's performance, but it.wiki was not included in any graph [18:33:38] Nemo_bis: it will not be reflected in ganglia, because the impact would be almost entirely client-side [18:34:09] hm? isn't it in the graphs on loaded JS size [18:34:30] yeah, i haven't expanded those to all wikis yet [18:35:05] csteipp: hey [18:35:25] csteipp: do you think that https://gerrit.wikimedia.org/r/96317 (Remove https redirect from CentralAutoLogin) should/could be backported to wmf5? [18:35:38] i don't mean to be prickly, but do you really need a graph to be satisfied that a request to retrieve a script from a different domain has a cost? [18:36:35] i'm not asking for it to be disabled; the change i asked for would prevent it from loading on pages where it wouldn't be activated anyway [18:37:36] maybe i'm looking at it the wrong way and you were looking for the graph as an aid for informing others, in which case sorry for the tone of my reply [18:38:56] paravoid: Did that not get into wmf%? [18:39:06] 5, that is [18:39:25] If not, yeah, we can bump wmf5's version of CentralAuth to master [18:40:21] ori-l: I'm not the maintainer of the tool nor of it.wiki's JavaScript; there are excellent people in charge of both and I leave such considerations to them [18:40:51] kk, sorry [18:41:17] i'm a bit annoyed because i pointed out the load-on-every-page issue a couple of nights ago and it was ignored [18:41:21] but not your fault [18:42:19] I don't think anything is on fire (any longer) [18:42:52] moreover, if you press "page up" a sufficient amount of times you'll see that concern was actually discussed here [18:49:10] csteipp: not afaik [18:49:12] the other one did [18:49:41] csteipp: https://gerrit.wikimedia.org/r/#/c/95609/ did, https://gerrit.wikimedia.org/r/#/c/96317/ did not, again, aiui [18:49:59] got to go now [18:50:00] bye [19:26:21] guillom: ^ [19:26:56] See #wikimedia-operations for more in depth discussion [20:15:09] Reedy: Hi! About https://bugzilla.wikimedia.org/show_bug.cgi?id=57616 I have made a clean fix in core https://gerrit.wikimedia.org/r/#/c/97835 [20:15:27] that is ever waiting for review [20:34:20] i hear leonard cohen and immediately think siebrand... [20:34:31] (he's playing on the radio now...) [20:34:40] jeremyb: eh? [20:36:53] 31 15:53:35 <+siebrand> jeremyb: you mean: First he takes Berlin, then he takes Manhattan? :) [20:37:53] jeremyb: heh :) Last time I was in Manhattan was in 1986. [20:37:59] wow [20:38:23] anyway, i have a strong association between you and leonard [21:05:34] Reedy: Hi! [21:07:16] Reedy: I've some bug fix for ProofreadPage to backport (I hope the last ones if you except the fatal error that have a pending fix in core). Will you accept to deploy them if I make the change? [21:09:17] Image scalers appear to be back [21:10:37] Tpt_: how fixed do you changes make it? [21:11:12] Tpt_: I'd prefer, if it isn't a total fix, to revert back to the previous version on production, and let you continue to fix it on the Beta Cluster, eg: http://en.wikipedia.beta.wmflabs.org/wiki/Main_Page [21:12:01] greg-g: It's a total fix if you except a fatal error that happens sometimes [21:12:18] heh, in what situation does the fatal happen? [21:12:51] When people plays with the contentModel parameter of the edit form [21:13:12] A clean fix for it is pending in core: https://gerrit.wikimedia.org/r/#/c/97835/6 [21:13:41] with a related change in ProofreadPage: https://gerrit.wikimedia.org/r/#/c/97835/ [21:14:37] Tpt_: gotcha [21:15:12] sorry, i've copy/paste twice the same link https://gerrit.wikimedia.org/r/#/c/97837/ [21:15:35] Tpt_: can you prepare the backports? to 1.23wmf5 and 1.23wfm6 [21:15:56] greg-g: ok [21:28:38] greg-g: core changes: https://gerrit.wikimedia.org/r/#/c/99518/2 and https://gerrit.wikimedia.org/r/#/c/99519/ [21:32:05] anyone able to answer a question about a possible change in bugzilla email subject format change? [21:32:16] Reedy: can you review Tpt_'s changes to fix proofreadpage? [21:35:13] any email list admins around? [21:35:50] greg-g: In the worst case, if the change looks good to you, I can +2 [21:36:05] Betacommand: you probably want andre__ for that [21:37:16] valhallasw: do you know if it its possible to change the format of email subject for new bugs in bugzilla? [21:37:43] I don't, but andre__ is our king of bugzilla [21:37:47] Tpt_: for the extension I trust you, I'll look at the core change [21:37:56] ok [21:38:04] (but now that you mention, I'd really like the [NEW] in the subject to go away) [21:38:30] valhallasw, [NEW]? there is only a New: for me. [21:38:43] Tpt_: it looks reasonable, but I'm not a dev :0 [21:38:44] :) [21:38:46] andre__: that was just from the top of my head :-) [21:38:59] valhallasw: exactly :P [21:39:02] it makes the mails not-group in gmail, which is a bit annoying. [21:39:22] it spawns two conversations [21:40:09] valhallasw, GMail could use mail headers for threading which have existed for 40 years or so... [21:40:30] (just a naive idea, why do so if Google could ignore standards) [21:40:43] andre__: because this is much more user-friendly for 99% of conversations. [21:40:55] to ignore mail headers?! How? [21:41:07] if it has a different subject, it *probably* shouldn't be grouped under the same conversation, even if the mail headers say so [21:41:43] well, most of the email hardcore people disagree :) [21:41:43] so that's working around those folks who clickiclick any email reply, remove subject and content, to "start a new thread"? [21:41:57] "No, I won't fix your broken GMail." [21:42:00] I hate those people [21:42:04] http://www.jwz.org/doc/threading.html [21:42:14] everyone should just implement jwz threading [21:42:14] Which is, basically, everyone who is not a hardcore emailer. [21:42:14] I hate people in general. Not just those, everybody! [21:42:31] * andre__ wonders if latest Bugzilla 4.4 had implemented something for Gmail [21:42:35] vaguely remembering [21:42:45] people should use real e-mail clients [21:42:52] mutt4ever [21:43:18] Yeah, let's stop supporting IE on wikipedia, too. People should use real browsers. [21:43:25] valhallasw, alright, once we upgrade to 4.4 this should be fixed: https://bugzilla.mozilla.org/show_bug.cgi?id=663747 [21:43:34] CC yourself to https://bugzilla.wikimedia.org/show_bug.cgi?id=49597 :) [21:44:16] greg-g: So, do you prefer to wait? [21:44:27] Tpt_: I want someone to look at the core change, yeah [21:44:32] Reedy: around? [21:45:11] Betacommand: See the discussion above -- I also added you as CC on the bug. [21:47:23] Sort of [21:48:57] Reedy: so, proofreadpage. we either need to review and deploy tpt's changes, or revert it and wait for them to be reviewed later. [21:50:36] Based on the error logs it's mostly not broken [21:51:08] yes, the error happens only when people plays with hidden fields of the edit form [21:51:20] valhallasw: so the GMail problem with bugmail should get fixed within the next two months or so :) [21:51:23] 1 line in the last 50,000 lines of the apache syslog [21:52:45] Backport fix for the other bugs that affect Wikisource is, I think, more important: https://gerrit.wikimedia.org/r/#/c/99522/ [21:53:18] Reedy: then I misunderstood andre and the bug reports' concerns [21:53:45] greg-g: AFAIK the update hashar did to the cluster earlier today fixed up most of the issues [21:53:59] Possibly a couple of bugs that should've been closed as resolved fixed [21:54:23] oh nice, if it's only one line that's great [21:54:26] Reedy there are some remaining issues that I've fixed this afternoon [21:54:30] https://gerrit.wikimedia.org/r/#/c/99522/ [21:54:35] Reedy: My problem sometimes with your quick error log pasting into bug reports is that I have no idea how often things happen. So it's hard for me to judge [21:54:46] Reedy: ahh, missed that completely [21:56:52] andre__: great :-) [21:57:07] Reedy: This backport includes: https://gerrit.wikimedia.org/r/#/c/99415/ https://gerrit.wikimedia.org/r/#/c/99412/ https://gerrit.wikimedia.org/r/#/c/99373/ https://gerrit.wikimedia.org/r/#/c/99386/ and https://gerrit.wikimedia.org/r/#/c/99379/ [21:58:19] Is it possible to deploy it? [22:05:10] greg-g: Could you deploy https://gerrit.wikimedia.org/r/#/c/99522/ , please? It fix most of the remaining issues with ProofreadPage (except the one that depends on the core change) [23:21:13] any wikitech mailing list admins around? [23:21:51] Thehelpfulone: ^ [23:22:53] Betacommand: mail wikitech-l-owner@lists.wikimedia.org [23:23:14] MatmaRex: irc faster :P [23:23:26] not really if there's no one around [23:28:22] <^d> Someone needs a list admin? [23:28:45] ^d: yeah [23:38:57] Who can I bother about irc.wikimedia.org stuff, namely blocking broken noisy bots on one channel? [23:38:59] see https://bugzilla.wikimedia.org/show_bug.cgi?id=54821 [23:40:56] file a bug [23:40:59] oh wait! [23:41:22] #wikimedia-ops, maybe? (that's 'ops' for 'irc ops') [23:41:44] MatmaRex: no. its an irc.wm.o issue [23:41:47] not freenode [23:42:01] what lego said [23:42:09] we have an irc network? :o [23:42:12] and, I'm not sure it's that big of an issue, they can just ignore joins/parts [23:42:15] :/ [23:42:24] well, i dunno, /kickban? [23:42:30] MatmaRex: it's announce only, of all edits, per project, eg: #en.wikipedia [23:43:35] yes it's not a big deal. still there might be somebody who could fix that? :P [23:43:39] <^d> I thought nobody could talk on irc.wikimedia.org. [23:43:40] well, I've also asked in ops [23:43:44] <^d> How is a bot being noisy? [23:43:49] join and quit [23:43:53] all of the time [23:44:02] well. see the bug report :P [23:51:02] is ignoring it absolutely not an option? [23:53:57] ^d, andre__ : I also thought irc.wm.o is read-only so to say [23:54:35] ah, quit and join [23:54:43] likely it is read-only. The bot joins and quits on every message [23:54:46] <^d> andre__: Anyway, that requires a root. Nobody else has access to it afaik (the config file is in the private puppet repo, looks like) [23:54:55] <^d> Actually, it quits on every message. [23:54:58] <^d> Then rejoins :p [23:55:07] <^d> Then quits on a new message. [23:55:11] Errm, "nobody else" than? [23:55:20] * andre__ only half-awake [23:55:52] <^d> andre__: I'm saying to fix your bug--you'll need a root opsen. I don't think it's anything we mortals can handle. [23:56:03] ah. alright. thanks :) [23:56:31] who is an IRCop there? "/stats o" anyone? [23:57:07] <^d> Nobody as far as I know. [23:57:11] ridiculous, trying "/stats o" I get "Permission Denied - You're not an IRC operator" [23:57:15] <^d> Since it's read-only, who needs ops? :p [23:57:24] heh :D [23:58:16] alright, thanks for explaining another part of our IRC empire to me. Appreciated (seriously)