[01:15:00] hmm [01:15:12] could the double redirect detection script be updated perhaps? [01:15:40] I'd like it to report self redirects and protected pages [14:07:56] @search Local [14:07:56] Results (Found 13): analytics, config, enableapi, errors, extranamespace, javascript, logo, memory, newusergroup, nofollow, subpages, tidyforidiots, vector, [14:08:07] @regsearch Local [14:08:07] Results (Found 13): analytics, config, enableapi, errors, extranamespace, javascript, logo, memory, newusergroup, nofollow, subpages, tidyforidiots, vector, [14:08:17] @regsearch Local|local [14:08:17] Results (Found 14): analytics, config, enableapi, errors, extranamespace, fileurl, javascript, logo, memory, newusergroup, nofollow, subpages, tidyforidiots, vector, [14:08:34] @regsearch LocalSettings [14:08:34] Results (Found 13): analytics, config, enableapi, errors, extranamespace, javascript, logo, memory, newusergroup, nofollow, subpages, tidyforidiots, vector, [14:08:40] weird [14:08:41] !ls [14:08:41] All configuration is done in LocalSettings.php (near the end of the file). Editing other files means modifying the software. Default settings are not in LocalSettings.php, you can look in DefaultSettings.php. See , , , and [14:08:46] aha [14:08:51] it's alias [15:57:49] are job queue statistics available somewhere? [16:15:03] can a dev please explain autotranslate to me? [16:17:27] also how can I disable the ? [16:17:31] "add topic" link [16:35:54] MaxSem: what sort of "statistics"? [16:36:08] MaxSem: on fenari: mwscript showJobs.php enwiki [16:36:33] Reedy, how large it is, how many items get processed per unit of time [16:36:35] thanks [16:37:06] We have http://ganglia.wikimedia.org/latest/graph_all_periods.php?c=Miscellaneous%20pmtpa&h=spence.wikimedia.org&v=51&m=enwiki_JobQueue_length&r=hour&z=small&jr=&js=&st=1339605412&z=large [16:37:47] MaxSem: https://gdash.wikimedia.org/dashboards/jobq/ might be more useful [16:38:10] 47-1143 a minute, average 507 [16:38:32] cool [16:48:33] ohhh [16:48:41] Reedy: did not know about that dashboard :-] [16:48:51] hashar: that's the pretty frontend [16:48:57] there's a complex one behind it.. [16:48:58] errrm [16:49:07] yeah graphite [16:49:21] wasn't aware of the newbie friendly frontend [16:49:21] https://graphite.wikimedia.org/dashboard/ [16:49:22] yeah [16:49:34] it's rather nice [16:50:50] * sumanah tries "guest" and "guest" - no [16:51:33] sumanah: labs account :-] [16:51:40] see you later, daughter duty [16:51:46] sumanah: hi -- you were looking for me? (~1am here in +800) [16:52:07] Hi there au! I was trying to find out what your email address is in Bugzilla so I can alert you to bugs you should know about [16:52:29] Sorry about the timezone mismatch, au. [16:52:36] sumanah: yep, its the labs user [16:52:41] ah. [16:52:49] np at all. my address is audreyt@audreyt.org as usual, though I don't recall having set up a bz account [16:53:07] (I did get the code review request and +1'd it, though.) [16:53:26] au: it'd be useful if you did, in my view, because sometimes people ask for help involving the parser, or submit patches via Bugzilla attachments [16:53:32] Thanks for the code review, au [16:54:24] sure thing. bz account created [16:54:44] please feel free to assign things my way and I'll do my best to respond timely. [16:55:06] au: sure, and often it won't be an *assignment* - just a cc [16:55:13] *nod* [16:55:31] <- feels quite a bit liberated working as a volunteer now (instead of via odesk) :) [16:55:47] au: ah, you're no longer a WMF contractor? good to know!~ [16:56:04] strangely, this now means I can ask you for MORE, since it cannot be perceived as a command ;-) [16:56:07] I actually didn't get notice on it, just a sudden termination of odesk contract [16:56:14] exactly my point ;-) [16:56:15] uh, what? that's weird, to me [16:56:22] au: did you ask why this happened? [16:56:36] or did you get a reason of any kind? [16:56:48] no, and no [16:57:01] I kinda assumed it was my 1.5month of absolute inactivity following svn->git (due to external circumstances) [16:57:20] but anyways, I'd like to keep volunteering as is [16:57:24] ok [16:57:38] have you been able to submit even a test commit into gerrit, au? [16:57:48] sure [16:57:54] did that right after the switch [16:58:13] (d3602bb on March 28) [16:58:49] ah, that's good to know. also, au, would you be comfortable looking at old patches against the parser & against the ParserFunctions extension, to see whether any of them are worth following up on, and nicely saying "sorry the revamp makes this obsolete but please do join us as we work on new stuff" to authors whose work is now obsolete and sadly useless? [16:58:57] au, there are at least 25 such cases [16:59:19] where should I look for them? [16:59:20] au: we were remiss in not responding to them. I think Gabriel meant to but didn't have time [16:59:27] au: I can give you a search, just a moment. [16:59:38] *nod* [17:00:08] 35 bugs found, au. https://bugzilla.wikimedia.org/buglist.cgi?keywords=patch-need-review%2C%20patch%2C%20&keywords_type=allwords&list_id=122727&resolution=---&resolution=LATER&resolution=DUPLICATE&query_format=advanced&component=Parser&component=ParserFunctions&product=MediaWiki&product=MediaWiki%20extensions Does that search work? [17:00:41] subbu: you might be interested in the recent backscroll here [17:01:13] sumanah: 27 (minus the reopened ones) [17:01:35] Good morning/evening. [17:01:39] au: not sure why we'd exclude reopened ones. Maybe I'm missing something? [17:01:41] Hi Dereckson [17:01:43] sumanah, ok. reading. [17:02:10] er sorry, thinko. [17:02:14] minus the duplicate ones [17:02:19] but including the reopened ones [17:02:34] saved those 27 as FollowUpWithSorryTheRevampMakesThisObsoleteButPleaseDoJoinUs [17:02:42] will triage tomorrow. [17:02:53] * sumanah involuntarily laughs aloud at Java-esque name :) [17:03:05] :D [17:03:14] tfinc: sooner today, we got a farsi Wikipedia contributor who wanted to submit some extension translations ; it appeared it's the source code of mwlib involved. [17:03:54] au: Some of the ones that are marked duplicate will also need looking-after, but it can certainly wait until after the 27 non-dupes. I appreciate your time. Thanks. [17:04:06] tfinc: could you ask PediaPress if they could update language files on our servers? You've the details on https://bugzilla.wikimedia.org/show_bug.cgi?id=37544 [17:04:54] (oh Reedy already cc you in this bug, sorry for the double post irc/bugzilla) [17:04:55] ori-livneh: hi, wanted to ask whether you have any SignupAPI-related questions or concerns I could possibly help with [17:05:53] sumanah, is this because all this will be supplanted by the new VE and parser? [17:06:24] subbu: I think the sequence of events that happened was: [17:06:48] 1) MediaWiki had bugs and some of them attracted bug reports and some of those attracted patches [17:07:03] sumanah: ok, I assumed "resolved duplicate" was an end-state. it looks like all but one of the is duplicate of a closed bug; I'll follow up with the one that's left open ( 21994, dupe of 7264 ) . [17:08:02] au: OK, I may have mis-searched! Usually "resolved dupe" is an end state, yes. Go ahead and follow your judgment; after your pass through the swath of patches & bugs, I'll look again. [17:08:13] wilco! [17:08:28] subbu: 2) Senior developers did not have/make time to review patches against many components, including the parser [17:08:58] 3) we decided to overhaul the parser, so I was unsure which of the parser & Extension:ParserFunctions patches might still be applicable in the new VE/Parsoid era [17:09:47] (this step led me to wait in a holding pattern for a few months) [17:09:48] 4) Now that au is a volunteer and we're nearing completion of the first parser deliverable, I can safely ask au to judge what to do with those 25-30 patches [17:10:04] subbu: I may have overexplained there, but does that help? [17:10:12] okay. yes, that clarifies things. [17:10:14] thanks :) [17:11:49] subbu: We had a peak, at one time, of more than 200 patches in BZ awaiting review, languishing. [17:11:58] although if there are any critical fixes/patches, they should probably go in .. au might know better. [17:12:48] subbu: I helped get developers to review and respond to many of those patches, especially the ones affecting MediaWiki core and extensions that WMF deploys. But I held off, mostly, re the parser, because it seemed like the relevant experts did not have time to do code review and response [17:13:00] ah, that is a lot. the parsing code is hairy overall from what i can tell, just looking at the scenarios parsoid has to handle. [17:13:35] subbu: we have, now, 101 bugs against MW core and 57 against WMF-deployed extensions, that have patches awaiting code review, in BZ [17:14:04] although in about 20 of those cases, the patch is actually in Gerrit, which is better [17:15:48] subbu: this is one reason why we instituted the "20% time" policy https://www.mediawiki.org/wiki/Wikimedia_engineering_20%25_policy and why I am glad that we moved to Git and Gerrit; at least when any volunteer, even without Subversion access, can submit a change into Git and Gerrit, we have less of this second-class citizenship problem [17:15:55] and everything is in the same dashboard. [17:15:59] i am not in a position yet to comment on what to do about those patches, but i am happy to help once i learn more. [17:16:27] subbu: as of a few days ago, our Gerrit backlog is also not doing so great: [17:16:28] Current statistics on all MediaWiki (core and extensions): [17:16:29] 30 that have received a positive tentative review (+1) but have not been merged (+2) [17:16:29] 216 that received neither -2, -1, +1, or +2 reviews (but might have textual comments) [17:16:29] 38 received a negative tentative review (-1) with issue to be addressed by the original contributor [17:16:29] 6 that have been rejected (-2) but not yet abandoned by their original authors [17:17:12] subbu: my understanding is that you're working full-tilt on hitting the VE deadline in a few weeks. but once you have some breathing room (and more experience!) I shall certainly talk with you about help :-) [17:17:35] ok :) sounds good. [17:17:40] hey Nikerabbit got a moment to talk? [17:19:51] sumanah, thanks for helping me with understanding some of the context as well. [17:20:16] subbu: I'm very glad to! Sometimes I forget how much of this sort of context is transparent to me but opaque to someone newer [17:20:20] Thanks for the reminder [17:20:21] :) [17:21:31] sumanah: I guess so [17:21:42] RoanKattouw: Looks like Chad abandoned the various test/example extension commits so you don't have to. [17:21:58] sumanah: Ah, I thought I abandoned those and merged one [17:22:09] RoanKattouw: maybe I misread who did it [17:22:12] At least the ones in test/extensions/examples [17:22:27] Nikerabbit: I hope I did this rightly; I saw that various i18n-related commits were languishing in Gerrit so I added the l10n team as reviewers [17:22:43] Nikerabbit: hope it doesn't seem like a command to review, just a reminder/invitation [17:23:01] Nikerabbit: in case these commits had not come across your team's radar. [17:24:07] sumanah: dunno know about others, but I follow core and some extensions anyway [17:24:46] Nikerabbit: ok. But I figure, since you are SPQRobin's GSoC mentor, your first priority in code review, when you have time to do code review, is his code, right? [17:26:09] sumanah: yeah, or even commenting before submission [17:26:15] Nikerabbit: right [17:26:46] Nikerabbit: I see he has a bunch of stuff open right now: https://gerrit.wikimedia.org/r/#/q/owner:SPQRobin+status:open,n,z [17:27:21] obviously since nobody else dares to review that stuff and I have been on vacation [17:27:55] Nikerabbit: do you have a backup mentor? [17:28:34] sumanah: not an official one, but I asked Amir and Santhosh to be available while I'm away [17:28:57] SPQRobin: ^ just to make triply sure you know that you can talk to santhosh & aharoni when Niklas is away [17:29:28] well, I'm not away anymore... until Wikimania [17:29:36] Nikerabbit: "bus factor" :-) [17:29:45] Nikerabbit: sometimes people have medical issues, win the lottery, whatever. [17:29:58] or miss the flight... [17:30:10] right [17:30:44] Nikerabbit: I saw SPQRobin had a question about whether to store a particular list of information (the list of languages, in some format, I forget) in his ext or in core. SPQRobin did you get a chance to decide that? [17:30:52] sumanah / Nikerabbit: yeah, I know I can talk to Amir or Santhosh if needed [17:31:13] SPQRobin: ok, good to know. And even if you don't need. :-) [17:31:36] ok, meeting time [17:33:11] I'll probably put it in core, but I haven't committed it yet [17:33:19] * sumanah needs to buzz off [18:47:02] The bug #37225 has been assigned a Highest/Major priority/importance, as it "affects a number of users and impedes their [18:47:05] use of the site" on "Wikimedia's servers in production". [18:47:47] Thanks Dereckson [18:47:50] Reedy: robla ^^ [18:48:04] !b 37225 [18:48:04] https://bugzilla.wikimedia.org/show_bug.cgi?id=37225 [18:48:17] I'm just looking, it's apparently not new [18:48:50] "Several history entries for the same content" [18:49:04] If Tim is right, the code has been changed since Jan 2011 [18:49:16] What in the world? then why is it just causing a problem now? [18:51:16] * robla looks [18:51:40] Not sure if it's Highest/Major [18:53:05] It qualifies to the http://www.mediawiki.org/wiki/Bug_management/How_to_triage#Highest definition, isn't it? [18:53:23] Dunno. Can't say I've ever read that [18:53:26] Well... I started with the assertion "the history integrity and tidyness is a core functionnality" [18:53:41] Sure, but it's not missing data [18:53:46] It's excess/duplicate [18:56:36] so...it appears as though 37225 and http://en.wikipedia.org/wiki/Wikipedia:Village_pump_(technical)#Diff.2Fwatchlist_difficulties may be related, but maybe not the same thing [19:01:23] I chatted with a few recent changes patrollers people. [19:01:36] They are annoyed because they have to control diffs who aren't modifications, so waste some time. [19:04:42] but +- half of them would find acceptable to stay in this situation one more week ; one of them addd "if it doesn't take 6 months to fix that because it wouldn't be important enough" [19:15:24] * robla pokes through http://www.mediawiki.org/wiki/MediaWiki_1.20/wmf3/raw [19:18:21] robla: is that a script you got to generate change logs ? :-D [19:18:32] looks great [19:19:04] yeah, it's checked in /mediawiki/tools/release.git [19:19:10] I am happy to see the git commit guidelines are well respected ( http://www.mediawiki.org/wiki/Git/Commit_message_guidelines ) [19:19:14] "make-deploy-notes" [19:20:33] * robla can't find a69fb4827409786203ee6f4befc5aa4009d13ef8 in gerrit [19:24:34] How the heck do I view an extension on gitweb? [19:25:33] * robla disappears in a few minutes [19:25:49] StevenW: typically using a web browser :) [19:26:02] Yeah but where is the link? [19:26:29] so...one thing that really, really sucks about gerrit right now is the visibility of the projects link [19:26:44] super seekrit link: https://gerrit.wikimedia.org/r/#/admin/projects/ [19:26:56] Thanks! [19:27:35] and then....go into branches, and then the tiny "gitweb" link to the right of the branch you're interested in (probably "master") [19:40:50] I usually just take the gitweb for core and modify it [19:41:23] someday gerrit will sucka lot less. or we won't use it. either way it will be a glorious day [19:41:40] (I hate navigating around to find anything in there. really anything.) [19:41:53] The last update gave navigation back from gitweb to the gerrit interface [19:41:54] that's useful [19:43:00] slowly and surely [19:43:06] and at least it's under active development [19:44:12] StevenW|lunch: https://gerrit.wikimedia.org/r/gitweb?p=mediawiki/extension/ExtensionName.git;a=summary [19:45:15] <^demon> apergos: Too bad we couldn't find a solution that worked 100% off the shelf with zero bugs :) [19:46:19] you know I can live with it having bugs [19:46:26] I would just like it to suck a lot less [19:46:40] or at least not to take away existing functionality [19:46:45] [19:47:00] <^demon> It's only java and they're super friendly to contributors ;-) [19:47:17] sweet. who do we know that writes java? :-P [19:47:23] (answer: notme) [19:47:31] <^demon> Anyone who's been through at least one "Intro to programming" course :p [19:48:02] not me. I have taken exactly one programming course in my life [19:48:09] it was the second half of an assembly language course [19:48:18] (tested out of first half) [19:48:25] that was um [19:48:48] 30 years ago :-D [19:48:52] 30+ [19:48:56] <^demon> Well, filing bugs is always good. It's actively developed, so getting stuff done isn't impossible. [19:49:17] <^demon> At least it's a project nobody's committed to in 6 months and nobody's seriously maintained it for 2+ years. [19:49:26] <^demon> *it's not [19:49:30] hahaha [19:49:39] <^demon> Negatives are hard. [19:49:52] how about double negatives? [20:01:37] ;-) [20:01:51] apergos: please do open bug about gerrit ;-:D [20:02:07] one I recently opened got fixed! [20:02:13] (the ability to abandon someone else change) [20:04:34] ok maybe I'll keep a running list of things I hate and go file a batch of them at once [20:04:59] <^demon> apergos: http://code.google.com/p/gerrit/issues/list is their tracker :) [20:05:15] tab opened [20:05:28] hm let's see how quickly ff eats through my extra 4gb these days [22:50:50] ori-livneh: Comment: "2048-bit RSA, converted by Olivneh@WMF from OpenSSH" [22:51:01] ori-livneh: I'd like to actually get it in OpenSSH format if you don't mind [22:51:50] RoanKattouw: sure, sec [22:54:06] RoanKattouw: updated the page [23:09:08] easiest bug ever: https://bugzilla.wikimedia.org/show_bug.cgi?id=37573 [23:19:29] hey, I get to poke devs like they poke me! [23:19:32] https://gerrit.wikimedia.org/r/#/c/11177/ [23:19:49] RoanKattouw, Reedy, etc, etc: ^^ [23:20:03] and: https://gerrit.wikimedia.org/r/#/c/11175/ [23:20:14] and https://gerrit.wikimedia.org/r/#/c/11169/ [23:20:27] and https://gerrit.wikimedia.org/r/#/c/10560/ [23:22:35] review me! :) [23:28:00] I've got a merge conflict in a .gitreview file [23:28:18] Reedy: in mine? [23:28:30] I don't think so, just updating my checkouts [23:29:03] you put a −1 on 11177. was that the reason? [23:29:24] nope [23:29:30] you need to document your new hook :p [23:30:53] damn [23:30:56] forgot about that [23:31:42] heh [23:34:40] Reedy: ok, amended. [23:36:12] Reedy: have I mentioned that this is the fourth place in the code base that you need to extend to do the same thing? [23:36:22] of course all four places are done totally different ways [23:36:40] Sounds about right [23:36:53] *none* of them use HTMLForm [23:37:23] at least one of them was written almost a year after HTMLForm, and re-does one portion of what HTMLForm was originally written for [23:37:32] <3 [23:37:59] and my solution is to add another hack ontop of it. yay mediawiki. [23:38:22] of course the alternative is to refactor all of the auth code, and the fuck if I'm gonna do that [23:38:54] ffs [23:40:20] maybe I can get someone else to takeover my other projects for a couple months and I can refactor all of the auth code [23:41:32] Looks like some of these checkout update scripts we have dont do that [23:41:47] don't do what? [23:42:14] update the checkout [23:51:54] mediawiki-1.19.0-IIS.zip | Bin 0 -> 22349161 bytes [23:51:56] wtf [23:54:38] apparently the maximum size of an IPv6 block on enwiki is still /64 [23:54:59] and people have been asking for days "how do I block a school now?" [23:55:01] Shouldn't that have changed with the recent deployment? [23:55:05] AaronSchulz: ---^^ [23:55:23] it's in 1.20wmf5 I think [23:55:30] Yeah, your change is [23:55:36] Are many schools using IPv6 already? [23:55:38] Aaron's too presumably [23:55:50] I did the research and wrote the commit before I saw Aaron's change [23:56:08] I figured I may as well commit it anyway at that point [23:56:30] Reedy: probably not many [23:57:50] I'm going to merge it to 1.20wmf4