[06:48:32] why does gerrit constantly log me out now? [06:54:35] there have been at least two upgrades in the last couple of days; each upgrade invalidates the session cache [06:55:34] that's my conjecture, anyway, since i've noticed the same thing and it has reliably followed ^demon's announcements on the SAL [06:59:32] ^demon mentioned clearing all disk caches yesterday too [07:01:33] ah, that explains why commits were so slow to load too [07:01:34] thanks [07:26:01] or maybe it's just gerrit not loading changes on firefox [07:26:52] no; that's an actual, bona-fide bug, unrelated to login [07:37:15] I want to use the prop=coordinates API but the documentation doesn't specify what is the format of the titles paramter when you want to give more than one title, i.e. is it comma separated? | separated or what? Nor https://www.mediawiki.org/wiki/Extension:GeoData#prop.3Dcoordinates or https://en.wikipedia.org/w/api.php give any hint to that the titles parameter values should look like [07:38:18] ShiningThrough: it's pipe separated throughout all api modules [07:39:37] ShiningThrough: see https://www.mediawiki.org/wiki/API:Query#Specifying_pages [07:45:28] liangent: thanks [08:00:06] not too bad, it seems Tool Labs has reached about 1/30 of Toolserver's hits http://tools.wmflabs.org/awstats/cgi-bin/awstats.pl?month=06&year=2013&output=urldetail&config=tools.wmflabs.org&framename=index https://meta.wikimedia.org/wiki/Future_of_Toolserver#Why_Toolserver_is_important [09:02:42] where does http://www.nedworks.org/~mark/reqstats/ live nowadays? [09:03:20] nowhere [09:06:07] too bad, it was such a friendly mate [09:06:55] I'm trying to make 10 million HTTP requests to get coordinates for articles on English Wikipedia, however after 30k requests I get "ConnectionResetError: [Errno 104] Connection reset by peer" even though I only make 1 request at a time and 1 second pause between requests, any parameters I need to adapt to maintain the connection and not get reset by peer from the server? [09:08:16] ShiningThrough: in a single connection? [09:08:29] there's probably a limit to the nr of requests per connection [09:08:53] I think the client I use has some kind of internal HTTP Connection Pool... [09:09:53] http://torrus.wikimedia.org/torrus "Error: Error initializing Renderer" expected? [09:10:45] To bad I can't get a list of page name which have coordinates... [09:11:07] ShiningThrough: you are BadDesign I suppose? [09:11:14] I've grabbed the coordinates from the enwiki-latest-page.sql.gz [09:11:24] err [09:11:26] the page names [09:11:31] Nemo_bis: yes [09:14:08] so I guess the API is not usable and all I have is parse the coordinates from XML dumps which is a horrible thing to do [09:16:07] Nemo_bis: sorry Nemo_bis I need the coordinates as soon as possible I can't wait to do what you said [09:20:00] ShiningThrough: have you started at 6 UTC? [09:21:25] Nemo_bis: no, just a few minutes ago [09:22:01] roughly, half an hour ago I started the Python script [09:23:22] I've used this library: http://docs.python-requests.org/en/latest/ [09:23:41] which says that "Keep-alive and HTTP connection pooling are 100% automatic" using it [09:25:20] ShiningThrough: Why do you say it will take 10 million requests? [09:25:52] are you setting a custom user-agent? [09:27:00] legoktm: because I have ~ 10 mil pages extracted from enwiki-latest-page.sql [09:27:13] legoktm: I have a user agent specified [09:27:13] Not all of those pages have coordinates though [09:27:19] Plus you really only want mainspace coordinates [09:27:32] You should be able to fetch up to 50 pages at once [09:28:03] yes, I do that [09:28:18] On enwp, any page with coordinates uses {{coordinates}}, so you can use the list of pages that use that [09:28:25] rather than checking each page [09:29:45] legoktm: in which dumps should I check for {{coordinates}} ? [09:31:34] you could use the tempatelinks dump I think. It might be easier if you use pywikipedia since all these API commands are coded in [09:32:01] https://github.com/wikimedia/pywikibot-core [09:43:33] legoktm: I'll try that, thanks; for reference here's the code I used http://paste.kde.org/786980/ which gets Connection reset by peer [09:44:12] well you really shouldn't be spoofing fake headers [09:44:41] something like 'ShiningThrough on freenode' would have been better [09:47:29] ShiningThrough: if you do use pwb, your code would look something like "for page in pywikibot.Page(site, 'Template:coord').getReferences(onlyTemplateInclusion=True, namespaces=[0]): print page.getcoordinfo() [09:49:14] i'm gonna go afk, you can probably get help in here or #pywikipediabot [09:49:43] legoktm: ok, thx, but that would only include pages with template {{Coord or all the pages with main coordinate set ? [09:50:21] It would include all pages with {{coord}} which is going to be all pages with a main coordinate set. [09:51:10] There are ways to specify the main coordinate of the article other than {{Coord:..}} [09:51:17] and many articles do use that [09:52:19] anyway it'll have to try and see [13:46:40] apergos, u thr man.? [13:46:56] need some help on the last topic. [13:47:03] about that dump thing.! [13:56:20] shanu1991: [13:56:22] sort of here [13:56:26] (lunch break in a few) [13:56:32] okay [13:56:45] so i'll not take much time i guess [13:56:56] no i have completed the sql commands [13:57:02] and have got the database ready [13:57:21] now how to read that without phpmyadmin [13:57:30] cause its huge amount [13:57:35] u know what i mean. [14:12:08] csteipp: curl security issue [14:12:54] matanya: A new one today? [14:13:20] fix published today [14:13:43] http://www.ubuntu.com/usn/usn-1894-1/ [14:14:08] csteipp: ^ [15:08:37] apergos, ur lunch over man? [15:08:57] over enough [15:09:00] so what's up? [15:21:41] Nemo_bis: ok switching :) [15:23:06] Nemo_bis: no that line is theoretically not needed because the categories are empty, but also no reason the remove them at the moment [15:24:12] and full disable is approved by "Bei inhaltlicher Ablehnung: Vollständige Deaktivierung" with a transition period to process the exiting feedbacks without getting new ones [15:25:42] Reedy: do you feel comfortable with that change? [15:25:44] looks like one more reason to create a bugzilla ticket where to copy this info :) [15:26:18] but the change in itself seems non-problematic [15:27:00] I'm logged out of gerrit :( [15:27:07] who isn't [15:27:41] Means I've got to go find my yubikey [15:31:35] my Strg+C didn't work corretly since some days. mostly won't copy it at 1st try :( meh [15:50:14] Did the VisualEditor get disabled? I don't see it in Special:Preferences#mw-prefsection-editing anymore on enwiki [15:50:39] oh, it got activated for all accounts? [15:54:27] enabled for logged in I guess, not for anon yet? [15:55:18] yes and yes [15:56:48] Reedy: do you found your key? I gotta go in half an hour and would love to be online. quick testing possible cache issues and so [15:57:58] Nope, let me go find it [15:58:04] oh [15:58:07] It's on my desk under paper [16:09:46] Reedy: Is the change live? [16:14:05] se4598: yes [16:17:28] The change doesn't do what it should do, it will be still displayed on articles with the category, (lottery-disable apparently worked) :((( [16:17:50] Purge? [16:17:53] The pages that is [16:18:08] e.g. https://de.wikipedia.org/wiki/Schachverein [16:18:20] tried already purge and nulledit [16:19:54] wmgArticleFeedbackv5Namespaces [16:19:59] 'dewiki' => array( NS_MAIN ), [16:20:12] That line wants removing [16:20:14] incremental dumps? [16:21:25] apergos: hello [16:21:49] hey Svick [16:22:01] lemme ping our third co-partner in crime [16:22:13] Reedy: last resort but that will also disable the special pages (if not set to an alibi ns) and I don't know the consequenzes with cache because the JS-module wouldn't loaded serverside any more [16:22:24] apergos, ok, i'll wait [16:23:16] should be showing up any sec [16:23:23] se4598: Purge it again? [16:23:34] Reedy: apparently the change didn't break anything. I think that is an issue through a new way of feedback activation by hand, but which can't be disabled by config [16:23:51] parent5446: hello [16:24:04] howdy parent5446 [16:24:04] Hey [16:24:04] guess the gang's all here [16:24:30] Svick: any needs for gerrit or is it set up the way you want? [16:25:07] i haven't actually tried committing yet, but i think it should work [16:25:10] Reedy: even removed the relevant category by hand, but nothing, fortunately feedback submission won't get redirected by server side (which should, but so didn't break anything) [16:25:47] ok, let me or qchris or ^demon know if it's not working [16:26:01] what do folks think about the email discussion? [16:26:51] I'm still pretty hesistant about using XML, but then again it's not my area of expertise so... [16:26:55] Reedy: will have to wait for mmuillie to come back from vacation in a week. I'm off now [16:27:04] not much to report today; i was mostly dealing with the mail thread yesterday and i will do that a but today too; aftter that, i think i will continue working on my initial code (so far, it's parsing the output of dumpBackup.php in C++) [16:28:05] apergos: i really don't see how could incremental dumps that are easily readable work [16:28:16] Although I do like the idea of possibly having a post-conversion tool to XML [16:28:34] we pretty much will have to have a conversion tool [16:28:45] because there is an entire ecosystem of stuff out there that expects the xml format [16:28:54] we absolutely cannot break that without giving folks a long [16:29:01] (many months. a year?) heads up [16:29:07] Mhm [16:29:08] yeah, i had that conversion tool in my "optional" list, i underestimated how important it is [16:29:34] if it can't fit in the timeline, this is ok but [16:29:49] it owuld mean you hack on that in your volunteer time after the project is done [16:29:55] (secret plan to keep you as a volunteer) [16:29:59] ;) [16:30:30] yeah, if there is some unfinished bussinness at the end of summer, i would like to continue [16:30:36] good! [16:31:01] are there questions you have for us (or lurkers in the channel) about anything (code. architecture. timeline. anything at all)? [16:31:24] though i will have to go to school too, so i won't have sa much time as now (but it won't be as bad as the past month, when i had to study for exams) [16:32:23] yes, that's understood [16:32:37] apergos: i don't have any questions [16:33:11] anything where you need us to run intereference ? I guess it's early in the process but still [16:33:31] apergos: what do you mean? [16:33:56] uh where you need something done and it would be easier for one of us to poke someone (for example) [16:34:17] I'm more bringing this up so you know it's something we can do [16:34:20] right; no, nothing like that [16:34:34] parent5446: any questions, comments..?? [16:34:37] I'll let you know if anything comes up [16:34:43] great [16:34:52] None that I can think of. We're pretty early on. [16:34:55] yep [16:35:09] well I'm good, looking forward to first commit :-) [16:35:20] Same here [16:35:42] see folks tomorrow? although parent5446 you said tomorrow you can't be here. so thursday for you [16:35:44] just don't expect anything spectacular :-) [16:35:47] Yep [16:36:03] hahaha I expect not only spectacular but awesomely awesome... by the end that is ;-) [16:36:24] ok, see you tomorrow [16:36:26] all right, happy coding all! [16:39:09] oh, funny, I didn't know "Tntnet" was still in use as a name :) [17:11:17] How're folks holding up? :) [17:16:13] hi, does anyone here know anything about list admin functions? [17:17:10] as in? [17:17:33] I would like to create moderators (not admin), but have no idea what the website is they should go to. [17:18:11] regular is "listinfo", admin is "admin", what is moderator? [17:18:24] for the API? [17:18:27] idk [17:18:31] someone else in here would know [17:18:44] anyone in particular? [17:19:11] * Jasper_Deng knows a few people but they're probably idle atm [17:31:23] nevermind i found out already [17:52:13] hey apergos [17:52:38] sorry had gone to some urgent work.! [17:52:40] u thr? [17:52:52] idle for about half an hour [17:53:19] ?? [17:54:41] Jasper_Deng, were u saying that to me? [17:54:45] yes [17:54:54] apergos may be busy or away [18:22:51] Jasper_Deng: i assume the question was about mailman not the API [18:23:04] taketa seems to no longer be here [18:52:01] my search icon is not lined up correctly with the search box, on all wikis. [18:52:18] is this ULS-related? [18:59:18] Sounds plausible [19:06:43] hi [19:06:48] I have a quick question ... [19:06:52] regarding collapsable templates [19:08:34] how would I be able to get a template to be collapsed while on the bottom of every page it appears on BUT to be uncollapsed when people access it directly [19:08:36] ? [19:09:51] WikiJunkie: Switch on the namespace [19:10:40] multichill: I am not sure what that means [19:10:45] See for example https://en.wikipedia.org/wiki/Template:Monty_Python [19:11:04] On https://en.wikipedia.org/wiki/Monty_Python_Live_at_the_Hollywood_Bowl#External_links it's collapsed [19:11:24] It uses {{Monty Python|state=collapsed}} [19:11:54] thanks [19:12:12] Seems to be part of the logic of https://en.wikipedia.org/wiki/Module:Navbox [19:12:24] What language wikipedia do you want to use this on? [19:13:06] On the Hebrew Wikivoyage [19:14:18] http://he.wikivoyage.org/w/index.php?title=%D7%AA%D7%91%D7%A0%D7%99%D7%AA:%D7%94%D7%A2%D7%A7%D7%A8%D7%95%D7%A0%D7%95%D7%AA_%D7%95%D7%94%D7%A7%D7%95%D7%95%D7%99%D7%9D_%D7%94%D7%9E%D7%A0%D7%97%D7%99%D7%9D_%D7%94%D7%9E%D7%A8%D7%9B%D7%96%D7%99%D7%99%D7%9D_%D7%91%D7%95%D7%95%D7%99%D7%A7%D7%99%D7%9E%D7%A1%D7%A2&oldid=23964 [19:14:30] Ok, I'm out.... [19:14:37] multichill: that is the template [19:16:11] multichill, I've slightly changed the code but now it is uncollapsable on all articles it appeares on [19:16:25] It's some magic css [19:18:37] I am about to run scap [19:19:13] does anyone know what the simplest solution would be ? [19:26:06] anyone? [20:42:54] some help with mediawiki database.? [20:42:58] anyone [20:43:42] shanu1991: #mediawiki might be better; also, just ask your question [20:45:14] i have to change the database of the a newly created mediawiki installation [20:45:26] #mediawiki [20:45:54] i have exported the newly created database and imported into the one I want to use [20:46:21] but I still couldn't see the pages of the old database? [20:46:40] which(if any) extensions do i have to enable? [20:47:10] shanu1991, they said go to #mediawiki [20:47:27] oaky [20:47:28] *okay [22:18:02] * jorgeluis gently links https://en.wikipedia.org/wiki/Wikipedia:Village_pump_%28technical%29#Unresponsive_script [22:42:07] thanks jorgeluis -- James_F ^^ ? [22:46:21] ooo [22:46:30] I was wondering if there was a page about the VE [22:46:48] Something here needs to change. The WMF cannot be allowed to ram change down our throats. I think the best course of action is for them to immediatly implement an easy opt-out function, via sitewide banner and admit they made a mistake here. [22:47:16] Isn't this exactly what Elsie told us would happen… [22:47:23] well, there is http://www.mediawiki.org/wiki/VisualEditor [22:47:41] and https://en.wikipedia.org/wiki/Wikipedia:VisualEditor/Feedback [22:48:08] deargod [22:48:19] I would've bought popcorn if I knew there was an angry thread about the VE [22:48:42] > 197 Citations adder is incomprehensible (LET ME OUT OF THIS) [22:48:44] lawl. [22:50:53] jorgeluis: that's probably ULS, actually, not VE [22:51:04] jorgeluis: there was a similar ULS bug reported, i think [22:51:23] Maggie, with respect, VE is not here. It isn't ready. There are too many problems and bugs that need to be resolved before it can be declared here. [22:51:34] I'm actually suprised it wasn't Elsie who wrote that. [22:51:43] Elsie: waaaakeee uuuuup! [22:53:05] * odder declares undying love to his four-year-old browser [22:53:18] it loads the normal editor by default! yay! [22:53:38] odder: only the article and user namespace load it right now [22:53:53] MatmaRex: Interesting. Do you want to make a note there or should I? [22:54:27] jorgeluis: I tried editing a random article, still got the normal editor [22:54:39] Weird. [22:54:46] jorgeluis: frankly i have noidea what might be causing this [22:54:54] jorgeluis: but i vaguely remember ULS having such troubles [22:56:58] odder: As of about 72 hours before the deployment yesterday, there were nearly 300 confirmed and unfixed bugs for VE. I think that alone should be a huge red flag. [22:58:28] jorgeluis: I'm CC'd to a couple of VE bugs, and I've seen many suggestions it's much too early to enable it on such a massive scale [22:59:12] but since my browser does not support VE, and I'm not planning on upgrading it in a foreseeable future, it's not really my problem [22:59:35] heh. [23:00:26] from a third-party position, though, I think that the WMF should be awarded a "We haven't learnt any lessons about how to communicate with the community properly" award for 2013 [23:00:55] jorgeluis/odder to be fair, mediawiki itself has hundreds of unresolved bugs [23:01:21] odder: WMDE is a big competitor here ;-) [23:01:52] odder: nah, they are learning [23:01:59] it could have been worse [23:02:16] I think it could have been worse, too. But it's pretty bad :) [23:02:37] it's kinda sad though that afaik almost everyone apart from the VE team themselves were suggesting postponing the deployment and it still happened [23:03:03] i guess deadlines are the most important [23:03:56] https://en.wikipedia.org/wiki/Wikipedia:Petition_to_the_WMF_on_handling_of_interface_changes is interesting [23:04:36] meh, that's frankly bullshit [23:04:43] en.wp has a vocal group of users who oppose any and all changes [23:04:45] petitions usually are :) [23:04:47] even the most trivial ones [23:05:06] like the moving of section edit links to the left [23:06:30] * DaBPunkt always liked that change :-) [23:07:01] i think that change was good, but it took some getting used to [23:08:09] # of bug reports is surely a reliable way of determining the quality of any software :/ [23:09:05] greg-g: i'm sure you don't need me to point out the ones which should have been considered blockers [23:09:14] Well, esp. given that it randomly removes or reformats code you didn't touch [23:09:18] (but i can do it if you wish) [23:09:20] that's probably important. [23:11:33] MatmaRex: that is a different thing. [23:12:48] greg-g: VE should not have been deployed in its current state, that's the thing [23:13:04] i'm pretty sure it will be much better next week [23:13:17] and you all should have waited for that week [23:13:20] but oh well. [23:13:29] That's an opinion stated as a fact, which doesn't leave room for discussion. [23:13:53] and it's way past the time for discussion, too. [23:14:09] so, then what are we doing? [23:14:15] damage control [23:14:34] if we're not discussing and you're just yelling at us, then that doesn't help. If we're going to really have a discussion then we should. [23:14:38] VE community team are handling it pretty well considering the circumstances, though. [23:14:47] http://enwp.org/User:Technical_13/Userboxes/Anti-VE [23:15:19] greg-g: so what we gonna do now, undeploy it? [23:15:24] there's nothing to discuss anymore [23:15:26] Releasing too early caused ^^^^^^ [23:15:28] there's just damage control. [23:15:40] any release would have caused that userbox, and you know it [23:16:04] that's opinion presented as fact [23:16:06] that userbox says "appearance" not "too early" [23:16:16] greg-g: want me to link bugs i think should haveb een fixed before dpeloyment? [23:16:26] there are like five of them [23:16:30] MatmaRex: sure, but you should instead talk to James_F [23:16:32] not more [23:16:43] I don't manage that team [23:16:50] Hey. [23:16:52] greg-g: that wording is very mellow because of ironholds... [23:17:20] greg-g: you manage deployments? [23:17:40] MatmaRex: yep, based on the expert opinions of others. [23:17:54] I can't know *everything*, no one can. Thus, delegation happens. [23:17:59] odder: heh. everyone should be rewarded with the "noone can communicate with anyone else" award for 2013, really [23:18:02] and i've already talked to James_F, eh [23:18:09] Ryan_Lane++ [23:18:12] One question then I'll butt out... [23:18:21] because seriously, there's about 928309283404 ways to communicate with "the community" and all of them are *terrible* [23:18:23] MatmaRex: then you've talked to the right person. [23:18:27] echo/flow should make that easier [23:18:28] Big upgrade to VE next week? [23:18:42] Technical_13: I doubt it; we're fixing things that the community are asking us to fix, instead. [23:19:00] There are planned and in progress and already completed (but yet to be deployed) changes to VE, yes [23:19:14] But yes, lots still to come, obviously. [23:19:21] they aren't done :) [23:19:38] I'm waiting for some indication that there is a najor improvement before I try it again. [23:19:51] define "major" [23:20:01] just so I know when to ping you :) [23:20:26] hmmmm, i have an "enable feedback" link in my toolbox that links to action=protect [23:20:29] anyone see that [23:20:33] on enwiki [23:20:40] a gadget malfunctioning? [23:20:57] aude: i think there's a bug for that [23:21:10] aude: Feedback is actually managed via action=protect isn't it? [23:21:30] oh, ok [23:21:38] There's "Article feedback" on https://en.wikipedia.org/w/index.php?title=Main_Page&action=unprotect [23:21:46] * aude confused and does not think it warrants a toolbox link [23:21:51] <^d> I hated that when we did it in FlaggedRevs too. Silly place to put it :) [23:21:57] <^d> In protection, that is. [23:22:03] should be in cactions [23:22:41] or maybe there isn't. can't find it. [23:23:06] Technical_13: you should probably cc yourself to some bugs [23:23:20] well, not critical but would be enhancement [23:23:41] greg-g: it actually working and not destroying pages is a good start. Tables templates citations all working correctly... [23:24:06] Technical_13: look through https://bugzilla.wikimedia.org/buglist.cgi?quicksearch=%3Avisualeditor for some keywords :) [23:24:14] MatmaRex: any suggestions? Feel free to add me to the cc if you see something I may like. [23:24:55] Technical_13: approximately last two screens of that link are things reported after deployment [23:24:55] When is mw Bugzilla getting updated? [23:25:00] Technical_13: https://bugzilla.wikimedia.org/show_bug.cgi?id=49833 is improvements to template editing. [23:25:00] you'll like most of them [23:25:35] Technical_13: https://bugzilla.wikimedia.org/show_bug.cgi?id=50540 https://bugzilla.wikimedia.org/show_bug.cgi?id=50542 [23:26:48] Technical_13: I believe we recently upgraded BZ, no? [23:27:08] <^d> We're still on 4.2.4 [23:27:30] <^d> 4.2.6 is latest in 4.2 branch, but 4.4 is also out now. [23:27:42] ah [23:27:42] 4.5 is out too... [23:27:53] file a bug ? ;) [23:28:08] <^d> Oh? I didn't see it on bugzilla.org [23:28:17] 4.5 would be an experimental ver iirc [23:28:42] <^d> Oh yeah, we've always stayed on the even releases. [23:28:59] http://www.bugzilla.org/news/#release44 may 22 2013 \o/ [23:29:15] ah we were all busy w/ amsterdam i bet [23:29:23] didn't see the announce [23:29:39] I'd much rather see you guys drop Bugzilla and create your own ticketing system that works for everyone (make it use SUL)... [23:29:51] ..... [23:30:06] Technical_13: what [23:30:20] build it on wikidata, dude [23:30:20] that'd be just stupid [23:30:28] * MatmaRex <3 Bugzilla [23:30:53] <^d> That being said, BZ now supports OpenID, so we could (in theory) share logins with wikitech/labs. [23:30:53] I don't love bugzilla by any means, but a homegrown option wouldn't be wise given the # of devs we have [23:30:56] <^d> Would be kinda nice. [23:30:57] Technical_13: We're intending to switch Bugzilla over to SUL at some point, if our soon-to-be-done OpenID provider can stretch [23:31:11] ^d: Bah, getting there ahead of me! :-) [23:31:18] i hate reinventing wheels is all [23:31:21] it's cuz he has a short nick today [23:31:24] Universal log in works correctly on hundeds of languages, bur doesn't work on test.wiki or Bugzilla... [23:31:25] <^d> brion: You wrote code review. [23:31:27] special:code notwithstanding ;) [23:31:31] hah! [23:31:55] special:bug! [23:31:57] <^d> So did I, so I can't blame anyone ;-) [23:31:57] Technical_13: testwiki is a special case wiki. And bugzilla isn't a wiki [23:32:37] Cstiep is fixing the test wiki thing.. [23:33:33] <^d> James_F: SUL plugin for Gerrit? ;-) [23:33:47] ^d: That'd be nice. [23:34:01] <^d> The ;-) meant I was joking. [23:34:13] ^d: Wouldn't necessarily be too hard. [23:34:19] <^d> Yes, it would. [23:34:39] Technical_13: Bz has the big issue of "leaking" email accounts [23:34:55] <^d> James_F: Ideal world is having OpenID on wikitech/labs though, and having Gerrit use that. [23:35:01] * James_F nods. [23:35:03] <^d> So you don't need your username/password on two sites. [23:35:04] Then fix the bug.. :p [23:35:14] :) [23:35:24] Technical_13: We would, but unfortunately we're on IRC arguing about it rather than having the time to fix it. :-) [23:35:48] <^d> James_F: Problem right now is Gerrit can't use multiple account info sources. So if you use OpenID, we can't fetch groups and such from LDAP [23:35:54] <^d> c'est la vie. [23:36:03] * aude wants to login someday on openstreetmap with my wikipedia (open id provider) account :D [23:36:08] you could make an ldap that bridges to openid [23:36:10] the osm part is there already [23:36:38] ^d: So we just need MW to act as an LDAP source too? ;-) [23:36:47] aude: oh yes – someday :-) [23:36:48] we already have that [23:36:48] Touche James_F good sir.. [23:37:04] so, my plan is to have openid on wikitech [23:37:07] as a provider [23:37:12] then gerrit -> wikitech [23:37:17] and wikitech -> wikipedia [23:37:21] Technical_13: Honestly, feedback is helpful; please don't stop speaking because of my comment. It's hugely valuable. [23:37:21] cool :) [23:37:41] chained openid is possible and I've tested it with the openid extension for MW [23:37:45] Ryan_Lane: Ooh. [23:37:54] of course, we also need OAuth [23:37:59] yes yes [23:38:00] and we need keystone to support OAuth [23:38:14] <^d> But the annoying thing is still the groups issue. [23:38:23] then, when you log into wikitech, it will OAuth log you into keystone [23:38:41] so, if you chain auth to wikipedia, you can still use OpenStack [23:39:11] ^d: openid attribute exchange [23:39:19] we can pass the groups from wikitech to gerrit [23:39:28] <^d> Possibly. [23:39:38] we'd need to add support for that in the openid implementation of gerrit, though [23:40:08] one problem, though, is when we modify groups [23:40:23] really it would be nice to have openid for authn and ldap for authz [23:41:42] could do openid at the apache level as "optional" having the login link go to wikitech with a returnto value [23:41:51] James_F: you can't shut me up that easy.. distracted by my baby is all. [23:41:57] and then use web+ldap [23:42:00] Technical_13: :-) [23:42:11] Technical_13: A good thing by which to be distracted. [23:42:24] then we'd have openid for authn and ldap for authz [23:42:31] and wouldn't need to modify gerrit [23:42:38] I wonder how good the openid support is in apache [23:43:00] <^d> I never tested that auth mode. [23:43:02] <^d> I have doubts as to how robust it is. [23:44:41] I brought the Bugzilla thing up because I'm working on revamping the {{tracked}} template/script on enwp. Would be nice to be able to retrieve the resolved status from api. [23:44:55] Ummm [23:45:12] I just blocked an account at 2013-07-02T23:42:24, and then it proceeded to edit at 10 seconds later [23:45:26] 6 seconds later [23:45:37] (del/undel) 2013-07-02T23:42:24 Legoktm (talk | contribs | block) blocked [offensive username] (talk | contribs) (account creation blocked) with an expiry time of indefinite ({{uw-vaublock}}) (unblock | change block) [23:45:41] (del/undel) 2013-07-02T23:42:31 (diff | hist) . . (+80)‎ . . m Rex Ryan ‎ (Tag: VisualEditor) [23:45:44] Where was the edit legoktm ? [23:46:01]