[01:25:26] greg-g: is the https login rollout going to be covered by a central notice? [01:27:56] Hi sDrewth. [01:28:19] gday Elsie [01:28:47] sorry, with that moniker, I keep wanting to say "I like what you have done to your hair" [01:28:52] weird! [01:29:26] * sDrewth checks in for a checkout [04:27:49] is the JobQueue supposed to ever be at 0? because AFAICT it is. [04:29:09] lego ^ [04:29:27] PinkAmpersand: yes, thats a known bug that the redis job queue only displays 0 [04:29:35] oh ok [04:29:40] there's an open patch for it, but it hasn't been reviewed [04:29:47] was trying to figure out why http://en.wikipedia.org/w/index.php?title=Special:WhatLinksHere/Nostalgia_Wikipedia&limit=999 is still wrong [04:29:48] let me find the bug [04:30:53] PinkAmpersand: https://bugzilla.wikimedia.org/show_bug.cgi?id=50635 [09:16:16] i recall there was a (somewhat) recent change that renamed users to username~wikt or username~en or some similar format, in an attempt to make usernames more consistent; is https://en.wikipedia.org/wiki/Wikipedia:Changing_username not out of date? are usernames not a thing users can change globaly, yet? [09:25:52] gry: that change was to allow for global accounts to be used everywhere [09:26:15] some unused accounts had to be renamed so the global accounts could be attached everywhere [10:03:50] Hi, I probably need someone from WMF. I need access to Special:OAIRepository or information about alternate solutions.. [10:04:12] Who do I bribe to get access? =) [10:05:33] I guess it is to early for people at the SF office.. [10:05:33] Hi JeroenDeDauw ! [10:05:40] JeroenDeDauw: Stop playing games and do something useful! =D [10:05:57] jeblad, what exactly do you need? [10:06:05] * jeblad goes to lunch [10:06:17] Ah MaxSem ! [10:06:51] We wonder about how to include metadata about Wikipedia in Norvegiana and then into Europeana [10:07:21] We have a working OAI-PMH import and OAIRepository seems to be the natural choice [10:07:24] that is, you want to export WD stuff in realtime? [10:07:54] Wp stuff, but bulkimport each day or something [10:08:09] Preferably without doing any additional processing [10:08:50] OAI export code looks extremely creepy [10:08:52] Bulk download and some scripting could be an option [10:09:38] like, mentioning cur [10:10:00] srsly, it gives me the creeps that we're running it:P [10:10:28] Hm.. [10:10:34] you could ask someone in the platform about it [10:10:49] I doubt that there's someone who knows it well though [10:10:54] Ok, thing is, we need an OAI import [10:11:01] lol [10:11:11] An option could be to pay someone to clean up the code [10:11:21] I'd say, just grab RC and update the pages it mentions [10:11:51] but I could be wrong and OAI actually works [10:12:02] also, I'm not sure how secure it is [10:12:18] We need a bulk export of a subset of pages that list specific information [10:13:02] Ok, I can add a note that the extension needs further work.. [10:13:10] crap.. [10:13:13] ;/ [10:13:31] * jeblad goes to lunch [10:15:55] MaxSem: OAI feeds lucene I believe, so it does work [10:16:04] yup [10:16:27] but does it work in general? does it filter out non-public stuff properly? [10:43:22] does anybody know which wmf code change could cause an invalid edit token response on my framework since yesterday? [10:45:01] Merlissimo: not sure, but I think the name of the session cookie was changed yesterday or so [10:45:06] there was an email to wikitech-l about it [10:45:13] not sure if that could be it [10:45:47] sure, because that name is hard coded because it is created from api response [10:46:12] thx [10:48:40] mmh, http://www.mediawiki.org/wiki/API:Login#Construct_cookies still describes the cookie names i use. counldn't that be on api or api-announce mailinglist? [10:53:44] MaxSem Who do I poke to get access to the special page? [10:53:55] robla [10:54:05] We would like to check if this is a viable solution [10:54:20] "we" is the Norwegian Arts Council [10:54:37] oki, I'll write him an email [10:54:45] Thanks [10:54:57] Merlissimo, that was an emergency WMF-specific hack, no point in describing it in generic MW docs [10:55:11] that's what you get for constructing the cookies yourself [10:56:00] * jeblad steals Merlissimos bike [10:56:23] MaxSem: thoughts on getting rid of the 'construct cookies yourself' parts from that page? Sounds like asking for trouble [10:56:45] I've already opened it for editing [10:57:04] to mark it with "if ya wanna shoot yourself on foot, fine" [10:57:40] however, I like your solution better [10:58:19] implemented [10:58:30] :D [11:03:16] jeblad: i have no bike [11:04:35] of course you don't! because he stole it [12:06:16] jeblad: lolwat? [12:06:54] jeblad: "doing something useful" is not part of my job description [12:08:07] * MaxSem gives JeroenDeDauw a barrel-sized mug o'beer [13:16:45] * JeroenDeDauw trades the beer for clubmate and throws the barrel at jeblad [14:13:24] JeroenDeDauw: Moahahahaaa [14:13:37] JeroenDeDauw: Let me repeat that [14:13:45] JeroenDeDauw: Moahahahaaaa [14:14:12] ..and the chance is that I will be back in Berlin! :D [14:43:22] * JeroenDeDauw readies WMDE lazer defencez [15:34:53] http://wikistream.wmflabs.org/ does not work :/ [15:35:59] Stryn: don't report that here... [15:36:05] Stryn: try #wikimedia-labs [15:36:12] ok, thanks [15:38:51] it angers the opsen if you report labs outages [15:39:00] * greg-g kids [15:42:01] It sounds like wikistream needs a stacktrae. [15:42:04] stacktrace [15:44:01] Elsie so funny!!!! [15:44:29] Elsie: javascript and java are different! only java needs stacktraces! [15:45:04] Don't ruin the joke. The typo already tried. [16:16:47] parent5446: hello [16:16:57] apergos: hello, are you here, or still sick? [16:17:07] I think she might be sick still [16:17:23] sumanah: ok [16:17:38] I am here [16:17:55] started out the day a bit unsure but migraine has jus lurked in the background so it was a productive day [16:17:59] Hey [16:18:28] <^d> Hey parent5446 :) [16:18:48] "day" in past tense. so weird [16:18:56] so I hear parent5446 that your classes are over [16:19:05] yay! and... what's your meeting time availability now? [16:19:20] Yep. I'm available all day pretty much [16:19:59] how would folks feel about starting an hour and 15 mins earlier? does this bork your schedules? [16:20:07] (trying to get back a little of my evenings) [16:20:21] that's not a problem for me [16:20:39] That's fine [16:20:57] if you would still be getting coffe or rolling out of the bed at that time, just say so [16:21:26] Nah that's like 11am my time. I should be OK [16:21:31] sweet [16:21:45] how's diff dumps going? [16:21:47] ok, so today i'm working on fixing bugs in diff dumps; i hope to have version that actually works later today [16:22:04] anything particularly problematic? [16:22:12] now that I have a working brain again... [16:22:48] i think i have found the last bug, it should be only a matter of fixing it and testing that it works [16:22:56] so no, i don't need any help with that [16:23:48] yay! [16:24:22] anything you want to bring up? [16:25:05] after that, you should be able to try it out too [16:25:33] i can't think of anything [16:25:38] * apergos rubs hands together and cackles evilly [16:25:56] ok, well I look forward to it, and I got nothing pending myself [16:26:08] parent5446: [16:26:21] I don't have anything in particular [16:26:22] we should start looking for reviewers with c++ chops [16:26:42] (of which I am not one, I do C without the pluses) [16:26:47] though you probably will need some instructions on how to use it from me (i will really need to write some sort of manual when we get near the end) [16:26:51] :-D [16:26:57] maybe a bit before the end! [16:27:04] maybe [16:27:24] so the thought behind getting someone to get eyeballs on it now is that review around wmf is traditionally slow, there's a lot of folks competing for our time [16:27:54] and it's better if they have any comments that involve large structural changes to get that in sooner rather than later [16:28:09] parent5446: any ideas? [16:28:39] Unfortunately not. I have yet to find any way to make review faster on MW. [16:29:01] all right, I'll start asking around [16:29:25] in which case... [16:29:32] see folks tomorrow, one hour and 15 mins earlier? [16:30:00] just prepare them that it's a relatively big amount of code [16:30:24] Yep see you tomorrow [16:30:25] see you [16:31:06] see yas [17:27:44] yo [17:27:51] who is Tyler Romeo? [17:32:48] domas: parent5446 [17:36:47] domas: What's up? [17:38:56] parent5446: ;-) reading blogs? [17:39:16] domas: Lol I found that article a while ago. If it's not accurate just comment. [17:39:28] Ah you already did. [17:39:31] it is accurate in a way [17:39:54] but "you may observe stalls" is something you would only see in certain workload scenarios [17:39:58] Mhm I understand your point [17:40:30] Designing databases shouldn't be this hard. :/ [17:40:54] what do you mean? [17:41:26] anyway, it is kinda funny, auto-inc table level locking will induce more locking contention than global lock [18:23:55] running scap in a second [18:56:10] ^d: is there any place I can see the new HTTPS preference in action? [18:56:32] (short of installing my own copy of MediaWiki) [19:01:30] <^d> robla: It's enabled on beta. [19:01:49] <^d> Known issue: https on beta is unstyled due to bits not having ssl. [19:04:16] <^d> I'm fine with throwing the switch on test2 as well if we want. [19:04:20] is the https change also affecting loggedin users using api like bots? [19:05:15] ^d: let's put it on test2 [19:05:26] greg-g: ^ [19:07:33] Is there a page/blogpost/mail explaining why the wmf doesn’t use PFS from the beginning? [19:08:24] Merlissimo: I guess they will be redirected. Not sure what happens for POST requests, though. [19:09:03] DaBPunkt: I'm assuming you mean this: https://en.wikipedia.org/wiki/Perfect_forward_secrecy [19:09:10] yes [19:09:50] is this accurate? "In theory, Transport Layer Security (TLS) can choose appropriate ciphers since SSLv3, but in everyday practice many implementations have refused to offer PFS or only provide it with very low encryption grade.[5] OpenSSL supports perfect forward secrecy using elliptic curve Diffie–Hellman since version 1.0,[6] with a computational overhead of approximately 15%.[7]" [19:10:51] robla: That's about right [19:10:59] but it means that i have to add cipher capabilities to my framework [19:13:42] ^d: you wanna do the test2 deploy today? [19:13:59] <^d> Yeah, sooner the better. [19:14:04] cool, fire at will [19:14:15] <^d> Ok, merging our stuff to relevant branches now. [19:14:43] robla: so teh answer is: Because wmf has not enough CPU-power? [19:15:01] Merlissimo: It depends... existing logged in sessions will not be affected [19:15:45] If you're loging int by submitting a POST to Special:UserLogin, then that will redirect to https [19:16:30] my bots uses api login [19:16:31] DaBPunkt: this is the first I've been aware that it was even a consideration. My understanding after a brief conversation with csteipp is that the startup overhead and lack of ubiquitous support were reasons for not going there right away. [19:17:07] Merlissimo: I don't think we're redirecting the API [19:18:07] DaBPunkt: also, Ryan L wrote more on this topic on wikitech-l [19:18:09] could you change "I think" to "I am sure" before changing to https? [19:18:29] he wrote: "I'm not saying that PFS is worthless, but I am saying that implementing PFS without first solving the issue of timing and traffic analysis vulnerabilities is a waste of our server's resources." [19:18:45] robla: ok, I will take a look into that, thanks :-) [19:36:54] chrismcmahon: it looks pretty likely that the SSL switch messed up Selenium tests. is that the case? [19:37:14] robla: one of those yes, but situations. [19:37:55] robla: selenium tests are messed up, but zeljko and I also hosed the builds running them temporarily while we get the CirrusSearch tests [19:37:58] shall we switch the testing https site to only test2 and turn it off on beta labs? [19:38:03] huh [19:38:28] greg-g: yes, it is the lack of bits on beta that's a hassle [19:38:35] greg-g: yeah, I think so [19:38:38] <^d> Why don't we just fix bits? [19:38:49] well, right [19:38:49] <^d> Rather than turning this back off each time. [19:39:07] how easy is it to "just fix bits"? [19:39:14] ^d: by all means, if that's easy to do [19:39:37] I'm assuming that since it's been visibly busted for 3 hours that it's not easy to do [19:40:38] my assumption is something "fun" involving ops [19:41:50] I see nothing in bugzilla about this... https://bugzilla.wikimedia.org/buglist.cgi?title=Special%3ASearch&quicksearch=ssl%20bits&list_id=227363 [19:42:18] <^d> Haha, it does work. [19:42:24] <^d> It's our browsers being funny. [19:42:43] nor noting in RT [19:42:45] <^d> Open this url: https://bits.beta.wmflabs.org/en.wikipedia.beta.wmflabs.org/load.php?debug=false&lang=en&modules=ext.gadget.PagesForDeletion%2CwmfFR2011Style%7Cext.rtlcite%2Cwikihiero%7Cext.uls.nojs%7Cext.visualEditor.viewPageTarget.noscript%7Cmediawiki.legacy.commonPrint%2Cshared%7Cmw.PopUpMediaTransform%7Cskins.vector&only=styles&skin=vector&* [19:42:53] untrusted [19:42:55] <^d> Click through the "certificate can't be trusted" crap. [19:43:08] <^d> beta's ssl has never had trusted ssl. [19:43:20] <^d> Then go back to https://en.wikipedia.beta.wmflabs.org/wiki/Main_Page [19:43:54] <^d> Repeat for https://upload.beta.wmflabs.org/wikipedia/en/b/bc/Wiki.png [19:43:55] styling [19:44:01] <^d> To get images. [19:44:32] <^d> I now have a fully styled & imaged https beta. [19:44:32] so, is Ops decidedly against getting wmflabs a real cert? [19:44:42] <^d> It's warning the heck out of me, but it works ;-) [19:46:16] greg-g: well i think the problem is that its "secret" won't be secret [19:46:20] so it will be false security [19:47:04] LeslieCarr: 1) who cares about it being false? [19:47:13] it's kinda unusable in its current state [19:47:32] I think hasher/ryan were talking about setting up a separate project for ssl termination [19:48:01] ah, that'd be nice [19:48:06] So limited admins there, and we could issue a real cert [19:48:27] well i guess if we don't mind that fact we could give it a cert [19:48:41] so, not an easy fix, got it [19:48:41] ;) [19:48:46] csteipp: do you know of an issue/ticket that they're using to track this? [19:51:31] greg-g: https://bugzilla.wikimedia.org/show_bug.cgi?id=48501 ? [19:51:56] grrr, just reported one [19:52:28] There are 4 bugs with about the same request... 48501 is probably the best one to track though [19:52:34] robla: huh, Ryan isn't on that one [19:52:37] * greg-g nods [19:52:41] how about a procurement ticket in rt for a cert ? [19:52:52] https://rt.wikimedia.org/Ticket/Display.html?id=5184 [19:52:57] with an acknowledgement that "yes, we know this will mean it's not actually secure" [19:54:10] greg-g: it's not *just* the browser tests, it's just a hassle in general [19:54:38] chrismcmahon: right. concur. but browser tests are pretty important ;) [19:54:57] LeslieCarr: isn't the suggestion to make ssl termination happen on a limited access box enough? [19:55:09] greg-g: the browser tests handle it more gracefully than humans, actually [19:55:12] well the private key will be in the not private repo [19:55:22] so nope [19:55:25] greg-g: but only in Firefox :( [19:55:42] LeslieCarr: not necessarily....it'll be wherever Ryan (or whoever) puts it [19:55:50] * greg-g nods [19:56:58] where is Ryan_Lane, btw? [20:15:08] The current CentralNotice ("Technical maintenance will be performed soon. [...]") will not properly "Hide" for me at any project. Is this on purpose? (It vanishes when I click "Hide", but reappears as soon as navigate to a new page or refresh.) [20:15:43] no, it's just broken, and no one cares [20:15:50] i hid it with site css on pl.wp [20:15:58] https://pl.wikipedia.org/w/index.php?title=MediaWiki:Common.css&diff=37316324&oldid=37104573 [20:16:40] MatmaRex: people care, we're just a little busy right now with the whole 'getting the HTTPS stuff to work properly" bits ;) [20:18:33] Okay, as long as it's known about. Sorry to poke the beesnest ;) [20:18:38] the "https stuff" is not deployed on any normal wiki yet, no? [20:18:39] the banners are [20:18:42] quiddity_: no worries, thanks! [20:18:46] i'dsay they have higher priority [20:19:03] MatmaRex: do you know how they work? do you know what I can do to fix it? [20:19:10] I have no experience with it, any help appreciated. [20:19:52] no, no idea [20:20:00] i hoped you know how they work, or at least who makes them [20:20:20] if iknew, i'd habve proposed a solution [20:20:42] Guillom made the one in question, and he's asleep [20:20:46] the current one is coming from (and translations requested at) https://meta.wikimedia.org/wiki/CentralNotice/Generic_maintenance_notice [20:21:31] parent page has oodles of links. might be what is needed. [20:21:41] greg-g: roblaLUNCH: LeslieCarr: i think the private key could be unpuppetized [20:26:08] sitenotice is broken! if you click it away it re-opens again and again and again [20:26:41] Romaine: known, no one cresponsible cares [20:26:42] Romaine: do you block cookies? [20:26:52] Romaine: i hid it insite css for pl.wp: https://pl.wikipedia.org/w/index.php?title=MediaWiki:Common.css&diff=37316324&oldid=37104573 [20:27:03] responsible*, in site* [20:27:26] MatmaRex: please stop. [20:27:31] MatmaRex: I'm reporting the bug now, you aren't helping. [20:27:55] MatmaRexL I already had shut it off on all wiki's for myself, so annoying it was [20:28:03] MatmaRex: there, you're cc'd on the bug [20:28:19] Romaine: thanks, it is being looked into [20:28:27] for your reference: https://bugzilla.wikimedia.org/show_bug.cgi?id=53116 [20:28:52] thanks [20:29:21] greg-g: can't you kill the notice or something? [20:29:36] MatmaRex: the notice needs to go out, I'm not disabling it all together. [20:29:43] MatmaRex: I'm working on it. trust me please [20:29:59] MatmaRex: patience [20:30:37] this has been happening repeatedly withthe central notices lately [20:31:10] MatmaRex: good to know, can you add that information to the but without being accusatory? [20:31:17] s/but/bug/ [20:32:11] next month a big campaign is coming, I hope it is solved before then [20:32:19] Wiki Loves Monuments [20:33:09] greg-g: i don't think it had a bug reported lat time it happened [20:33:18] also, accusatory? i haven't even accused anybody yet [20:33:58] maybe accusatory is a poor word, but the repeated comments of "no one responsible cares" even after you've seen that A) I can be considered responsible and B) I care, is disheartening [20:34:59] possibly related to https://bugzilla.wikimedia.org/show_bug.cgi?id=17108 [20:35:12] is that notice "emergency" level? [20:35:27] according to last comments these are indismissible [20:36:19] (but back to accusatory part, someone should be told something about setting up things and then going to sleep. ;) ) [20:37:04] MatmaRex: yes, it is marked emergency: https://meta.wikimedia.org/w/index.php?title=Special:CentralNotice&method=listNoticeDetail¬ice=Generic+maintenance+notice [20:37:18] well, unmark it emergency? [20:38:03] and then maybe get someone to finally fix that bug, eh [20:38:15] especially since it seems somehow possible per https://bugzilla.wikimedia.org/show_bug.cgi?id=17108#c4 [20:38:16] MatmaRex: I'm talking with kaldari (who knows CentralNotice well) right now about what to do. [20:41:01] i updated the bugs. [20:41:37] thanks MatmaRex [20:41:51] Bug in Sitenotice. if i klick remove and navigate to a other site the sitenotice is not removed [20:41:57] known [20:42:02] https://bugzilla.wikimedia.org/show_bug.cgi?id=53116 [20:42:03] James A is also helping now [20:42:08] MatmaRex: thanks [20:42:57] greg-g: can't you temporarily unmark it as emergency until further coyurse of action is figured out? :/ [20:43:24] MatmaRex: give it 5 minutes and all will be fixed, best not to get in the way of the person who knows what they are doing. [20:43:33] ~5 minutes :) [20:43:41] I think greg-g means *Greg* doesn't want to get in the way [20:43:52] right [20:44:07] I'd just be fumbling around getting in the way of the experts :) [20:45:12] alright, thanks [20:53:42] greg-g: okay, can i ask about progress now? :) [20:57:19] ah, there's a comment on the bug, nice [20:58:07] * Jamesofur waves [20:58:21] for those here for the https banner: I've turned it off to let the cache clear but also fixed the issue [20:58:38] in 10-15 minutes I will turn it on (after testing) with the ability to hide the banner [20:58:49] thanks much Jamesofur [20:58:54] For those interested in the steps involved I added them to the bug [21:00:06] thanks James_F [21:00:10] thanks Jamesofur :) [21:30:20] wikimedia error by varnish (error 500) on git.wikimedia.org normal or not? [21:36:16] <^demon> Nemo_bis: I'm getting ok responses from it. It's behind varnish now. [21:42:14] ^demon: I guess it's just me asking wrong things? https://git.wikimedia.org/history/operations%2Fpuppet/HEAD/templates%2Fexim%2Fexim4.conf.SMTP_IMAP_MM.erb [21:42:36] <^demon> Ah yeah, I think Gitblit did something lame there before anyway. [21:43:04] also if I search SPF [21:43:16] good thing I can just google it over the repos... oh wait ;) [21:44:18] Heh. [21:44:21] GitHub, FTW. [21:44:41] You can do searches like "sfp @wikimedia" to search Wikimedia's repos on GitHub. [21:44:47] Though the index is a bit out-of-date, I think. [22:17:09] question for documentation purposes. What is the serverside timeout of a session ? the php default of 24 minutes ? [22:17:40] gn8 folks [22:52:09] Saludos! [22:52:13] Alguno habla espa�ol? [22:52:24] Hey! Somebody speak Spanish? [22:52:32] Need help with a bot for upload monuments to wiki page. (For wikilovemonuments) [22:52:46] TDK: uno momento por favor [22:52:52] sumanah, :) [22:53:22] no hablo espan~ol :( [22:53:32] sumanah, Don't worry. [22:54:16] I can read english. [22:54:58] I believe there are also some -es channels, TDK [22:55:27] TDK: have you already tried making the bot? what problem are you having? [22:55:44] I need help getting the monuments that we will use in wlm [22:56:16] * Technical_13 wonders how you run out of English... [22:56:25] lol [22:56:47] I need help to automate the process of uploading to the wiki of monuments we already have [22:57:08] TDK: what format is the data in right now? [22:57:17] a comma-separated file, like a spreadsheet? [22:57:44] In .xls document [22:58:10] spreadsheet. [22:58:15] Yes [22:58:38] We have a spreadsheet with this data. [22:59:09] TDK: have you already looked at https://commons.wikimedia.org/wiki/Commons:Monuments_database ? [22:59:48] TDK: https://commons.wikimedia.org/wiki/Commons:Monuments_database#Create_the_lists_from_the_excel_file [22:59:49] I have the knowledge that there bots to automate the process of uploading to a wiki of these data [23:00:11] TDK: how many rows do you have in your spreadsheet? [23:02:12] 7 [23:03:46] so, 7 monuments in total? [23:04:18] No, no. [23:04:33] ok, how many columns are there? [23:05:29] Have, 22 states. With monuments Total: 6996 [23:05:54] OK, I understand now why a bot would be nice :-) [23:06:25] TDK: Have you already asked the Wikimedia Loves Monuments mailing list or IRC channel for this bot? [23:06:29] I don't know where this bot is. [23:06:34] Dx [23:06:37] Yeah! :) [23:07:02] No. I don't ask. [23:07:30] I don' t suscribe in the Wikimedia Loves Monuments mailing list. [23:08:44] Whitch IRC channel sumanah ? [23:09:20] TDK: I'm looking in https://commons.wikimedia.org/wiki/Commons:Wiki_Loves_Monuments_2013/FAQ#Where_can_I_find_information_about_monuments_database.3F to try to find out. [23:09:48] sumanah, Ok. [23:09:54] And I'm looking at https://meta.wikimedia.org/wiki/IRC/Channels . [23:11:39] TDK: the #wikimedia-commons channel might be able to help you. Also you could simply subscribe to https://lists.wikimedia.org/mailman/listinfo/wikilovesmonuments and ask them. [23:12:40] Aha! [23:12:41] #wikilovesmonuments [23:14:16] TDK: you can join #wikilovesmonuments to ask for help [23:14:28] sumanah, Ok! Thanks! :) [23:17:42] Dmcdevit may be a good resource for TDK. [23:18:24] Amgine, �Dmcdevit? [23:23:47] sv.wikipedia just reached 1.5 million articles ... by bot. [23:25:25] Are you accusing them of cheating? [23:25:35] just reporting facts... [23:25:45] :D [23:25:51] this finally kills the meaning of counting articles [23:26:16] or rather, we should count the articles that people *read* rathern than those people (and bots) write. [23:26:43] why? then only the Facebook article matters [23:27:01] popularity contests aren't useful either ;) [23:27:04] article(s) [23:27:41] I mean: this month, Swedish Wikipedia had 200K articles that anybody cared to read (and 1.8 million bot created stubs). [23:28:02] When is an article not an article? [23:28:23] When a tree falls in the woods and no one reports in Wikipedia.