[00:18:00] !stalk en.wikipedia page .*a.* [00:18:15] Rule added. [00:18:17] [[User talk:Qwyrxian]]; Softlavender; /* Your quick response. */ wlink; https://en.wikipedia.org/w/index.php?diff=571716198&oldid=571716081 [00:18:18] !unstalk en.wikipedia page .*a.* [00:18:31] [[Talk:East African Community]]; BattyBot; [[WP:AWB/GF#Talk page general fixes|Talk page general fixes]] & other cleanup using [[Project:AWB|AWB]] (9466); https://en.wikipedia.org/w/index.php?diff=571716200&oldid=566766575 [00:18:32] [[Talk:Sverre M. Fjelstad]]; LaraBot; [[Wikipedia:Bots/Requests for approval/LaraBot|Bot]]: New page.; https://en.wikipedia.org/w/index.php?oldid=571716202&rcid=600818713 [00:18:33] [[Harry Potter and the Philosopher's Stone (film)]]; MatthewHoobin; /* External links */; https://en.wikipedia.org/w/index.php?diff=571716201&oldid=570176201 [00:18:35] Tina2095 new user account; https://en.wikipedia.org/wiki/Special:Log/create [00:18:35] [[List of Japanese voice actors dubbing Disney characters]]; Funandtrvl; removed [[Category:The Walt Disney Company]]; added [[Category:Disney characters]] using [[WP:HC|HotCat]]; https://en.wikipedia.org/w/index.php?diff=571716204&oldid=569662830 [00:18:37] [[Talk:Verónica Carreón Cervantes]]; LaraBot; [[Wikipedia:Bots/Requests for approval/LaraBot|Bot]]: New page.; https://en.wikipedia.org/w/index.php?oldid=571716205&rcid=600818718 [00:18:37] [[De Bruijn notation]]; Mark viking; Added wl; https://en.wikipedia.org/w/index.php?diff=571716207&oldid=555701005 [00:18:38] [[Template:Obama Executive Office]]; F111ECM; [none]; https://en.wikipedia.org/w/index.php?diff=571716208&oldid=567183757 [00:18:39] [[Shiran]]; Carlossuarez46; create; https://en.wikipedia.org/w/index.php?oldid=571716209&rcid=600818722 [00:18:40] [[Geography of Iceland]]; 107.219.151.33; [none]; https://en.wikipedia.org/w/index.php?diff=571716210&oldid=566350810 [00:18:41] [[Battle of Sardarabad]]; Yerevanci; [none]; https://en.wikipedia.org/w/index.php?diff=571716211&oldid=571714158 [00:18:42] [[Talk:Patrick Hoban]]; LaraBot; [[Wikipedia:Bots/Requests for approval/LaraBot|Bot]]: Added {{[[Template:WPBiography|WPBiography]]}}.; https://en.wikipedia.org/w/index.php?diff=571716212&oldid=571688244 [00:18:43] [[Hizgil Avshalumov]]; Boxes12; [none]; https://en.wikipedia.org/w/index.php?diff=571716213&oldid=571602089 [00:18:44] It seems to be working. ;-) [00:18:44] [[List of legendary creatures from Japan]]; 99.32.18.134; /* N */; https://en.wikipedia.org/w/index.php?diff=571716214&oldid=570973440 [00:18:45] [[Living for the Weekend (album)]]; Lil-unique1; /* Track listing */ fixing credits; https://en.wikipedia.org/w/index.php?diff=571716215&oldid=571715969 [00:18:46] [[Talk:Lee Harvey Oswald]]; BrandonTR; /* Deficiencies: incivility, unsupported assertions, NPOV, etc. */; https://en.wikipedia.org/w/index.php?diff=571716216&oldid=571715904 [00:18:47] It'll stop in a moment. [00:18:47] [[Bratz: The Movie]]; 207.6.120.232; /* Plot */; https://en.wikipedia.org/w/index.php?diff=571716217&oldid=571716103 [00:18:48] Elsie: lol, indeed [00:18:49] [[List of tallest mountains in the Solar System]]; Fotaun; /* Gallery (in order of decreasing base-to-peak height) */; https://en.wikipedia.org/w/index.php?diff=571716218&oldid=561468708 [00:18:49] Jujutacular deleted "[[File:SlamDunk-promo.jpg]]": CSD F5: Unused non-free media file for more than 7 days; https://en.wikipedia.org/wiki/Special:Log/delete [00:18:50] [[List of people known as the Good]]; Clarityfiend; +1; https://en.wikipedia.org/w/index.php?diff=571716220&oldid=476109227 [00:18:51] ... [00:18:51] [[National Law Institute University]]; Orangemike; needs a lot of work; https://en.wikipedia.org/w/index.php?diff=571716221&oldid=571714881 [00:18:52] [[Just Dance 2014]]; Funshine97; /* Track listing */; https://en.wikipedia.org/w/index.php?diff=571716222&oldid=571715987 [00:18:53] [[Talk:Budweiser Clydesdales]]; DKqwerty; /* Soap content */ add comment; https://en.wikipedia.org/w/index.php?diff=571716223&oldid=571714049 [00:18:54] Rule deleted. [00:18:55] domas: Hi! [00:18:58] hi [00:20:04] domas: You successfully killed this channel. ;-) We just debugged an IRC issue in -operations instead. [00:20:20] how did I kill it?! [00:20:23] And by "we," I mean I just complained until someone else found the problem and fixed it. [00:20:34] All the cool kids moved to -operations. :P [00:20:45] oh [00:20:57] thats kind of nice [00:20:57] :) [00:21:21] The /topic of -operations still contains "serious stuff", heh. [00:22:40] \o/ [00:22:52] I guess there's nothing much to talk about wikimedia technology [00:22:55] :) [00:28:51] Elsie: Want to open a bug for MW core? As it should really be fixed there too... [00:29:50] Reedy: I'm not sure what the bug in core would be. [00:30:00] removing the leading space? [00:31:23] the patch that I'd like to see us take is to extend vvv's approach to other log streams [00:32:09] rather than have UDPRCFeedEngine.php delegate to wfErrorLog, rename UDPRCFeedEngine.php to UDPLogEngine.php, and have wfErrorLog delegate to it [00:32:12] and deprecate wfErrorLog [00:32:23] then deprecate UDP [00:32:28] path [00:32:35] Though "patch" was poetic. [00:32:37] path [00:33:58] https://www.mediawiki.org/wiki/Talk:Requests_for_comment/Structured_data_push_notification_support_for_recent_changes [00:36:43] we'd have to ensure serialization was generic [00:37:12] when are we migrating to 5.4? [00:37:29] w/5.4 we get http://php.net/manual/en/jsonserializable.jsonserialize.php [00:37:48] Is "we" MediaWiki or Wikimedia? :-) [00:37:56] ori-l: Deprecate MediaWiki whilst you're on a roll [00:38:07] we could require that every class that wants to be loggable (e.g., RecentChange, MWException) implement jsonSerialize [00:38:47] Ideally, we'd want the last Ubuntu LTS (ie what the apaches would be on) to have the version of PHP.. [00:39:07] 5.3.10-1ubuntu3.6+wmf1 (apache2handler) [00:39:27] you mean for the distro to have it as the default? or for us to have it packaged? [00:39:27] so... After 1404 mebbe? [00:39:42] default ideally [00:39:49] ori-l: I may submit a commit to document that .ltrim(). [00:40:04] PHP 5.4.9-4ubuntu2.2 in 13.04 [00:40:09] Elsie: I couldn't merge it [00:40:14] it's operations/puppet [00:40:16] Right. [00:40:18] There's no rush now. [00:40:38] I'm not really concerned about it being removed, but it's fragile and should probably have a pointer to your commit message. Maybe. [00:40:40] Reedy: ah, I seeeeee. Well, it looks like 14.04 is coming in April, so that's not tooo far away [00:40:50] Indeed [00:41:16] 14.04 in April 2014? That's awfully clever. [00:41:23] ikr [00:43:17] "The x64 builds of PHP for Windows should be considered experimental, and do not yet provide 64-bit integer or large file support" [00:43:18] lol [01:28:35] ori-l: Did you log a bug for your propsed changes? [02:06:37] Reedy: no :/ [02:06:46] I should, though [07:18:51] API request failed (backend-fail-synced): The file "mwstore://local-multiwrite/local-public/6/60/Usedom_St._Marienkirche_2013-08_Turm.JPG" is in an inconsistent state within the internal storage backends at Fri, 06 Sep 2013 07:18:40 GMT served by mw1116 [07:19:28] i cannot work with commons.... every day.... every day.... such errors. [07:21:48] Steinsplitter: sorry, it must be frustrating, but we're fully aware and working on it [07:22:27] ori-l: https://commons.wikimedia.org/wiki/File:Usedom_St._Marienkirche_2013-08_Turm.JPG can you pleas delete this fiel from the database? :-) thx [07:24:42] ok I hesitate to just delete the row or whatever [07:25:32] this was an attempt to do the speedy delete? [07:26:02] Steinsplitter: [07:26:15] is a duplicate [07:26:30] and when you tried to delete you got the above error? [07:27:08] Error deleting file: The file "mwstore://local-multiwrite/local-public/6/60/Usedom_St._Marienkirche_2013-08_Turm.JPG" is in an inconsistent state within the internal storage backends [07:27:15] right [07:27:16] yepp :/ [07:27:50] can you please bz this? I don't know the 'right way' to fix it, nor if there is an ongoing issue, so I'd rather that someone who knows look at it before we just delete things from the back end [07:28:02] you can add me on the cc (ariel@wm.o) [07:30:42] https://bugzilla.wikimedia.org/show_bug.cgi?id=53838 [07:31:34] ok great [07:31:59] thy for helping, apergos :) [07:32:06] thanks for reporting [08:34:31] not [08:35:03] oops. accidental enter press.. appologies for the traffic [08:39:56] irc dos not use a lot of traffic :-) *evil* [08:51:39] Y'know, I hate to say it, but thinking about VE? [08:52:19] ...A lot of the issues seem to come down to a culture divide between coders and the public. [08:53:57] For example, if a coder says "VE is a new HTML-based editor, because we couldn't use Wikitext for what we wanted", the coder means "We need to mark up the Wikitext to clearly mark what's created by templates, where the bold and italics are, and so on, and HTML is the obvious choice because we can display HTML right on the browser." [08:54:16] The public hears. "THEY CAN'T USE WIKITEXT?! THEY HATE WIKITEXT!"" [08:56:04] So... [08:57:51] ...I suppose the question is, how do we bridge the divide? [08:58:43] AdamCuerden: I don't know much about the VE editor but isn't it more like a case "Onoes.. Someone would have to make sure the VE doesn't clutteer the wikicode view with shit code.. that's not gonna be easy..I'm going to the bookies to bet this ain't gonna work.." [08:59:30] AdamCuerden: I like working in wikicode, me personally, it reminds me of my youth [08:59:39] Jubo: There's a reason that example sprang so readily to mind, actually. That was the mistake I made that turned me (temporarily) against VE. [08:59:56] Oh, aye, I still use Wikicode. [09:00:03] I've been on Wikipedia since 2006. [09:00:16] me Spring-Winter 2003 [09:00:31] ...Didn't it only launch in 2004? [09:00:46] AdamCuerden: dohh, no.. 2001 I think [09:01:05] Oh, right [09:02:03] AdamCuerden: I got an RSS item from the aggregator that said "Wikipedia, the free encyclopedia reaches 100,000 articles.", went over to check it out and I was like "Whoa. Awesomeness.." [09:02:08] Heh [09:02:43] I don't remember why I looked at Wikipedia; remember I got hooked when I found a really bad article on my favourite playwright [09:02:47] ...And fixed it [09:02:49] If they wouldn't have let me read my news aggregator I might have been ignorant of the existence of the 'pedia for a longtime [09:02:55] *at work [09:03:19] (It's been an FA since early 2006.) [09:03:38] AdamCuerden: we are cluttering the screens and buffers of people here with our off-topic.. /j #wikipedia-en ? [09:03:56] Sorry, I suppose I could bring it back to topic [09:04:23] I've started writing the tech report for the [English] Wikipedia Signpost. [09:26:16] It makes you think about bridging the cultures, because, I have a little code knowledge, but mainly use it to understand what's being done and take it out to others [09:34:53] I think there are at least two 'public's [09:35:22] one is the established editor who says 'wait what? no wikitext? but but, I have optimized my editing for the last 5 years based on that, plus all my bots and tools' [09:36:13] one is the potential editor who says 'wikitext? what is that? html? what is that? I want to point and click, this is too complicated, what are all these [[ {{ | thingies?' [09:36:58] oh, I would have to go to some wp channel for this discussion... nm [09:37:04] sorry for the noise [09:37:49] it would probably have done a lot for community goodwill if the VE had not been released in a half-finished state [09:38:51] but this appears to be in vogue with "Web developers" these days; release something half-working and force your end users to be your testers [09:39:56] if they complain they're just old fuddy-duddies who hate change [09:40:59] BlastHardcheese: the rapid obsoletion drives the publishing of half-cooked things [09:41:35] if you tune and tune urr product it'll be obsoleted by more advanced and/or economical models from competitors [09:41:47] before it's release [09:43:47] Hmm, so Wikimedia is competing against other groups now? Better start running ads so we can fund more development. [09:53:22] Honestly, I think the problem was Parsoid was really good, but VE - the interface - wasn't as good as they thought it was. [09:53:46] And there wasn't enough hard testing. [15:03:47] apergos: hello [15:03:51] hello [15:04:39] pinged parentabcd [15:04:45] ok, i'll wait [15:10:13] i guess we'll start without him [15:10:53] yep [15:10:59] no response, maybe bad connection [15:11:05] i've been trying to run the trwiki dumps, but they were running out of memory [15:11:23] mm that's not good [15:11:28] turns out, it's because i'm reading from XML and not from mediawiki [15:11:46] because i have to read revision and its text at the same time [15:12:00] you should be able to do tha and fit in memory though [15:12:31] what i do is something like: read all revisions from XML; write page; write revisions [15:12:40] oh all of them [15:12:41] mm [15:12:42] *all revisions of a page [15:12:57] ok well you'll have to restructure that [15:13:09] there are pages with huge numbers of revisions as you know so that's untenable [15:13:16] when reading from MediaWiki, this should work, because i call fetchText only in the third step [15:13:35] sure.. but we want to convert xml to the new format too [15:13:45] hmm, ok [15:14:03] otherwise, we're going to have to what, do a full dump from scratch? for en? [15:14:06] it will take ages [15:14:22] so to jumpstart that, use the xml, convert to new format [15:14:34] i could write the revisions first and the page after that; but that means more seeking when reading [15:14:51] which is why i did it this way [15:15:17] but you're right about enwiki [15:15:19] I feel your pain, as a famous politician once said, but [15:15:31] these things are big and will only get bigger [15:17:03] for now, i've tried to solve it by compiling as x64 (which doesn't have the ~2 GB limit) [15:17:21] heh [15:17:21] but that won't work for enwiki [15:17:24] er no [15:17:54] so i'll just write the page after revisions [15:17:57] ok [15:19:22] so, i'll run these over the weekend [15:19:52] and i'll also test updates with LZMA groups on elnews today [15:21:34] ok great [15:23:29] hello, i am looking for an extremely simple to use wiki to build a platform for people with learning disabilities. I was thinking, maybe wikimedia has a "extremely simple theme" ... ? Or can you recommend anything else? [15:24:42] apergos: i can't think of anything else, do you? [15:24:55] so svick, typically we want to be able to run at least 4 of these in 8 gb leaving memory for other crap like puppet, rerunning some broken step, etc, in general leaner is always better, so if it could be 6 in 8 gb so much the better [15:25:05] nah, it sounds like you have your work cut out for you [15:26:01] if we could do more than the 27 pieces we do now for en wiki, even better (64 gb ram) [15:26:30] right, if i don't store all the text of revisions of a page like i do now, that shouldn't be a problem [15:26:45] how many cores does the machine have? (since compression is CPU-bound) [15:27:56] 32 [15:28:20] but we can't hog them all cause the kernel, someone sshing in and puppet and cron need a little bit :-D [15:28:25] hog 31 of them [15:28:38] and only 8 GB of memory? [15:28:41] oh. [15:28:44] no that host has 64gb [15:28:48] oh [15:28:50] that's the host for en [15:28:58] right [15:29:01] the hosts for the others have 12 cores now with 16 gb [15:29:08] no. [15:29:10] not 16 [15:29:15] emc: define "entremely simple" [15:29:17] meh mixing old specs and new... [15:29:34] emc: you could just take any of the core skins and kill their CSS (styles) :) [15:29:37] 12 cores, 12 gb [15:30:31] MatmaRex: simple to use… i.e. would be great if it was very minimal from the functionality but also still not technical (i.e. use common user interface paradigms) [15:30:41] (like word-like button for editing text etc.) [15:30:56] !visualeditor [15:30:56] VisualEditor is a project to create a simple editor for MediaWiki that would make it far simpler for people to contribute to Wikimedia projects. See also https://www.mediawiki.org/wiki/VisualEditor and #mediawiki-visualeditor channel. [15:31:08] emc: ^ it sorta kinda usually works, you could try it out [15:31:14] what's interesting is that LZMA seems to use multiple threads (i didn't look into that in detail); on my 4-core machine, one process uses about 30% of CPU (something like 1.5 cores) [15:31:31] emc: i don't think any functionality-limited skins exist (at least i don't know of any) [15:31:43] MatmaRex: ok, great… thanks for that, that sounds interesting. [15:32:25] apergos: so, see you monday, i should have all results by then [15:32:36] ok, looking forward to it [15:32:40] have a great weekend [15:34:41] thanks, you too [16:27:31] ^d around? got a git/gerrit problem I've never seen before [16:28:18] <^d> Yeah. Sup? Willing to bet I can guess your problem's root cause :) [16:28:45] [remote rejected] HEAD -> refs/publish/master/update_for_new_release (n/a (unpacker error)) [16:29:03] I found http://stackoverflow.com/questions/4025708/git-cant-push-unpacker-error might be relevant [16:29:38] ^d: ^^ [16:29:45] <^d> Yep. [16:29:54] <^d> So, that's related to the migration. [16:29:57] <^d> There's some bugs open. [16:30:04] <^d> We're still trying to sort things out. [16:31:00] ^d: anything I can do to advance the cause? this change is updates for browser tests to pass against the new VE release that just went out. [16:31:26] <^d> Nothing really. I'm hoping to finish getting things fixed today. [16:37:56] oh, Reedy, what's your flight schedule? When will you be in town? (asking cuz I'm doing the deploy calendar edits for next week, and, curious if you'll be available during Monday's deploy :) ) [16:49:59] greg-g: I'm supposed to be in SF 18:30 on sunday [16:50:04] Barring delays [16:50:27] gotcha [16:50:35] so in the air during the deploy window :) [16:50:43] no..? [16:51:02] oh, you're talking utc? [16:51:15] No, in SF on Sunday, 1830 PST [16:51:30] oh, my lord, where's my second cup of coffee [16:51:44] thanks :) [16:51:55] TripIt is apparently already concerned with my ORD layover. Arrive 1310, leave again 1600.. 3 hours to immigrate claim and re-check baggage... [16:52:07] could go either way [16:52:09] brb [16:54:09] Last time through ORD my plane was very late (it didn't get to my airport for at least 5 hours after it's departure time), and hit at a time with a lot of other flights coming in.. Hopefully not so bad this time [16:58:56] Reedy: "there's plenty of room for improvement"! [17:38:47] are there currently any issues noted with the image servers? [17:39:29] hello hello [17:39:59] I noticed some annoying problems the other day [17:40:16] I was renaming a bunch of images, and it took several minutes -- in some cases, upwards of an hour -- for the files to catch up [17:40:30] Puppy-afk - also, I have no idea if anyone on this channel right now is paying attention [17:40:43] * Puppy-afk bounces annoyinging up and down to gain attention [17:41:14] hello, neverendingo ? [17:43:01] basile ? [17:43:03] * Dragonfly6-7 waves arms [17:43:05] hello hello [17:45:35] * Puppy-afk gnaws quiety on basile's ankles. [17:46:37] * Dragonfly6-7 beginsi setting fire to people on the channel, in alphabetical order [17:47:12] * Puppy-afk starts in reverse alphabetical order [17:47:22] we'll meet in the middle and roast marshmellows. [17:47:32] hello, demon [17:47:54] have there been any recent changes in mediawiki ? [17:48:15] especially those which might affect how amperands are handled in filenames? [17:53:47] hello, mark? [17:53:53] Hi [17:54:02] Can you link to an example? [17:54:07] moment [17:54:16] And have you filed a bug? [17:54:19] https://en.wikipedia.org/wiki/Marcel_Diallo [17:54:26] notice that dead image link? [17:54:28] No, because I'm not sure what happened [17:54:58] Dragonfly6-7: And the file exists on Commons? [17:55:51] No, it does not. [17:55:58] Or if it does, it does under another name [17:57:37] And if the name has been changed, we haven't been able to determine when that happened [17:58:05] OK, so the file _used_ to exist there, but no longer does? And there's nothing in the move log to indicate where it went? [17:58:13] Right. [17:58:42] Or at any rate, I have no idea how to search the move log KTC did [17:59:19] https://commons.wikimedia.org/wiki/File:Diallob%26w.jpg [17:59:34] deleted for copyright looks like [18:00:20] nods, why the freak didn't that show for me when I went looking for the file? [18:00:44] * bd808 blames gremlins [18:00:45] I enter File:Diallob&w.jpg, and I didn't get that page. The ampersand [18:00:51] %26 [18:00:59] * Puppy-afk curses all ampersands [18:01:20] ok, mystery solved. Pity whoever deleted it didn't clean up and remove the link from the article. [18:01:25] * Puppy-afk frowns sternly [18:01:48] and who the hell moved it to commons in the first place [18:02:27] that shouldn't matter, though, Dragonfly6-7 [18:02:40] but deleting a file and then leaving the redlink is a bit sloppy [18:02:51] * bd808 notes that the links provided in the delete banner aren't properly url escaped [18:02:53] and since the & screwed the links up, I wasn't able to determine what happened [18:03:24] ok, thank you so much, everybunny! [18:03:56] who moved it to commons *without realizing that it wasn't free* [18:04:03] and that it should have been left on commons [18:04:37] shrug, not my mystery, mine has been solved. :-D [18:36:48] did red links become redder? [18:45:11] greg-g: I'd like to request an exemption to the no deploy friday rule -- I have a patch that will cool off some centralnotice users tempers -- and I'd rather not wait till monday to push it [18:45:24] mwalker: link plz [18:45:39] it's ready nowish so I'll have the rest of the day to pull it if required [18:45:47] * greg-g nods [18:46:24] it's this https://gerrit.wikimedia.org/r/#/c/82804/2 which depends on https://gerrit.wikimedia.org/r/#/c/79184/12 [18:47:39] mwalker: "A near complete rewrite of the banner class." this worries me [18:47:45] :D [18:47:52] it's worrysome [18:47:57] but I tested the heck out of it [18:49:23] and it's a high enough usage of class that if it does cause problems on the cluster; I'm going to know about it within 30 minutes [18:49:38] mwalker: what kind of testing? [18:49:46] did you have unit test coverage? [18:49:48] :) [18:49:51] I wrote a bunch of unit tests for it [18:50:06] and I walked through all of the functionality [18:50:13] *that I could think of [18:51:48] so, this is so that some people can block some banners based on their own something or other? [18:52:05] now, they're seeing banners they don't want to see and are pissed? [18:52:27] ish; it's mostly so that the WLM people and the LCA people can stop glaring at each other [18:52:42] because right now they have to share a hide cookie [18:53:18] so they're mutually concerned that people are closing the other groups banner and not seeing theirs [18:53:29] ahhhh [18:53:32] grrr [18:53:35] agreed [18:53:35] :) [18:54:32] blugh [18:55:17] mwalker: are there db changes needed? [18:55:24] this patch deprecates the cn_templates.tmp_fundraising column in favour of cn_templates.tmp_category. [18:55:26] I was just about to mention that :) [18:55:40] yes -- the schema change works in both directions [18:55:52] e.g. the presence of it wont cause an older version to break [18:56:26] and without the schema change nothing in the new version breaks? [18:56:43] without the schema change the new version wont work [18:56:51] and the schema change has been made? [18:56:57] nope [18:57:03] not yet [18:57:13] (I'm really making my case here arent I? :) ) [18:57:17] :) [18:57:50] asher's gone, py's gone, springle is asleep.... [18:58:10] oh wait, no, asher's still here, right? [18:58:21] not online, eitherway [18:58:24] yep [18:58:33] no dba's present [18:58:53] blugh.... [18:59:01] I have to say no, I think, man. :/ [18:59:06] kk [18:59:14] have the annoyed masses email me [18:59:15] :) [18:59:21] you really dont want that [18:59:24] hahaha [18:59:37] greg+straighttotrash@wikimedia.org [19:00:05] oh cause that's not obvious at all :) [19:00:18] hide it in a element [19:00:22] ;) [19:00:46] so! can I get a slot monday then! [19:01:07] :) yeah [19:01:32] LD is open, or early morning (if you can get the db changed in time) [19:02:27] the change is fast to deploy; but we might have the same problem with no dba's early in the morning [19:02:43] yeah :/ [19:02:49] I mean, springle'll be awake [19:02:54] i was going to say springle is in the aus [19:02:55] well, the night before [19:02:58] But is he actually in the office? [19:03:00] "night" [19:03:02] oh right, dangit [19:03:18] where is everyone's flight info? [19:03:46] no on their respective calendars, of course :/ [19:03:49] not [19:03:52] Stupid Wikipedia trick: Once VE's copy-paste arbitrarily formatted text is working, paste Time Cube into [[Time Cube]] as a test. [19:03:53] I haven't seen it altogether [19:04:20] greg-g: Go poke travel? [19:04:58] Reedy: I guess..... [19:05:02] probably easiest just to do it in the LD then [19:05:14] ksnider isn't on irc to poke [19:06:03] mwalker: yeah, let's plan on that, or lunch. or morning after Reedy's done since that window is rarely ever fully used [19:06:52] kk; I'll add the changes to the LD then and we can move them to when it actually happened when we determine when that is [19:08:30] word [19:14:12] mwalker: thanks for humoring me and all [19:14:22] no worries [19:14:38] it's nice to have someone to hit you over the head when you really want to do something silly [19:15:23] :) [20:36:09] sorry to pester again about the API, but I have another question, how should I encode my text string for the MD5 parameter in action=edit? [20:56:05] MartijnH: I guess it's whatever you put in the text string - you could verify by looking at the source code, and maybe submit a patch for the documentation to clarify? [20:56:32] I can take a look, but my PHP is really bad [20:58:27] so I might ask some more questions later on PHP handles that kind of stuff [20:58:31] MartijnH: You can also omit that parameter - if it's not set it won't affect the behaviour [20:59:43] Hey. Is anyone with access to the servers available and willing to help me dig on https://bugzilla.wikimedia.org/show_bug.cgi?id=43046 ? [20:59:57] I have some contradictory results from my runs :) [21:01:00] it takes the string as it arrives from the POST, but I assume that it does some internal transcoding [21:01:08] the question is, to what [21:01:43] strainu: You might try in wikimedia-operations and be a little more verbose about what the problem is [21:02:09] marktraceur: thank, will try. I can't be more verbose, I don't know much more for the time being [21:02:26] I mean [21:02:29] If I knew I would have wrote a comment in the bug [21:02:32] Say the words "gateway timeout" [21:02:48] oh, ok :) [21:24:55] * greg-g wonders over here to avoid crypt/nsa discussion [21:25:23] manybubbles: ^d : when do you think "as soon as possible" is for mediawiki to have the new search as primary? [21:25:57] greg-g and ^d: early week after next [21:26:12] gotcha [21:26:19] after we're back and settled [21:26:21] <^d> I was thinking of doing it sometime next week so people could bang on it during tech days :) [21:26:31] also, are there any specific non-english wikis you want to test on before rolling out more widely? [21:26:58] <^d> We're working on a list. [21:27:12] * jeremyb wraps greg-g in a jumbo frame [21:27:24] jumbo frame? [21:27:27] <^d> Ooh, manybubbles, how is greek support? [21:27:46] ^d: claimed [21:28:00] <^d> Yeah, I'm seeing it in docs now. [21:28:13] http://sais.se/mthprize/2007/ntais2007.pdf [21:28:26] <^d> I'm wondering if apergos would know of any el wikis that would want to play guinea pig with search :) [21:28:38] hm [21:28:53] I think you could ask the el wp folks, I don't know off hand or not [21:29:05] greg-g: huh, i never saw a talk page (ns1) rank so high in google results before. https://www.google.com/search?q=jumbo+frame+ipsec [21:30:27] <^d> manybubbles: So, maybe we could turn mw.org to primary backend like...monday? We should rebuild the index probably between now and then. [21:30:32] jeremyb: ah, what it sounded like [21:30:55] how about tuesday? Announce today. [21:31:47] <^d> I'm down with that. [21:32:04] don't want to spring it on everyone on monday without an announce :) [21:34:14] ^d and greg-g: sure! let me kick of an reindex on beta so I can reverify it soon to make sure there isn't something unexpected going out. [21:35:15] So, I'd love for Search deploys to be well announced, so if we can't confirm and announce today, then lets delay until post all hands. [21:36:14] greg-g: lets confirm next monday and announce then. [21:36:22] ^d: ^ [21:36:35] I feel comfortable with that assessment [21:36:38] <^d> Let's doooo iiitttt [21:36:45] :) [21:36:47] pusher [21:37:31] ^d: beta'll be reindexed later tonight and I can check it then. We can announce this monday if you really want [21:37:55] <^d> We'll announce today to do it tuesday. [21:38:14] <^d> If beta looks good, I'll reindex mw.org monday. [21:38:21] <^d> And we'll throw the switch tuesday. [21:38:42] by monday you mean next monday, after all hands? [21:39:24] <^d> I mean this monday, week of all hands :) [21:39:42] <^d> Since we'll have a ton of people around, maybe we can convince people to bang on it during tech days. [21:39:59] not a bad idea [21:40:31] also, I've been talking to Amir about other languages too - his home wiki is Hebrew but we don't support that out of the box. [21:40:41] and we have [21:40:49] haven't heard any other good candidates [21:42:08] <^d> Hmm. https://github.com/elasticsearch/elasticsearch/issues/3395 [21:43:00] <^d> Leads me to https://github.com/synhershko/HebMorph [21:43:09] <^d> Promising, it's actively maintained. [21:43:12] so, if you want to announce today, do it soon. But, please do Tuesday, Monday is pretty full now (due to the short week) [21:43:39] <^d> Will do, I'll get an e-mail sent out within an hour. [21:43:46] word, thanks [21:43:53] ^d: HebMorph isn't really done, iirc [21:44:06] * greg-g forearm muscles are starting to get tight, I should stop typing and stretch for a bit [21:44:14] <^d> But if it's active it could be something worth helping along :) [21:45:01] <^d> brb. [21:46:27] ^d: maybe. It'll need a less crazy license for us to use it, though. [21:47:28] ^d: though it does look reasonably active [21:49:40] what's wrong with the AGPL? [21:49:46] manybubbles: oh, this part? "The hspell dictionary files distributed with HebMorph are provided with [21:49:50] the license to be used ONLY for search by HebMorph. To get an official [21:49:52] hspell distribution under the GPLv2 license, visit their site." [21:51:04] greg-g: I didn't even see the weird hspell thing. I'm not up on my AGPL knowledge but doesn't using it imply accepting some odd license terms? [21:53:21] manybubbles: simply, it closes the webpage loophole in the gpl. By that I mean: [21:53:46] If you use AGPL'd code in a web app (or something else a user interacts with), you have to make the code available to the user. [21:53:57] with the GPL, you don't [21:54:13] <^d> Well I wouldn't mind that :) [21:54:19] since all our code is open and available, we're compliant [21:54:39] and AGPL is compat with GPL [21:55:10] Is it OK to link to Apache licensed code? [21:55:24] yah [21:56:14] in that case, cool [21:57:56] ^d: gonna head out now - 6 on a Friday and the family wants me to come have fun [21:58:13] <^d> Enjoy your weekend. You get in monday? [21:58:23] ^d: yeah, I'll be in around lunch [21:58:28] <^d> See you then. [21:59:02] you too [22:31:20] Reedy: around? are you in a position to merge https://gerrit.wikimedia.org/r/#/c/83012/ ? [22:53:10] if someone happens to pass by who knows the details of how post variable strings are encoded, I wouldn't mind hearing about it