[09:06:09] http://labs.wikimedia.beta.wmflabs.org/wiki/Problem_reports [09:06:28] wondering if js changes are mentioned on 1.19 page [09:06:38] because people are really having troubles there [10:37:07] can I get a bit of subversion help? [10:38:04] I need to duplicate svn://svn.wikimedia.org/svnroot/mediawiki/trunk/lucene-search-2 to svn://svn.wikimedia.org/svnroot/mediawiki/trunk/lucene-search-3 [10:41:00] OrenBo: probably: svn copy URL1 URL2 [10:41:24] or you can make a tag [10:41:47] which is a copy actually :-) [11:40:31] hashar: I get network connection closed unexpectedly [11:40:52] OrenBo: probably: svn copy URL1 URL2 [11:40:55] that should work [11:41:18] yep - I lookd it up [11:41:24] you are right [11:41:28] that is how we do tag [11:41:57] ie: svn copy svn://something/trunk svn://something/tag/version1 [11:42:21] not I copy svn+ssh://oren@... [11:47:20] with svn+ssh I get the above error [11:47:55] wih just svn:// I get connection refused [12:14:58] OrenBo: what error do you get? (I don't see it in my scrollback) [12:27:15] with ssh urls I get network connection closed unexpectedly [12:27:33] with svn: urls I get connection refused [12:33:00] <^demon|zzz> OrenBo: svn+ssh:// [12:34:09] <^demon|zzz> svn:// won't work at all. http:// doesn't auth, it's anon only. [12:34:30] <^demon|zzz> If svn+ssh:// isn't working, I'd confirm that your key's in your ssh agent [12:48:19] Localisation and internationalisation bug triage in 15 minutes in #wikimedia-dev. See http://etherpad.wikimedia.org/BugTriage-i18n-2012-01 for details. [12:50:35] <^demon> siebrand is canvassing! [12:59:13] ping 1m [12:59:26] hi srikanthlogic [12:59:36] santhosh: hello! [13:00:00] Etherpad: http://etherpad.wikimedia.org/BugTriage-i18n-2012-01 [13:00:03] pong! [13:00:03] *srikanthlogic was confused when gerard blogged this triage is at 1500 UTC [13:00:17] oh. Did he? [13:00:29] my UTC click is blacked out [13:00:32] clock [13:00:38] siebrand: http://ultimategerardm.blogspot.com/2012/01/mediawiki-i18n-triage.html [13:01:06] srikanthlogic: Please fill out yout nick on the etherpad [13:01:54] hexmode: done :) [13:02:12] Anyone want to suggest some webfonts before we dive into Narayam? [13:02:30] srikanthlogic: thanks for the notice on tawiki :) [13:02:55] hexmode: more the merrier :| but sadly few takers.. [13:03:14] could you put the URL where you posted it on etherpad? [13:03:16] hexmode: our community want some special support for chinese in webfonts [13:03:43] the situation is, users don't have font for some rarely used chinese characters [13:03:43] liangent: is there an open bug? Are their font files for this? [13:04:16] if we add a font file containing all chars, it can be too huge [13:04:38] liangent: cool. We know Chinese is very complex. Can you briefly describe, and can we then continue outside of this triage? [13:04:41] liangent: do you have any alternate suggestions? Size is a problem for webfonts? [13:04:50] ideally it can pick those rarely used chars used in the article, and send the font file only with these chars to users [13:05:29] run time created font files? [13:05:38] ugh [13:05:47] or deliver that rare chars and depend on font fallback? [13:05:48] hey sorry being late [13:06:00] maybe, or create when parsing the page [13:06:01] or, one individual file per character? [13:06:03] *siebrand welcomes hashar [13:06:13] do you have any conf call setup or is triage only handled in this channel + ether pad ? [13:06:19] hashar: we just started [13:06:27] hashar: the latter [13:06:39] santhosh: of course it should be working in this way [13:07:08] only on channel + etherpad... though maybe we should try google hangout in the future? [13:07:24] I think having 40k or so files for Chine characters might be a tad inefficient (pun intended) [13:07:50] I bet 40k is not enough for all rare chars [13:07:59] liangent: how many chars are there of these "rare ones"? [13:08:15] hexmode: I guess most of chars above BMP [13:08:22] exactly??? It's worth exploring this, because we've been pushing zh web fonts forward because it was "too difficult". [13:08:26] hundreds of thousand? [13:08:51] surely not [13:08:56] I agree we should explore later. Can we agree on that? We'll make an appointment of when to discuss this this week or the next. [13:09:14] Nikerabbit: much more that that [13:09:24] siebrand: ok [13:09:30] yes, if liangent can help, we can have a test page and get a better idea about what can be done [13:09:33] *hexmode gets scared [13:09:39] next! [13:09:54] * https://bugzilla.wikimedia.org/31904 -- Bamini keyboard map needs fix [13:10:01] liangent: can you drop me a mail at smazeland AT wikimedia.org, so I have your e-mail address? [13:10:14] *srikanthlogic is trying to find bamini users. [13:10:15] liangent: I'll get back to you with a proposal for date and time. [13:10:41] siebrand: you must have my email addr as you sent me money several times... [13:10:55] liangent: heh :) I'll look it up. [13:11:02] bamini is used by mostly anon / old school users from srilanka. most logged in users (whom we can easily communicate) dont use bamini [13:11:25] Does it have its own script? [13:11:33] srikanthlogic: do you know any with JS skills? Or is that too "new school"? [13:11:41] siebrand: its just a input method [13:11:52] siebrand: It is Tamil inputmethod- Tamil script [13:12:03] ah, I didn't know we had moved on from webfonts.... [13:12:03] hexmode: i know some JS, but i dont know the input method // i only use transliteration [13:12:37] siebrand: anyway I sent you a mail' [13:12:54] hey aharoni [13:12:59] tried mailing the user who reported issues, but no response :( [13:13:01] srikanthlogic: if we send a mail to tamil wiki mailing list , will it help? [13:13:06] aharoni: ! [13:13:33] hi hexmode, siebrand, srikanthlogic [13:13:39] srikanthlogic: :( can only solve issues if we have some back-n-forth [13:13:41] santhosh: nope, I had even sent once to thamizha, no reply //most 'savvy' users dont use it.. but supposedly widely used by old school srilankans [13:14:03] hexmode: agree, lower priority, I will continue to hunt someone .. [13:14:31] ok, next... [13:14:32] aharoni: happy birthday! [13:14:41] thanks :) [13:14:54] https://bugzilla.wikimedia.org/32029 -- Some vowel combination in Sinhala Wijesekara need to be corrected [13:15:04] next ^^ [13:15:31] This is also pending without any feedback from community [13:15:52] lack of documentation on the input method is the isssue. [13:16:27] For everyone's information: the next iteration of the Narayam extension features centralized help (with the possibility to override that on local wikis). [13:16:49] Tamil can for example point Narayam from Wikipedia and all sister projects to the Tamil Wikipedia help page. [13:16:50] one probable solution is to ping sinhala maintainers at Red Hat [13:16:54] siebrand: added the last point on Narayam related to that.. [13:17:04] https://www.mediawiki.org/wiki/Help:Extension:Narayam [13:17:16] siebrand: we chose these bugs hoping to get more people from the community here, right? So maybe focus on things srikanthlogic can help us with? [13:17:30] yep, fine with me. [13:17:48] anyone else here that isn't a dev? [13:18:03] hexmode: liangent [13:18:09] *hexmode sees a few nicks he doesn't recognize [13:18:17] (well, not "dev-ish" :) [13:18:49] siebrand: [13:19:18] srikanthlogic: I don't understand what you meant by "added the last point on Narayam related to that..". Or was that the addition of "hexmode: agree, lower priority, I will continue to hunt someone .." in the pad? [13:19:43] santhosh: lack of documentation for 32029 (Sinhala)? [13:20:02] siebrand: not bamini.. // i was mentioning about ** Script to automate transliteration help maps (like http://www.mediawiki.org/wiki/Help:Extension:Narayam/Tamil/Transliteration) for Narayam help [13:20:32] siebrand: I didnt log a bug yet for it, but felt it might be good thing to do, esp since community isnt all that active in many languages [13:20:43] hexmode: yeah, 32029 need documentation or feedback from community [13:21:10] srikanthlogic: arch. That table look horrible. Use in script! [13:21:12] santhosh: you said in the bug that you can refer to layouts used by linux? I'm confused by what you need re feedback [13:21:50] ^demon: svn+ssh:// fails as well [13:22:03] siebrand: apparently tamil folks dont use inscript. // no one knows to type, even the 'developers' [13:22:15] how do I check about auth - I got ssh forwarding in my ssh client [13:22:18] yes, comment 3 of that bug asks some questions, by feeback, I meant answer to that [13:22:31] <^demon> OrenBo: What's the exact command you're trying to do? [13:22:32] and I can see it worjs [13:22:33] and I can see it worjs [13:22:38] sec [13:22:44] if we dont get it, refer the tables in linux. [13:22:48] siebrand: table can be beautified, but felt such might be required.. after hearing from a newbie at a outreach session [13:22:51] :( You guys are so severely hampered by a lack of (adopting/adopted) standard. I admire that you still get stuff done in native languages. [13:23:16] srikanthlogic: why a script to create that table? It's not changing all the time, is it? [13:23:16] aharoni: are you on the etherpad? If so, could you identify your nick? [13:23:22] tamil99 is a standard adopted in tamil. inscript was a failure [13:23:25] (so once created, it just _is_) [13:23:32] svn --username oren copy svn+ssh://svn.wikimedia.org/svnroot/mediawiki/trunk/lucene-search-2 svn+ssh://svn.wikimedia.org/svnroot/mediawiki/trunk/lucene-search-3 [13:23:36] siebrand: not for tamil. but for all other languages [13:24:03] *srikanthlogic assume vowel / consonant combination is similar [13:24:10] santhosh: can you confirm ? [13:24:18] I think that creating transliteration tables may be a manual job in any case, because there may be reasons to present them differently for whatever linguistic reason in each langauge. [13:24:23] what's the etherpad url? [13:24:24] Or does that sound dumb? [13:24:24] <^demon> OrenBo: Drop the --username oren and try svn+ssh://oren@svn.wiki.... [13:24:33] http://etherpad.wikimedia.org/BugTriage-i18n-2012-01 [13:24:40] aharoni: ^^ [13:24:51] siebrand: atleast when i did tamil i felt its a mundane thing and could be automated [13:25:00] siebrand: I'd be interested in transliteration schemes [13:25:06] you develop [13:25:21] OrenBo: hey. [13:25:24] hey [13:25:29] OrenBo: we have a shitload of them already. [13:25:30] yeah, transliteration cannot be automated, it is regex rules and cannot get meaningful table information [13:26:00] OrenBo: they are in https://www.mediawiki.org/wiki/Extension:Narayam [13:26:09] whoever hasn't identified on the etherpad, please do. [13:26:28] It a bit down the road but I''d like to store/index transliterate foriegn words/names when analyzing [13:27:03] santhosh: oh! i felt it follows patterns across //my language skills are limited to tamil [13:27:15] ;P [13:27:35] anyway, we are creating documentation for Wikimedia deployed key mappings now. [13:27:52] The documentation is scheduled to go live next Monday. [13:28:07] it's still WIP and help is appreciated :) [13:28:18] should we continue to the next issue, hexmode ? [13:28:23] yep [13:28:24] https://bugzilla.wikimedia.org/33243 -- Narayam IME on fails to replace english characters on mobile [13:28:50] that got atleast 2 different bugs //might need to be split [13:29:10] ls [13:29:30] srikanthlogic: yeah, if you could help me split it up, that would help [13:29:57] hexmode: i needed the cases to be tested on different handsets to start identifying issues before it could be split [13:30:09] ah [13:30:34] else symbian bugs will always be treated badly ;) [13:31:11] ^demon: that also does not work - I get connection closed unexpectedly after 5 seconds [13:31:28] <^demon> I really don't know OrenBo :\ [13:31:42] I've even seen strange behavior of iPad input in Lqt. Mobile is horror in the sense that it even has more operating systems than PCs. [13:32:58] Anything more to be said on https://bugzilla.wikimedia.org/33243 -- Narayam IME on fails to replace english characters on mobile now? [13:33:29] ^demon: can you try from your end ? [13:33:32] siebrand: or should we be doing something different than Narayam for mobiles / tablets ? [13:33:40] (we're about half way the session, and have covered 25-30% of the prepared issues :P) [13:33:56] srikanthlogic: don't know yet. [13:34:04] <^demon> OrenBo: Will try, sure. [13:34:04] something more suited for touch maybe //wild guess [13:34:14] srikanthlogic: I'm planning to discuss that further with the mobile product managers at wikimedia. [13:34:38] next: https://bugzilla.wikimedia.org/33300 -- Unwanted activation which disables user's ability to type [13:34:45] here it is less the user name :: svn copy svn+ssh://svn.wikimedia.org/svnroot/mediawiki/trunk/lucene-search-2 svn+ssh://svn.wikimedia.org/svnroot/mediawiki/trunk/lucene-search-3 [13:35:08] <^demon> OrenBo: https://www.mediawiki.org/wiki/Special:Code/MediaWiki/109361 [13:35:15] more configuration is a solution to it? [13:35:38] the diversity of screen sizes, resolutions and aspect ratios is so diverse, that a standard solution like that for desktop will probably not ever work. [13:36:03] siebrand: and then languages themselves ;) [13:36:22] well, first the framework... [13:36:32] then one language at a time... [13:36:45] hmm.. [13:37:03] making things bigger makes them exponentially harder to solve, so we like to keep our stories small :) [13:37:19] :) [13:37:31] so, for 33300: is it possible to have an input toggle attached to the currently focused input field? [13:37:54] hexmode: yes, but most(atleast in commons) dont know the hot key [13:38:26] santhosh, what's your take? [13:39:04] siebrand: are these tranliteration schemes phonetic or arbitry ? [13:39:08] srikanthlogic: I was thinking of something visual that would be next to the input instead of the current small icon in the heading [13:39:13] well, its not only textarea, but inputboxes too. so we need to be clear about the UI [13:39:18] OrenBo: both [13:40:11] I see [13:40:12] hexmode: that was the old design. apparently the new UI was suggested as UX improvements [13:40:23] OrenBo: http://svn.wikimedia.org/viewvc/mediawiki/trunk/extensions/Narayam/Narayam.i18n.php?revision=109231&view=markup line 26 and on [13:40:54] srikanthlogic: ah... I didn't realize that. [13:40:57] OrenBo: looks like only have one specifically tagged phonetic [13:41:20] k, this bug is too much design and we don't have any comments on the bug from anyone about that [13:41:24] santhosh: i think even with new UI, this might still exist, should we not enable by default and just load on common wikis like commons ? [13:41:42] indeed ^ [13:41:48] maybe I'll get one after today when the enwiki blackout is done [13:42:20] final narayam bug: https://bugzilla.wikimedia.org/33480 -- Add Telugu Transliteration input method to Narayam [13:42:31] I'm not for enabling input methods on multilingual projects automatically, by the way... [13:43:14] for that one, it looks like I need to find a tewiki person to help [13:43:16] moving on [13:43:18] I refer here to my own experiences that it's very disorienting when visiting wikis as anonymous, and finding that I'm typing in a different script. [13:43:25] anu Telugu speakers here? it's really easy to do. [13:43:44] Telugu has this inputmethod in js already [13:43:50] *srikanthlogic lives in telugu land but doesnt know to speak [13:43:53] we just need to port it to Narayam [13:44:03] so, even easier. [13:44:12] srikanthlogic, you don't need to speak. [13:44:24] I think I'll just have to go on-wiki and find the person(s) who wrote it [13:44:31] you only have to be not afraid of foreign letters :) [13:44:45] hexmode: the person left was what i heard [13:44:56] srikanthlogic: ah, I see [13:45:08] aharoni: :) [13:45:59] Moving to translation [13:46:26] https://bugzilla.wikimedia.org/31632 -- When re-marking translatable page for translation, old version of the pages might be shown [13:47:01] I think that's been resolved with the latest deployment (?). Nikerabbit ? [13:47:19] *hexmode adds chasing down tewiki user/dev to his todo list [13:48:12] Nikerabbit is ignoring us :( [13:48:23] *hexmode hopes pings bring Nikerabbit [13:48:25] *siebrand pokes Nikerabbit [13:48:26] :) [13:48:28] next... [13:48:32] we'll get back to this one. [13:48:41] or should i provide a little more background? [13:48:42] ** https://bugzilla.wikimedia.org/31695 -- Support Google Translate V2 API [13:48:50] I'm here [13:48:55] I expect the external translation system to be wikimarkup-aware [13:48:56] On Google Translate V2 API. [13:49:11] Google has deprecated its free Translate API recently. [13:49:19] siebrand: I haven't confirmed myself, but haven't receive any new reports about that issue [13:49:25] It may still be there, but it's been capped, and no longer usable for translatewiki. [13:49:49] v2 is a paid service, but we might convince Google to have translatewiki and Wikimedia use it for free. [13:49:53] I prefer not to use the autotranslate because it messes up tags and I have to clean it up manually [13:50:04] is the v2 API something we need to talk to goog about? [13:50:12] actually, I've been in contact with a Google product manager, and he already promised it for at least transltewiki. [13:50:28] siebrand: k, so you're on that [13:50:42] liangent: we call MT translations suggestions. You can insert them, and clean them up, or not use them at all... [13:50:46] implmentation... is it much different? [13:50:52] hexmode: unknown. [13:51:03] If there's someone who'd like to work on that, that would be great. [13:51:15] I'm gonna live in a fantasy world where that is trivial [13:51:17] :) [13:51:20] Currently only Bing Translator and Apertium remain for machine translation. [13:51:35] next https://bugzilla.wikimedia.org/32983 -- Page protection leads to issue in translatable pages [13:51:39] ;) [13:51:39] siebrand: yeah but usually I find cleaning up is more complicated than writing a new one [13:51:51] are people really using it? [13:52:16] liangent: could be true for Chinese. For Dutch it's a real time saver, but not as big a time saver as translation memory is. [13:52:34] I don't understand 32983 from the summary it doesn't make sense. [13:52:38] *hexmode re-reads [13:52:39] on page protection: [13:52:47] Translate triggers on content edits. [13:53:02] page protection is not an edit and throws special:page translation off. [13:53:13] Nikerabbit: what kind of help is needed on this? [13:53:29] let's see if I can create an example.. [13:53:51] siebrand: Nikerabbit: looks like something someone like ^demon or Roan could help with [13:54:03] if only to provide information and insight [13:54:29] siebrand: Translate should probably listen on more hooks about those events [13:54:40] or a better hook [13:54:59] <^demon> Huh? [13:55:04] *^demon hasn't been following anything [13:55:48] ^demon: your help may be needed after we're done with the triage. [13:55:55] ^demon: https://bugzilla.wikimedia.org/show_bug.cgi?id=31632 [13:55:56] ^demon: could you look at https://bugzilla.wikimedia.org/32983 [13:56:14] k, final one! [13:56:15] ^demon: sorry, my link as bad. [13:56:43] https://bugzilla.wikimedia.org/33647 -- Translate popups with insufficient height [13:56:59] <^demon> Bug 32983 I don't know how much help I'd be. page protection code is one of the few things I've never touched. [13:57:22] ^demon: k, just hoping you had :) [13:57:58] this seems to happen from time to time [13:58:05] ^demon: do you know who would understand it well? [13:58:20] Nikerabbit: "this"?? [13:58:22] *^demon shrugs. Aaron? [13:58:23] the dialog is sometimes like 100x100px for some reason [13:58:48] I use $( window ).width() * 0.9 to set the width for example [13:59:08] haven't been able to reproduce myself consistently [13:59:13] ugh... so maybe a few more reports here would help? [13:59:20] Nikerabbit: JS ninjas like crinkle have helped out yet? [13:59:27] and a way to produce it consistently? [13:59:49] darn autocorrect... [13:59:57] heh [14:00:03] hexmode: it's not so much about the way but about particular users [14:00:19] I tested this in IE6, but the size was not 100px anyway, it was narrow and positioned in a wrong place. but IE6 [14:00:32] I think Nemo_bis is able to reproduce it [14:00:44] Nikerabbit: could the Assistant languages be involved? [14:00:48] no [14:00:48] not consistently [14:00:49] (a preference) [14:01:00] it seems to happen randomly [14:01:05] Nemo_bis: how often do you see it? [14:01:14] santhosh: it should not happen in IE6 at all, I only set max-width and IE6 should ignore it [14:01:20] on TWN I have fr,es,pt; on sandwiki fi [14:01:22] we're 1 minute over the scheduled time. Should we start wrapping up? [14:01:49] sometimes I'm not able to reproduce it even in 500 messages, sometimes it happens every time... [14:02:05] it's very weird [14:03:00] siebrand: yep... I think this has been good, but we do need more participation from wikis. :( [14:03:17] thanks, srikanthlogic, for your input :) [14:03:32] hexmode: happy to help [14:03:43] siebrand: when are we going to discuss on Univ lang sel ? [14:03:55] Thanks Nemo_bis, Nikerabbit, aharoni, Nikerabbit, santhosh, siebrand, srikanthlogic, ^demon, OrenBo, liangent for your contributions. [14:03:55] seems like many issues on narayam / webfonts related to it. [14:04:01] srikanthlogic: not in this session. [14:04:03] *hexmode is off to tewiki to hunt [14:04:22] srikanthlogic: it's basically just tucking those features away in an options screen. [14:04:48] we've had another preliminary design this week that's not been shared widely, not even in my own team. [14:05:16] siebrand: ah ok.. good to wait if its something in the eyesight. [14:05:48] But it looks like there may come a single simple O-shaped icon. What's the name in English... [14:06:12] Gear [14:06:22] ah. just like google :d [14:06:24] Many application and sites use it, GOogle, too, IIRC. [14:07:30] The screen we will use may be a little larger, and with grayed out background, but I think that's the way I would like it to go. [14:07:58] As for selecting a user language: no clarity yet, same for how to fit web fonts preferences and IMEs in there. [14:08:19] I think it would be cool if that gear would "follow" in the corner of a page. [14:08:41] This might mitigate the "need to scroll to top of page to change settings" issue [14:08:52] hmm [14:09:03] so something, but nothing we're developing... [14:09:16] I think this might take another month or two before we actually get to developing. [14:09:44] I've been promised proper UI design resources a few times, but apparently those people are overcommitted. [14:10:01] and have not been able to come up with something I could get started with. [14:10:05] oh! [14:10:37] Anyway, as soon as there are designs, we'll most definitely share them on mediawiki.org. [14:10:50] Keep https://www.mediawiki.org/wiki/Universal_Language_Selector on your watchlist. [14:11:03] cool.. added [14:11:04] [14:11:11] Thanks again. [14:11:15] siebrand: do you have time to talk about that webfont issue now? [14:11:21] liangent: nope [14:11:23] thanks! [14:11:34] liangent: didn't have much sleep last night and am actually really tired. [14:11:48] liangent: but if santhosh is up for it, please feel free to do it without me [14:12:49] Nikerabbit: Jenkins now runs PHPUnit tests for the translate extension :b [14:13:05] hashar: and they work? [14:13:08] nice [14:13:08] yup [14:13:17] sent you the link to a build result [14:13:22] link is probably outdated though [14:13:23] *siebrand jumps up and slaps a thundering high5 with hashar ! [14:13:29] *hashar high5 [14:13:40] *hashar offers a round of beer (and coke) [14:14:05] so you should all be able to add ton of tests for Translate [14:22:22] *siebrand nods. [14:22:37] Nikerabbit: those page translation parser tests, they're not run as part of the unit tests? [14:24:23] siebrand: the unit tests? [14:24:31] they are not included if you run unit tests for core [14:24:59] Nikerabbit: is that somehow possible? (because I don't think we run those test regularly, do we?) [14:27:09] Nikerabbit: ergo, this is all that's run now, right? https://integration.mediawiki.org/ci/job/MediaWiki-sqlite-phpunit/5948/testReport/%28root%29/MessageGroupBaseTest/ [14:28:05] hashar: url for translate extension tests? [14:29:02] the parser tests might not be in the unit test format, can't remember [14:30:13] https://integration.mediawiki.org/ci/job/MediaWiki-sqlite-phpunit/lastBuild/testReport/(root)/MessageGroupBaseTest/ [14:30:31] /lastBuild/ means you can bookmark it [14:30:36] until someone change the job name :b [14:30:44] :) [14:31:07] hashar: I was able to log in to Jenkins and from what I can see I have write access to the config. Scary :) [14:31:32] yeah everyone has access [14:31:44] another wiki engine... [14:32:02] liangent: oing [14:42:45] 10,269 Comments is just insane... [14:48:03] Especially when you realize that each one of them was manually approved. [14:49:59] :o [14:50:21] now molliug will say that he was right when I asked to open comments [14:54:49] Well, we're not getting too much "really bad" stuff (about 5% spam/trash), but those 5% are really horrible and we really wouldn't want those to be on our blog. [14:56:01] Yes, I only asked not to require registration. [14:56:22] Without that there would probably be way less comments to approve. :-p [15:02:23] hehe, true [15:02:43] But it's nice to give people a venue to express themselves. I don't regret it. [16:17:59] OrenBo: what is the bazzar module in lucene-search-3 ? [16:18:17] btw, the first build in jenkins : https://integration.mediawiki.org/ci/job/Lucene-Search-3/2/console [16:18:18] ;) [16:18:32] note jenkins use "maven --offline" [16:18:44] that will be the package of the solr code [16:19:26] ps it's not ready for mave just yet - will take a few more hours [16:19:39] ok ok [16:19:56] or minutes [16:19:57] I just create a very basic jobs in jenkins. It is disabled for now [16:20:06] will poke it later tonight or tomorrow :-b [16:20:27] I'ts got some subversion issues. [16:22:27] hi OrenBo [16:28:29] re [18:00:42] Platonides: you have a chrome? [18:00:53] maybe we could try it using that [18:50:34] alolita: hey, can you put your Thursday deployment(s) on https://wikitech.wikimedia.org/view/Software_deployments so I can schedule around it? I need a meeting with Benny on Thursday afternoon. [18:57:18] sumanah: is the meeting about the practice audience? [18:57:44] bsitu: this is the meeting in which you are the practice audience for Tomasz's tutorial, yes. [18:58:01] sumanah: :) [19:01:28] and things that screen scrape wikipedia need shooting [19:01:28] repeatedly [19:01:28] why [19:01:28] Screen scraping is stupid and evil to start with [19:01:28] not with wikipedia, you can add a extra param to the URL and it will strip out all the shit [19:01:28] rage [19:01:48] yes they need to be shot [19:01:52] and then burned [19:02:02] and then a stake driven through their evil hearts [19:02:13] and then reanimated with a printout of API:Mainpage in their hands [19:02:16] why have an api at all [19:02:17] really [19:02:26] I'm trying to find out what "shit" said parameter strips [19:02:38] hehehe [19:02:57] better you than me... (I am on the last round of zip4 -> cl_zip5, dunno if I'll get it done) [19:03:02] Ughhh [19:03:15] To be honest, if you wanted to scrape the article text anyway, grabbing it from the api makes more sense to begin with [19:03:26] it sure does [19:03:31] who wants all that crappy html in there [19:03:32] yuck [19:04:51] People trying to be smart over the sopa banner are getting amusing - "But if you do X, it stops it!", "Yeah, and?" [19:08:16] heh [19:35:26] Reedy - how do webservers index wikipedia - scrape scarpe ... [19:40:10] make that search engines [19:58:08] Hmm, no Neil or Ian [19:58:28] sumana: will do [20:21:22] raindrift, quick question... Is the uploadstash table supposed to keep finished jobs? [20:22:05]