[00:09:14] does the &$parser object in the functions hook give access to the whole current article page to add something in the end? [00:10:29] do you want to add it as if it were at the end of the text of the page, or under the content area like below the catlinks? [00:12:26] I am no php programmer, but if I wanted the former, I'd find a handy dandy http://www.mediawiki.org/wiki/Manual:Hooks#Page_Rendering hook here and simply append to it. Which hook would depend on what type of text I was inserting [00:13:05] it should be part of the article like it would be than it's in the editable article text. So above the kat links [00:13:37] right, so maybe just http://www.mediawiki.org/wiki/Manual:Hooks/ParserBeforeStrip [00:13:52] $text = $text . $yourtext; [00:15:11] but this won't enable be to create a parser function which adds some text like {{#addoutputtoarticleend: text}} [00:15:47] .= [00:15:58] ... that seems an odd thing to do, if you're going to be adding a parser function, doesn't that mean manually editing each page to have it? [00:16:11] as in, why not just put 'text' at the end of each page [00:16:46] because it's a thing only for a few pages and very dynamic [00:17:02] it will use it in some templates [00:17:21] so you couldn't just put the template calls at the end of the page? [00:17:57] no. the user should use the template where he needs it and the template should add something at the end of the page (another template call in fact) [00:18:18] and no i can't do it with css. it's not a style thing, more a content thing [00:18:29] ahh, odd [00:18:44] dunno then, since Simetrical is the know-it-all he can help ya ^_^ [00:18:49] *Splarka hides behind said sim [00:18:51] very odd. I have to write a seccond plugin which prints the text from the end to the beginning of the page later :D [00:19:19] you might have to use multiple hooks, registering a parser function and then interrupting page rendering for insertion, or something [00:20:32] yes perhap that would work. but to do it this way i have to know another thing: is there a function or something where I can make a proper template call or even more MW markup and the function returns the parsed output of that function?? [00:21:01] not of that function, i meant of that markup^^ [00:25:08] I guess I have to initialize another parser object for that or something [00:26:49] dunno, but seems inefficient, probably easier to append the raw wikicode before expansion... [00:29:54] but I can't add the code before because at this time my parser function wasn't called [01:04:59] who wants to do some bugzilla work for me, looks at werdna [01:05:23] how do i pick which way is easiest to set up multiple languages ? http://www.mediawiki.org/wiki/Manual:Wiki_family [01:06:16] there's like 6 ways to do it. should i just try the 1st one first? or [01:13:12] What namespace is Template:? [01:13:38] 10 [01:13:42] Like what number? or what? [01:13:51] Number, yes. [01:13:54] 10 [01:14:21] I am trying to search for templates that contain two characters in their name. [01:15:16] Through the API. [01:15:24] p858snake_: dude it's 3.15am [01:15:38] Unfortunately, I get no results. I am assuming MediaWiki hasn't indexed these. [01:16:44] werdna: its only about fourty or so bugs that need to be moved/re-Producted and re-componented >.> [01:16:56] be easy to do with a db query [01:17:18] OverlordQ: If I had DB access... [01:18:14] p858snake_: why would I do that? [01:18:25] so i don't have to [01:18:33] but I do other stuff :) [01:18:34] *OverlordQ doesn't think there's an easy way to search that through the API [01:18:56] that, and i would do it, but i realllly need to do some tafe stuff [01:19:10] then let somebody else do it ;) [01:19:29] svip: counting a 3-byte unicode string as one character? [01:20:55] werdna: thing is, no one will want to do this, i know that [01:21:08] how do i pick which way is easiest to set up multiple languages ? http://www.mediawiki.org/wiki/Manual:Wiki_family there are like 6! [01:21:22] p858snake_: ah well then it won't get done [01:21:32] if nobody cares enough to do it themselves [01:21:39] then it probably isn't very important, is it? [01:21:56] svip: which wiki? [01:22:16] werdna: not really [01:22:37] its just cleaning non wikibugs related bugs from the wikibugs componeant [01:26:39] werdna: i bet your maths classes wern't like this http://photos-f.ak.fbcdn.net/hphotos-ak-snc1/hs112.snc1/4821_119031680029_691045029_3334365_2379326_n.jpg >.> [01:28:45] 03(mod) Capital letters are always sorted first - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=19197 (10Drilnoth) [01:29:53] *Splarka prods svip [01:34:36] 03(NEW) SMW Comparators use sortkey instead of page name - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=19198 normal; Normal; MediaWiki extensions: Semantic MediaWiki; (christopherraylamb) [01:35:36] What's the fastest way to knock out an entire category of pages on a MediaWiki? [01:36:23] bash it in the head [01:36:26] 'knock out' has multiple contrary meanings there [01:36:37] knock out = create fast != kick out [01:37:36] :S I am confused about how action=raw works. When I do it in my browser it works as intended, but when I doing it through my Python script, it apparently returns the whole thing (HTML and all). [01:38:09] svip: svip: counting a 3-byte unicode string as one character? which wiki? [01:38:39] svip: also, are you accessing the same URL in both? because it depends how you access action=raw [01:39:24] compare: http://en.wikipedia.org/w/index.php?title=Foobar&action=raw and http://en.wikipedia.org/wiki/Foobar?action=raw [01:39:42] I only use the former. [01:40:40] which wiki are you using action=raw on? are you sending "action=raw" as POST? [01:41:05] I suppose it might be post. [01:42:46] hmm, seems to work with post though too [01:43:10]
[01:44:01] 03(mod) Lost bug report due to unknown keywords - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=12133 (10p858snake) [01:44:23] hi guys, I'm trying to write a python script that can take the title of a page in russian and retrieve it from ru.wikipedia.org -- but I'm getting an error page instead of the appropriate page back (though when I go the url manually in my browser it works...) for example if we try this url: in my browser it returns the JSON code for the britney spears page [01:44:31] Splarka: Ah, I forgot. The wiki was non-readable to unlogged in users, and I forgot to login through my script. [01:44:50] but if I request it using urllib in python I get an error page [01:44:52] 'knock out' has multiple contrary meanings there [01:44:55] what url? [01:44:58] *Splarka bites svip [01:45:06] I basically mean how do I wipe all articles in a category [01:45:08] http://ru.wikipedia.org/w/api.php?action=query&prop=revisions&titles=Спирс, Бритни&rvprop=content&format=json [01:45:34] 03(mod) Buttons merge into a box - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=12879 (10p858snake) [01:45:42] mib_z2cp92: try setting a user agent that isn't the default python, also try urlencoding the titles= parameter [01:46:02] Dr_Link: have to edit them all, there might be a pywikipediabot script for that [01:46:28] I'll try both of those thanks [01:46:53] 03(mod) Display of lower case rho as ϱ rather than ρ on Safari Browser - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=11663 (10p858snake) [01:46:55] Sparkla: Alright, thanks. [01:47:40] mib_z2cp92: I know wikimedia wikis block certain non-browser user agents by default, they like contact information and statement of purpose, dunno if python is one of them [01:47:58] 03(mod) invalid HTML: invalid width attribute value - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=11738 (10p858snake) [01:47:59] yes, you need to uniquely identify your script with the user-agent [01:48:10] Splarka but isn't the api made for non-browser agents? [01:48:24] the api is not just for wikimedia [01:48:36] mib_z2cp92: it doesn't matter, we want you to change the user-agent [01:48:37] what's the error page? [01:48:56] if you are sending a request from python and you haven't changed the user-agent, your requests will fail [01:49:05] wedna -- got it thanks [01:49:11] note that it will work for many non-wikimedia mediawiki installs just fine [01:49:20] it's blocked in squid [01:49:22] I set my 'Referer' but I'm guessing that different [01:49:29] no, User-Agent [01:49:43] 03(mod) Search possibilities cover up search buttons - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=16327 (10p858snake) [01:50:37] 03(mod) Moving a page back does not work as described - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=17228 (10p858snake) [01:51:42] 03(mod) sandbox display differs from live display, page contains racial comments - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=10368 (10p858snake) [01:53:36] 03(mod) Wrong URL to old image version - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=12981 (10p858snake) [01:53:54] one more and i'm done for awhile [01:54:43] 03(mod) On "log in" and "log out", a link to Special:Userlogin is displayed instead of a link to the page listed in the returnto= parameter - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=11339 (10p858snake) [01:59:27] Wikia runs MediaWiki trunk right? [01:59:51] no, 1.14.0. highly customized [02:07:55] any progress on [02:07:58] !16553 [02:07:58] --mwbot-- I don't know anything about "16553". [02:08:01] !b 16553 [02:08:01] --mwbot-- https://bugzilla.wikimedia.org/show_bug.cgi?id=16553 [02:09:55] does anyone know of a good technique to just strip off the wiki tags/mark up and the HTML and just get the text back? [02:10:01] (using python) [02:10:19] depends on your requirements [02:10:35] do you want the infobox stripped, do you want the contents of it? [02:10:38] hmm nevermind forget that I just realized I'm still getting the same problem I was having earlier [02:10:42] here is a good example: http://en.wikipedia.org/wiki/Foobar [02:10:54] [02:10:59] [02:11:47] but what I find most useful, is to visualize the content of a page, like via action=render, as a set of top level block level elements: view-source:http://en.wikipedia.org/wiki/Foobar?action=render [02:12:25] for example, this is structured as: div, div, p, p, table, script, p, h2, p, p, p, p, p.... [02:12:54] you'll find most of the useful content inside the top level

blocks (as that is the default block container for wikitext when parsed if it has no other) [02:13:21] ol, ul, table, div, are usually templates/markup/dabs/lists, etc [02:13:44] blockquote, pre are often content but often not very portable [02:13:46] is it ok to scrape wiki content without the api? [02:13:57] you can get this with the API, this is just a visual example [02:14:47] http://en.wikipedia.org/w/api.php?action=parse&page=foobar [02:16:14] so for kicks, try parsing the top-level dom of the pages, selectively grabbing

-

, and maybe
 and 
and dumping all the rest and their contents [02:16:37] well actually [02:16:48] http://en.wikipedia.org/w/api.php?action=parse&page=foobar&format=jsonfm [02:16:58] everything I need is under parse / text / '*' [02:17:20] does anyone know of a good technique to just strip off the wiki tags/mark up and the HTML and just get the text back? [02:17:32] what is in text there is exactly the same as what is in http://en.wikipedia.org/w/api.php?action=parse&page=foobar [02:17:40] they are just escaped in different ways [02:18:07] in format=xmlfm (default), the html entities are escaped, " & < >, etc [02:18:19] in format=jsonfm, the quotes are \" \' escaped [02:18:32] and newlines converted to \n [02:19:24] I guess now I just have to strip off the html -- I imagine there are libraries in python that will do that [02:20:33] hmm, shouldn't this find any single character (from 1 to 4 bytes) that is valid unicode? ([\x00-\x7f]|[\xc2-\xdf][\x80-\xbf]|[\xe0-\xef][\x80-\xbf]{2}|[\xf0-\xf4][\x80-\xbf]{3}) [02:40:16] is there any problem with < and > in preg_replace or with mediawiki? [02:40:45] I try to use preg_replace in ParserBeforeTidy hook but as soon as I use < or > inside it it won't work... [02:59:17] 03(mod) The search function on Wikipedia won't select - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=19194 (10Utahraptorostrommaysorum) [03:05:42] 03mrzman * r51858 10/trunk/phase3/ (RELEASE-NOTES maintenance/purgeOldText.inc): (bug 19160) Use DISTINCT instead of DISTINCTROW for compatibility with postgresql [03:11:20] 03(FIXED) DISTINCTROW with Postgres - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=19160 +comment (10mrzmanwiki) [03:12:56] When I view a diff of a null change, the diff table layout is wierd in Google Chrome. [03:13:50] Null edits don't add revisions, so it's fairly difficult to diff them. [03:13:57] Are you talking about things like page protections? [03:14:32] No, the diff is made across a revert [03:14:40] So there was no change to the page [03:15:00] What kind of weirdness are you having and can you take a screenshot? [03:15:20] Yeah; right column gets scrunched up, I'll get a screenshot [03:19:21] 03dale * r51859 10/branches/new-upload/phase3/js2/mwEmbed/ (10 files in 7 dirs): [03:19:21] * stubs for transition editor [03:19:21] * transition menu selector [03:20:00] 03(NEW) Granularize RevisionDelete (RevDelete) user rights - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=19199 enhancement; Normal; MediaWiki: Revision deletion; (public) [03:20:15] http://infinisoft.ath.cx/mwbug.jpg [03:20:48] Here is the screenshot [03:21:37] infinisoft: IE 7 does that alot more often [03:22:12] but I've never seen it happen on any other browser [03:22:15] Ahh. [03:22:27] I believe that issue has been resolved in SVN. [03:22:44] Safari 3 and 4 had the same problem, as I recall. [03:23:00] I know the IE7 one is a MSFT glitch though [03:23:07] It should be fixed when (if) Wikimedia ever updates their copy of MediaWiki. [03:23:18] Ok cool. [03:23:35] b/c when I hit refresh it magically goes away... 50% of the time [03:24:50] I guess webkit is getting confused by the colspan in the diff table [03:25:18] Yeah. [03:25:21] The fix is... [03:26:15] g'night all [03:26:39] http://www.mediawiki.org/wiki/Special:Code/MediaWiki/49688 [03:28:19] Sweet [03:28:34] It was bothering me too. ;-) [03:32:22] Hmm, did rollbacks always require a token to be given? [03:32:32] Yes. [03:32:38] ever since we discovered CSRF [03:32:58] Morning Tim. [03:33:15] hi [03:34:00] Was rollback the only thing modified to avoid cross-site request forgery? [03:34:43] heh [03:34:55] no [03:39:58] Ok; thanks. [03:44:39] I was looking for maybe an old user javascript to add rollback links to recent changes, it would have to use some ajax then. [03:45:19] Well, there are scripts that emulate JavaScript that are very old. [03:45:30] "emulate javascript"? [03:45:34] Err. [03:45:38] Emulate rollback. * [03:45:51] "godmode" or "godmode lite" are the names, I think. [03:46:03] I think I saw that one, it relies on a parameter given to the recent history and then does a link by link revert or something. [03:46:18] http://sam.zoy.org/wikipedia/godmode-light.js maybe? [03:48:20] Yeah. Wouldn't it be less work to run an ajax to fetch the token using Query API and add a link to the RC page ^_- [03:48:31] Prolly. [03:48:48] Splarka is the one to ask (he's not on right now). [03:49:59] I figured the ajax could wait for you to mouseover the rollback link before actually getting a token since it looks like you need to do a request for every recent change [03:53:15] Hehe, I think I will have a go at this one then :D [03:53:38] I'd like to use Cavendish skin with Mediawiki. DL, installed & update of the skin from CVS repo is easy. Of course, I want to 're-brand' the skin to my brand. No problem, just customize ... but, [03:53:47] Once I've customized the skin, is there a "clever way" to continue to sync/update with the latest-n-greatest from the Cavnedish repo, and NOT lose/overwrite my edits -- and NOT 'confuse' Mediawiki? [03:53:54] Is a simple, occassional 'cvs up' in my _customized_ Cavendish dir sufficient? [04:14:40] [03:48:20] Yeah. Wouldn't it be less work to run an ajax to fetch the token using Query API and add a link to the RC page ^_- [04:14:40] [03:48:48] Splarka is the one to ask (he's not on right now). [04:15:08] it'd be less work to just ask for 'rollbacker' rights (and/or convince the shellers on said wiki to create such a group) [04:15:38] emulating rollback in javascript accurately is non-trivial, as it has to check the history and find the first revision by another user [04:27:23] Splarka: It's on a Wikia; I'm not familiar with Wikia much though. [04:30:44] which wikia wiki? [04:30:57] Splarka: That godmode script thing already seems to perform emulation; I basically had a request from an admin user of a wikia to add RC rollback for convenience purposes I suppose [04:31:16] oh, for someone already with rollback? [04:31:22] ahh, right, they are a few versions behind [04:31:29] Yeah. [04:31:31] trunk has rollback on irc [04:31:37] er on rc [04:31:38] heh [04:31:40] Ah. [04:31:52] Well that's nice, how often does Wikia update? :p [04:32:18] *Splarka can't remember if 1.14 did api rollback [04:33:07] The query for getting the token seems to work [04:33:12] yah [04:33:17] then you could just submit that to the UI even [04:33:29] ahh, it does have action=rollback too [04:33:38] but the UI is more friendly, so yah, easy to do [04:33:48] though I'd do it on click, not on mouseover [04:34:19] [rollback] [04:34:28] drop that in for each
  • in RC [04:34:43] Hmm. Sweet [04:34:53] and then on doRollback, fetch the token via ajax [04:35:29] and if it returns successful, document.location to the title=Page_Name_Here&action=rollback&from=USERNAME&token=~~~3%2B%5C [04:35:36] so one-click with short delay [04:36:32] you may be tempted to do but is much more open intent, and user friendly for debugging, and you don't need a non-js fallback anyway for js-generated content [04:36:42] That should work perfectly then :D [04:36:49] *Splarka nods [04:37:13] *infinisoft starts scripting :p [04:37:36] if(wgCanonicalSpecialPageName && wgCanonicalSpecialPageName == 'Recentchanges' && wgUserGroups && wgUserGroups.join(' ').indexOf('sysop') != -1) addOnloadHook(initRcRollback) [04:38:25] Ok :p [04:38:55] oh, and there may be multile UL [04:38:59] http://test.wikipedia.org/wiki/Special:RecentChanges [04:39:00] spanning days, for example [04:39:28] so.. var ul = getElementsByClassName('ul) [04:39:31] ack [04:39:44] so.. var ul = getElementsByClassName(document,'ul','special'); [04:40:03] iterate over those, iterate over ul[i].getElementsByTagName('li') [04:40:25] and ew, iterate over the .getElementsByTagName('a') for those, finding one you like dealing with, probably the diff [04:40:38] if the (diff) isn't linked it isn't rollbackable, heh [04:40:40] (or a log entry) [04:42:06] grab the title and the user's name from the later , plop those into a new document.createElement('a') where the setAttribute('href','javascript:doRollback("' + encodeURIComponent(titlethingy) + '","' + encodeURIComponent(usernamethingy) + '")') [04:42:18] at least, that's how I'd do it [04:42:21] *Splarka afk [04:42:27] Thanks [04:44:28] Yeah, some crazy DOM walking/filtering will have to go on, should be doable thoguh [04:47:07] btw: http://svn.wikimedia.org/viewvc/mediawiki?view=rev&revision=45918 [04:47:44] plopping that revision into http://svn.wikimedia.org/viewvc/mediawiki/branches/?pathrev=45918 shows it was before 1.15 was branched, but after 1.14, so it is a 1.15 feature [04:49:32] You'll see that at http://wiki.conky.be/wiki/Category:Reference there are 3 pages categorized under the letter 'R' and not one of them starts with 'R', can someone explain why this happens ? [04:51:54] [[Category:Reference|Reference]] [04:51:56] at http://wiki.conky.be/conkywiki/index.php?title=Command_Line_Options&action=edit [04:52:02] the parameter in a category tag is the sortkey [04:52:13] that sorts it alphabetically as if it were 'Reference' [04:52:30] use [[Category:Reference]] if you want it correctly sorted [04:53:10] 03tstarling * r51860 10/trunk/phase3/skins/CologneBlue.php: Fix typo from r51820 [04:54:27] Splarka: Thanks ! [04:56:39] np..... [04:59:33] I like how Chrome lets you abort JavaScripts with alert boxes [05:00:17] eh? [05:06:25] If you have a script that throws more than one alert dialogue box then chrome lets you stop the script [05:06:57] don't all browsers do that? [05:08:23] The one I Was using before didn't :\ [05:08:58] flyingparchment: depends if they are at the same time or one after another iirc [05:11:48] Splarka: Awesome, I have it adding the doRollback links [05:52:29] Ug. I need to upgrade my test wiki. [06:27:52] any crats around ? I working my way through a global rename. [06:28:46] on mw.org, sure! [06:31:43] ialex: http://meta.wikimedia.org/w/index.php?title=User_talk:John_Vandenberg&diff=1517807&oldid=1517798 [06:32:27] Eek. [06:33:10] jayvdb: done [06:35:01] ialex: ty [06:35:56] Rowan: can you please stop /nick all the time? [06:36:15] ialex: Sorry. I'm actually done now, was regrouping nicks. [06:36:20] Because the old alts were dropped. [06:36:23] 03tstarling * r51861 10/trunk/extensions/ (4 files in 4 dirs): [06:36:23] Fixed some syntax errors: [06:36:23] * __call can't be static [06:36:23] * missing comma [06:36:23] * No such thing as "abstract static" [06:36:25] * __set can't be protected [06:37:30] *Rowan should have prolly opened another connection. [06:46:44] 03ialex * r51862 10/trunk/phase3/skins/Standard.php: we have SpecialPage::getTitleFor() for that [06:54:53] 03tstarling * r51863 10/trunk/phase3/includes/Export.php: Remove $resultset->free() calls, as in live hack. The calls are unnecessary, removing probably fixes some bug. [07:00:20] Yay, I think my RC Rollback script works :D [07:42:56] hey guys, i just did a new first time install and added a user while installing, how can i add it to the sysop or the bureaucrat group? [07:43:32] if it's the one you specified in the installer it already has crat and sysop [07:43:40] if it's a new user check [[Special:UserRights]] [07:49:22] 03siebrand * r51865 10/trunk/extensions/ (42 files in 34 dirs): Localisation updates for extension messages from translatewiki.net (2009-06-15 07:42 UTC) [07:51:27] 03siebrand * r51866 10/trunk/phase3/languages/messages/ (45 files): Localisation updates for core messages from translatewiki.net (2009-06-15 07:42 UTC) [07:53:46] OverlordQ, how come i cannot see 'Create a new page' then? [07:57:24] 03dale * r51867 10/branches/new-upload/phase3/js2/mwEmbed/ (4 files in 2 dirs): * more stubs for transition editor [07:58:11] not a clue [07:58:44] if you pull up a nonexistant page, what does it say? [08:04:08] OverlordQ, says "This page has no text" and theres an edit option [08:04:12] i should just add them like this? [08:04:15] still work? [08:06:05] 03siebrand * r51868 10/trunk/phase3/skins/common/shared.css: Label width from 20% to 30% to further reduce label wrapping. Is was for example wrapping on Special:AllPages for "Display pages starting at:" at 'at' in FF3. [08:13:36] 03thomasv * r51869 10/trunk/extensions/ProofreadPage/ProofreadPage.php: follow-up to r51788 [08:13:44] 03(mod) SMW Comparators use sortkey instead of page name - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=19198 (10markus) [08:29:36] if your user account didn't get created properly during the install you can [08:29:39] !createandpromote [08:29:39] --mwbot-- To recreate the admin user, run "php maintenance/createAndPromote.php" on the command line. [09:15:28] 03(NEW) Scap to r51863 broke edit summary label - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=19200 normal; Normal; MediaWiki: Page editing; (sco_scam) [09:17:27] 03(mod) Edit summary label broken - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=19200 summary (10siebrand) [09:20:27] 03(NEW) "Multi-Category-Search" in Mediawiki - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=19201 enhancement; Normal; MediaWiki: Categories; (daniel) [09:24:11] 03(mod) Edit summary label broken - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=19200 +comment (10siebrand) [09:24:45] 03(mod) Edit summary label broken - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=19200 +comment (10agarrett) [09:26:30] 14(DUP) "Multi-Category-Search" in Mediawiki - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=19201 +comment (10rainman) [09:26:34] 03(mod) Allow searching in intersections, etc. of categories - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=5244 +comment (10rainman) [09:32:10] 03(mod) Search in yi: should ignore diacritics and identify ligatures - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=18764 (10wiki.pedia) [09:32:52] is there a syntax highlighter that is self-contained, i.e. not simply a pipe to a binary highlighter? [09:33:39] SyntaxHighlight_GeSHi [09:33:43] (which uses GeSHi) [09:36:38] 03(mod) Edit summary label broken - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=19200 +comment (10lastword) [09:36:44] gah, enscript's configure is for unix only [09:37:19] rook2pawn: see what I said above :) [09:41:12] 03jojo * r51870 10/trunk/extensions/Collection/ (4 files): use hardcoded rights collectionsaveas{user|community}page. [09:43:39] 03(FIXED) Implement user permissions for saving books to the wiki - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=18902 +comment (10support) [09:45:08] 03rainman * r51871 10/trunk/phase3/includes/specials/SpecialSearch.php: Fix for r51574: did you mean.. links were broken, the original change seem to miss this one.. [09:45:38] ialex: i just read a geshi vs enscript "Better highlighting. Flat-out kicks enscript's ass." i guess im going with geshi :) [09:47:35] 03jojo * r51872 10/trunk/extensions/Collection/ (Collection.body.php Collection.php README.txt): rename wgLicense{Name|URL} to wgCollectionLicnse{Name|URL} [09:48:39] 03jojo * r51873 10/trunk/extensions/Collection/Collection.templates.php: fix indentation [09:50:04] 03jojo * r51874 10/trunk/extensions/Collection/Collection.hooks.php: replace empty() with isset() [10:00:35] 03jojo * r51875 10/trunk/extensions/Collection/Collection.hooks.php: fix indentation [10:03:33] 03jojo * r51876 10/trunk/extensions/Collection/Collection.hooks.php: fix for unset var $oldid [10:23:24] 03(NEW) Disclaimer link has broken URL - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=19202 major; Normal; MediaWiki: User interface; (daniel) [10:27:33] hm, seems like a whole bunch of UI stuff broke [10:27:55] Duesentrieb: MediaWiki:Summary went from raw html to wikicode only [10:27:59] so a lot of abuse of it has now broke [10:28:04] http://bg.wikibooks.org/wiki/MediaWiki:Summary?action=edit [10:28:06] like that, haha [10:28:21] yes, i just noticed that [10:28:28] *Splarka is making a list [10:28:33] but the edit toolbar is also broken for me. at least on dewp [10:28:36] know any stewies with free time? [10:29:14] seems to be local. [10:29:23] Dues: hmm, wfm [10:29:40] but the edit summary there is foob'd [10:30:19] yes [10:32:37] 03jojo * r51877 10/trunk/extensions/Collection/ (Collection.hooks.php collection/popup.js): fix escaping/quoting. shorten code by using Xml::encodeJsVar() and Skin::makeVariablesScript() [10:34:21] edit buttons gone :/ [10:35:01] 03jojo * r51878 10/trunk/extensions/Collection/Collection.hooks.php: optimize page-in-category check [10:36:49] dues: sure it isn't a preferences prob? [10:36:52] any js errors? [10:39:09] How would i get http://meta.wikimedia.org/wiki/Template:Index-table ? [10:39:55] It seems to use class="prettytable" .. so would I also create Template:prettytable ? [10:44:19] a class is not a tempate [10:46:07] 03jojo * r51879 10/trunk/extensions/Collection/ (Collection.body.php Collection.hooks.php): check result from Title::newFromText() [10:47:50] is there some kind of wiki page substitution when specifying 03(mod) Disclaimer link has broken URL - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=19202 +comment (10overlordq) [11:11:31] 03siebrand * r51880 10/trunk/phase3/includes/OutputPage.php: Follow-up to r50822. Add id="mw-returnto" to allow styling. [11:16:40] 03(WONTFIX) Edit summary label broken - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=19200 +comment (10siebrand) [11:24:14] 03(NEW) AbuseFilter does not fail gracefully with bad regex's - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=19203 major; Normal; MediaWiki extensions: AbuseFilter; (overlordq) [11:28:01] 03siebrand * r51882 10/trunk/phase3/includes/Skin.php: (bug 19202) Follow-up to r51819: fix incorrect conversion. Put wfMsgForContent back. [11:28:08] 03(mod) Disclaimer link has broken URL - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=19202 +comment (10siebrand) [11:28:25] 03(FIXED) Disclaimer link has broken URL - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=19202 (10siebrand) [11:36:36] 03jojo * r51883 10/trunk/extensions/Collection/ (Collection.body.php Collection.i18n.php): addArticle() is a static method. check return code. [11:40:25] 03jojo * r51884 10/trunk/extensions/Collection/ (Collection.body.php Collection.i18n.php): use return codes in removeArticle() etc. [11:41:11] 03jojo * r51885 10/trunk/extensions/Collection/Collection.body.php: remove calls to User::invalidateCache() (deprecated by the cache invalidation hook) [11:44:21] 03jojo * r51886 10/trunk/extensions/Collection/Collection.body.php: do not use hardcoded value for pediapress. do not pretend to return something. [11:46:29] 03(NEW) Special:GlobalUserRights's borked - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=19204 normal; Normal; MediaWiki extensions: GroupPermissionsManager; (koneko) [12:04:43] 03(NEW) Special:Undelete gives wrong links on pages with "&" in the title - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=19205 normal; Normal; MediaWiki: Deleting; (church.of.emacs.ml) [12:05:33] 03jojo * r51887 10/trunk/extensions/Collection/Collection.body.php: use static member as cmp function. use array_merge() when sorting items. [12:10:28] 03jojo * r51888 10/trunk/extensions/Collection/Collection.body.php: check that $index is an integer. do not issue notices on empty lines when parsing collection pages. [12:13:06] 03siebrand * r51889 10/trunk/phase3/includes/specials/SpecialUndelete.php: (bug 19205 ) Follow-up to 50351 and 51572: no double escaping. [12:13:31] 03(FIXED) Special:Undelete gives wrong links on pages with "&" in the title - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=19205 +comment (10siebrand) [12:21:32] is there a prefered markup for plugins which create some markup for themselfes in an early parser step which is replaced and interpreted later in the parser process? [12:22:08] for example xml [12:23:09] 03jojo * r51890 10/trunk/extensions/Collection/Collection.hooks.php: typo fix [12:23:26] 03(NEW) Articles created by autoeditors are not marked as autoreviewed - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=19206 critical; Normal; MediaWiki extensions: FlaggedRevs; (claymore.ws) [12:23:36] 03jojo * r51892 10/trunk/extensions/Collection/Collection.body.php: factor out line parsing function [12:23:37] 03shinjiman * r51891 10/trunk/extensions/ (2 files in 2 dirs): Localisation updates Cantonese and Chinese for extension messages [12:28:15] 03jojo * r51893 10/trunk/extensions/Collection/Collection.body.php: use CURLOPT_TRANSFER instead of ob_*() [12:29:08] 03(NEW) Bot edits not be marked automatically as sighted - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=19207 critical; Normal; Wikimedia: wikibugs; (danny.leinad) [12:31:07] siebrand: Dunno if you knew this already, but it might be helpful to disable translation of PrefStats messages for a while as the extension is still under development [12:32:03] RoanKattouw: I read your request. For now there are only 2 messages which are fine. [12:32:22] Right; in the future, however, unstable messages might be added [12:33:13] 03siebrand * r51894 10/trunk/phase3/includes/ (16 files in 2 dirs): Fix double escaping issues. Related to r51559 and friends. [12:33:15] mmm, whyfor: http://translatewiki.net/wiki/Special:Translations/Returnto doesn't work, but it pre-fills in the "Name:" box, which when simply submitted /does/ work? [12:34:47] Splarka: : it isn't selected [12:35:11] Splarka: something in there should probably have a default to make that work that doesn't. [12:35:20] Splarka: feel free to create a bug report in bugzillla. [12:36:18] siebrand: http://translatewiki.net/w/i.php?title=Special%3ATranslations/Returnto&namespace=8 also doesn't work, but correctly parses the parameter out [12:36:23] meh, am lazy [12:38:32] 03jojo * r51895 10/trunk/extensions/Collection/ (Collection.body.php Collection.templates.php): added token to save collection form to prevent CSRF [12:39:16] Splarka: of course that does not work. [12:39:29] Splarka: "/Returnto" is not part of the page name [12:40:06] wat? [12:40:25] title=Special%3ATranslations/Returnto&namespace=8 [12:41:26] Hello. I couldn't find a clear answer on this question: is one able to retrieve the pages that a user has viewed? [12:41:40] Or is it possible to write an extension for this [12:43:00] siebrand: but it correctly parse out the slashie parameter and sticks it into the Name box exactly as it needs to appear there to work, so I don't understand why it fails X_X [12:43:36] I add some < and > in my function hooks together with some xml in the page text. But how can I tell the parser that this is xml and the should not convert < to < ??? [12:44:03] Splarka: me neither. And I really don't care at the moment :) [12:44:12] siebrand: fair enough, <3 [12:44:45] Danwe: IIRC you register it as a tag extension? maybe: http://www.mediawiki.org/wiki/Manual:Tag_extensions [12:45:09] then it gets replaced with a strip marker during parsing and put back after, maybe [12:46:58] Splarka: I don't want to use the tags on my article pages, I'd like to add the xml as return value of a parser function [12:47:39] Danwe: Then you should indicate that what you're returning is HTML, not wikitex [12:47:42] Lemme look up how [12:48:12] http://www.mediawiki.org/wiki/Manual:Parser_functions#Parser_interface [12:48:33] great, thanks! [12:48:47] 03nikerabbit * r51896 10/trunk/extensions/Translate/SpecialTranslations.php: Possible fix for Special%3ATranslations/Returnto&namespace=8 requested by Splarka [12:50:28] 03(mod) API must be accessed through the primary script entry point error - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=13049 +comment (10alexsh) [12:51:17] Splarka: gosh, Nikerabbit fixed it for you :) http://translatewiki.net/w/i.php?title=Special:Translations/Returnto&namespace=8 [12:51:57] Splarka: http://translatewiki.net/w/i.php?title=Special:Translations/MediaWiki:Returnto will also work, btw. [12:52:15] *Splarka tries http://translatewiki.net/wiki/Special:Translate/mainpage [12:52:22] er [12:52:31] *Splarka tries http://translatewiki.net/wiki/Special:Translations/mainpage even [12:52:36] hmm [12:53:04] Any hint why the {{{1}}} does not work?
    {{{2}}}
    [12:53:22] 03(mod) Page shown as draft after being sighted - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=18419 (10gtisza) [12:53:48] ahh, siebrand: gotcha, needs the namespace [12:53:55] Nikerabbit: cool beans <3 [12:54:01] 03(mod) API must be accessed through the primary script entry point error - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=13049 +comment (10russblau) [12:54:58] Splarka: beans? [12:55:24] "good job" [12:57:15] I would still appreciate an answer :P [12:57:33] dergringo: https://bugzilla.wikimedia.org/show_bug.cgi?id=12974 possibly, is # the first part of the parameter? [12:58:00] Splarka, yes. I also tried rgb() and green. all do not work. [12:59:15] Joozey: I'd use a third party server, and some JS, personally [12:59:20] like google analytics [12:59:44] Hm okay, but it's not native to mediawiki [12:59:57] not really, no, not if you wanna associate usernames with page views [13:00:06] ok, thanks [13:00:09] something like document.createElement('img) [13:00:47] er, document.createElement('img'), and assign it a .src with some JS parameters ... wgPageName wgUserName and the current JS time [13:01:02] and have it load a remote script that logs the parametrs and returns a 1x1 transparent image [13:01:48] 03jojo * r51897 10/trunk/extensions/Collection/ (3 files): MW coding conventions [13:05:23] And how is flash support? Is it already possible to communicate with the wiki through as3? [13:05:35] should I use American English or or British English for parser function names o.O ?? [13:05:41] (perhaps it's like "duh, of course", but you never know :P) [13:06:44] 03nikerabbit * r51898 10/trunk/extensions/Translate/groups/ComplexMessages.php: Quick fix for duplicated magic words export [13:09:04] dergringo: WFM [13:09:16] Splarka, ? [13:09:49] http://test.wikipedia.org/w/index.php?title=Template_talk:Cat [13:09:49] What does that mean? [13:09:55] "works for me" [13:10:19]
    foo
    [13:10:26] hmmm [13:15:25] 03siebrand * r51899 10/trunk/phase3/languages/messages/ (14 files): [13:15:25] Update magic words: [13:15:25] * remove duplicates [13:15:25] * note consistently [13:17:05] 03(CLOSED) javascript error while playing clip in IE6 - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=19038 (10eran.kuziel) [13:20:53] 03nikerabbit * r51900 10/trunk/phase3/includes/ImagePage.php: Fix regression probably from r50975 - link was not parsed [13:21:21] 03(NEW) Disable 'missingsummary' feature on dewiki - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=19208 enhancement; Normal; Wikimedia: Site requests; (church.of.emacs.ml) [13:21:32] 04(REOPENED) clip does not play with native engine and with Java engine - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=19124 +comment (10eran.kuziel) [13:21:45] I'm getting an "unable to fork" error when trying to create thumbnails [13:21:48] any suggestions? [13:22:04] i.e. in the apache log- exec() [
    function.exec]: Unable to fork ['/home/cojs/publi [13:22:04] c_html/cojswiki/bin/ulimit4.sh' 30 102400 102400 ''\\''/usr/bin/convert'\\'' [13:25:14] 03(NEW) Capitalisation of tabs - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=19209 enhancement; Normal; Wikimedia: General/Unknown; (servien) [13:29:21] 03tstarling * r51901 10/trunk/phase3/languages/messages/MessagesMr.php: Removed empty-string magic word synonyms, were probably breaking the wiki [13:29:59] 03(CLOSED) Audio and video are not in synch - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=18929 (10eran.kuziel) [13:31:29] 03nikerabbit * r51902 10/trunk/extensions/Translate/groups/ComplexMessages.php: Check for empty (=== '') elements too [13:31:43] i'd like to add a "onclick" attribute to every text-link in my mediawiki... any suggestions how to do this ? [13:34:50] 04(REOPENED) using Java engine - scrubber does not move in progress bar when clip is playing - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=19123 (10eran.kuziel) [13:36:22] 03nikerabbit * r51903 10/trunk/phase3/languages/messages/ (MessagesHe.php MessagesKo.php MessagesMl.php MessagesQu.php): Removed empty elements from magic words [13:37:24] HeimdallAustria: lazy way is to add site js [13:37:50] Splarka: outstanding issues? [13:37:57] like this to MediaWiki:Common.js http://p.defau.lt/?bRh8l97ykSS0buRpzpCyEw [13:38:25] Nikerabbit: http://translatewiki.net/w/i.php?title=MediaWiki:Returnto/km&action=history maybe [13:39:04] 04(REOPENED) in FF 3.5b4 timer shows full time of original clip - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=19030 (10eran.kuziel) [13:39:16] other than that, all I can suggest are to have returnto behave normally, and temporarily revert Summary to allow raw html until the js and crap can be migrated properly [13:39:25] Splarka: http://translatewiki.net/w/i.php?title=Special:Translations&message=MediaWiki:Returnto looks good [13:39:47] 04(REOPENED) "waiting for video" label doesn't appear - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=19023 normal->low (10eran.kuziel) [13:39:48] Nikerabbit: yes, km was the only one in core that wasn't plain text, sie just fixed it [13:40:07] Splarka: well, it will be broken either way, we cannot allow both at the same time [13:40:37] k, so if summarytext is to remain broken, convince tim or a steward to fix them ^_^ [13:40:59] I'm not saying that [13:41:06] er, lemme rephrase [13:41:24] if summarytext is to stay as regular wikicode and not allow raw html: (I've spammed the list a few times, sorry) http://p.defau.lt/?dXjZXN0z9GCYl_55L85D2Q [13:41:43] if returnto is to remain as plain text: http://test.wikipedia.org/wiki/MediaWiki:Returnto http://en.wikipedia.org/wiki/MediaWiki:Returnto http://es.wikiquote.org/wiki/MediaWiki:Returnto http://mi.wikipedia.org/wiki/MediaWiki:Returnto http://ms.wikipedia.org/wiki/MediaWiki:Returnto http://mt.wikipedia.org/wiki/MediaWiki:Returnto [13:41:54] some of these were already fixed, but I am so sleepy... [13:42:01] There's another one [13:42:11] the cancel link is plain text instead of a link :P [13:42:15] heh [13:42:16] 03(CLOSED) seek does not work - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=18895 (10eran.kuziel) [13:42:17] we missed the announce changes part [13:42:32] Heh [13:42:34] OverlordQ: in defaults or when customised? [13:42:36] I am sure more will turn up over the next day or so [13:42:44] 03(CLOSED) Download function doesn't work - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=19029 (10eran.kuziel) [13:43:02] Nikerabbit: Defaults [13:43:03] Nikerabbit: anyway, you might wanna commit the change to /km (per the actual question rather than per my ranting) [13:43:20] so, I'm trying to heave the executive decision (either to revert it for all, revert it and do it again with announce, ???) to someone else [13:43:44] fair enough [13:43:51] it is a nasty quagmire [13:44:15] I just wanted it to not look horrible, most easily by deleting all messages that have the outstanding issue I see is the target=_blank thing on editsummary [13:44:51] those who have