[00:01:02] compwhizii_: depends... do you want a central image repository like commons on your site, for use of several of your wikis? [00:01:08] or do you want to use commons itself? [00:01:16] Hi, is it possible to use pictures from Commons on an other Wiki without uploading it there? [00:01:17] or do you want to be able to use some external images from somewhere? [00:01:35] using images that are on the commons [00:01:48] ok then [00:01:53] yes [00:02:00] the anser is "no, but"... [00:02:04] !instantcommons [00:02:04] --mwbot-- InstantCommons is a proposed feature for MediaWiki to allow any MediaWiki installation to use media on the Wikimedia Commons. However, several issues (some legal) need to be worked out, and it is only a proposal. See < http://www.mediawiki.org/wiki/InstantCommons >. [00:02:19] awwwww [00:02:22] :( [00:02:23] there's a working demo afaik [00:02:29] but the legal aspects are a real problem imho [00:02:39] getting the image to show isn't such a big problem [00:02:44] dealing with image deletion is [00:03:03] what if it's a copyright violation? who is responsibel? [00:03:07] etc... [00:03:35] okay then [00:03:54] ok, Thanks Duesentrieb [00:28:48] 03siebrand * r31698 10/trunk/extensions/ (7 files in 3 dirs): [00:28:48] * fix svn props for CategoryIntersection [00:28:48] * add extension credits for Gnuplot [00:28:48] * add support for FixedImage, and Gnuplot to Translate [00:33:53] 03siebrand * r31699 10/trunk/extensions/CategoryIntersection/CategoryIntersection.i18n.php: title(s)->result(s) [00:41:53] 03(mod) Magic word to add noindex to a page's header - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=8068 (10danny_b) [00:58:11] Where to find a list of all Special:command? [00:59:13] [[Special:Specialpages]] ? [00:59:25] Ok. [01:01:59] Charitwo, nop, not that I mena, as "special:blockip". [01:02:31] im not sure I understand, that's the only list of special pages I know, you didn't specify blocks [01:02:54] but Special:Blockip is one of them, restricted to sysops [01:03:15] Yes, it was so. [01:25:10] could the next MW update have a bigger search box? :P [02:12:34] 03aaron * r31700 10/trunk/extensions/ConfirmAccount/ (ConfirmAccount_body.php SpecialConfirmAccount.php): PG fix [02:20:21] 03(NEW) special: namespace issue - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=13289 04BLOCKER; normal; Wikimedia: wikibugs; (lilewyn) [02:21:04] 03aaron * r31701 10/trunk/extensions/ConfirmAccount/ConfirmAccount_body.php: Use $wgConfirmAccountRejectAge for token expiration [02:29:20] 03(NEW) enwiki arbcom private wiki has gone missing - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=13290 04BLOCKER; normal; Wikimedia: wikibugs; (lilewyn) [02:30:30] 03(mod) enwiki arbcom private wiki has gone missing - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=13290 (10lilewyn) [02:31:01] 03(mod) special: namespace issue - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=13289 (10lilewyn) [02:47:54] 03(mod) blank edit summary message is displayed when an auto summary is put in - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=8912 (10pink-ink) [03:34:38] Is there a feature/extension of MediaWiki that automatically uses a different skin for different browsers? I've been searching for a while but I can't find anything [03:35:49] not that I'm aware of [03:36:12] !extensions [03:36:12] --mwbot-- MediaWiki has been built so it can easily be customized by adding extensions. This is usually a simple process. See for instructions to install extensions, as well as for writing them. See for an overview of known extensions. [04:24:46] 04(REOPENED) enwikibooks: let sysops grant patroller and rollbacker permissions - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=12676 +comment (10mike.lifeguard) [04:36:09] 03(mod) special: namespace issue - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=13289 (10lilewyn) [06:00:19] greetings [06:00:26] (comingo from #wikimedia-tech) [06:00:46] I was wondering... is it possible to *not* capitalize the first letter from usernames? [06:00:50] Panora: usernames are case sensitive on wiktionary... [06:00:54] http://en.wiktionary.org/wiki/Special:Contributions/Splarka [06:00:55] http://en.wiktionary.org/wiki/Special:Contributions/splarka [06:01:05] as are page names [06:01:15] Splarka: AFAIK, everything in wiktionary is case sensitive [06:01:16] Panora200: It should be dependent on the settings of the wiki in question [06:01:23] exactly... [06:01:29] not *everything*, but lots of things [06:01:32] I just wanted the usernames, if possible [06:01:56] *just* the usernames? that might be hard [06:02:38] hmm [06:02:49] I tried to create 'splarka' and it failed [06:03:07] so usernames are still first-letter case sensitive in that respect, even if they are not in other respects [06:03:32] (well, it probably would have failed for a spoofing name, but it failed in this case because the name already existed) [06:07:36] this is what I want to achieve: [06:08:12] have the edits bound to the username (case sensitive, which means having the first letter lowercase, if the user wants), but display "real name" in history, recentchanges, etc [06:08:36] if I have to hammer something, I'd like to do it the right way... like an extension, if at all possible... [06:08:50] is anyone able to point me to any valuable direction? [06:09:02] (I was reading LDAP auth, but doesn't seem to help..) [06:09:58] Hey guys. I just migrated to a new server and upgraded my wiki, all images are working fine, however the main page (index) is giving me errors: Error creating thumbnail: convert: unable to open image and this onvert: missing an image filename. Can anyone point me in the right direction please? (http://www.hapes.org/library_1_11/index.php?title=Main_Page) [06:10:17] oh boy, after so many years, gaim/pidgin still childishly crashes [06:10:45] King_James: permissions? [06:11:01] I haven't edited any CHMOD settings, which would I need to set? [06:11:28] King_James: chmod 777 /home/hapes/public_html/library_1_11/images/ -ARE (hammer-lway) [06:11:36] 777 ? [06:11:43] sorry, text substitutions... last parameter is -R . [06:11:48] 777 would solve it [06:12:02] but it's best to chown to the running apache user [06:12:13] ah, lemme log in via SSH then [06:12:26] and set it chmod ug+rwx [06:12:35] perhaps it's good to setgid as well [06:12:39] so: chmod 777 /home/hapes/public_html/library_1_11/images/ -R [06:13:26] it's easier, though :-( [06:13:28] :-) [06:13:58] ok, so -R or just the ug+rwx ? [06:14:40] ok, the CHMOD cmd is done, but error still there [06:14:46] what is the setgid? [06:14:51] King_James: what's the current ownership? [06:15:08] King_James: you should understand how permissions work... [06:15:15] man chmod [06:16:05] is there some sort of plugin that is suppose to create these thumbnails? [06:18:43] Panora200 the ownership is my me (my user) [06:18:45] not quite, I think [06:18:56] King_James: what about the user? is the same as apache is running ? [06:18:56] and the chmod didn't fix it :( [06:19:02] checking [06:19:12] chmod 777 didn't fix it? [06:19:17] nope :( [06:19:37] wait [06:19:43] read: /home/hapes/public_html/library_1_11/images/c/c6/Branch-MoCulture.JPG': No such file or directory. [06:19:57] no such file... [06:20:00] pan: possibly you could grab the 'real name' field instead of username for displaying in history/rc, etc [06:20:05] hmm [06:20:11] i wonder if i can make a page totally inaccessible [06:20:17] by naming it to something with an & in it [06:20:54] however, if you remove the /home/hapes/public_html/ and put the domain www.hapes.org/ you will see the file IS there [06:21:11] ah, it changes it to & [06:21:32] Error creating thumbnail: convert: unable to open image & convert: missing an image filename [06:21:51] Splarka: that's hacking the code! I wonder if it's possible to do it as an extension...? [06:22:15] Random: that was one reason it was blocked, pages with & in the title couldn't easily be moved [06:22:16] Random832: there was a bug like that... [06:22:37] Random832: I tried creating "Tips & Tricks", and never got to the page anymore :-) [06:22:59] Splarka: it seems now if i move a page _to_ something with & in the title, it becomes an & [06:23:05] so why not turn the & in the request into an &? [06:23:11] right, but it didn't, a few weeks ago [06:23:16] *Splarka shrugs [06:23:17] ask domas! [06:23:21] he didn't wanna do anything about it [06:23:26] bah [06:23:31] what if someone searches for &? [06:23:40] even after I proved to him that url encoded amp (percent) was hampering my ability to send examples of buggy parser diffs to Tim [06:24:01] King_James: did you change the location of the website? [06:24:08] well then someone else fix it [06:24:26] King_James: it's a fact that apache (nor anyone on the system) can find /home/hapes/public_html/library_1_11/images/c/c6/Branch-MoCulture.JPG [06:24:30] *Splarka shrugs [06:24:40] King_James: so you should be wondering why is mediawiki looking for the image in there.. [06:25:49] well, enough for tonight, need rest, thanks for now [06:28:10] 03(mod) Create the Hungarian Wikinews, Erzya & Extremaduran & Gan Wikipedia, and Japanese Wikiversity - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=13264 (10alexsh) [06:29:24] Panora200: http://www.mediawiki.org/wiki/Manual:Hooks/PageHistoryLineEnding [06:29:29] that might be able to edit the link [06:35:17] I deleted the page the hard way [06:35:17] eheh [06:54:00] why is the sitenotice output via javascript document.writeln? [06:54:14] because brion [06:54:34] ? [06:55:54] if ya wanna review the history, see http://svn.wikimedia.org/viewvc/mediawiki/trunk/extensions/DismissableSiteNotice/DismissableSiteNotice.php?view=log [06:56:09] I mean [06:56:23] we have code to dismiss our watchlist notice on en without using docment.write to create it [06:57:00] and why doesn't it do it for users who aren't logged in? [07:00:18] because brion [07:00:38] and part of the watchlist notice code on en is mine ^_^ [07:01:19] well one of them anyways (the geonotice one is funky, it uses visibility:hidden, which makes it take up space after dismissal) [07:02:16] one reason you don't want your sitenotice in plaintext, is it pollutes the top of google summaries (for a site: with no text) [07:02:49] monobook is specially designed to have the firstHeading, the contentsub, the sitesub, and first paragraph at the top of every page [07:03:05] the watchlist notice method wouldn't fool google [07:03:20] 03(mod) Display message on user registration form if database locked - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=11771 (10N/A) [07:03:24] although document.writeln is not the best solution (ick), should use dom [07:08:50] eh [07:08:55] you'd have to user innerHTML in that case [07:09:32] since in order to let people use wikimarkup in Mediawiki:Sitenotice, you have to, if you're _not_ rendering it directly onto the page, have a string that consists of the html output of rendering the sitenotice [07:10:03] that's doable with a to the api with action=parse [07:10:09] but [07:10:20] you could also simply pre-parse it in a way the dom could utilize [07:10:43] ... [07:10:45] ick. [07:10:48] whine whine [07:10:56] well then [07:11:06] the sitenotice could be rendered at the end of the page content, and cloned [07:11:11] you could render it at the bottom of the page and MOVE it with dom [07:11:13] or with CSS [07:11:18] ... [07:11:29] ...our messages on a similar topic seem to have crossed [07:11:36] great minds think two seconds apart and all that i guess ;) [07:11:43] *Splarka dismisses yours [07:11:47] [hah] [07:12:10] the problem with doing it with CSS is, of course, height. [07:30:36] hi there.. i have a question regarding namespace ids. [07:31:44] if you look at the database layout documentation (http://www.mediawiki.org/wiki/Page_table) the namespace ids are defined as ranging from 0-15 and 100-255 [07:31:58] you know what would be nice? [07:32:04] a way to get the move log for a _page_ rather than a title [07:32:41] however, if you have a look at the dump files (did not have a look at the db yet) there are namespaces with ids -1 and -2 (Special and Media) [07:33:00] babilen: yeah, but those namespaces aren't for pages that appear in the page table [07:33:37] so they can't appear on page [07:33:57] ok, so the cited documentation is correct, although that does not mean there are no namespace ids less than zero [07:34:45] they just do not belong to pages but to media files.. [07:34:51] where are they saved? [07:35:39] (media files that is) [07:40:26] 03aaron * r31702 10/trunk/extensions/ConfirmAccount/ (3 files): [07:40:26] * Consistency cleanup, just in case [07:40:26] * Add message [07:41:14] 03aaron * r31703 10/trunk/extensions/ConfirmAccount/ConfirmAccount_body.php: Opps, wrong place [07:41:45] does the user table actually contain entries for every single possible IP and range? [07:41:56] ok, not range, nevermind [07:42:11] but, how is a block on an IP represented in the database? just range start = range end? [07:49:20] Random832: http://www.mediawiki.org/wiki/Ipblocks_table [08:04:04] where in the database are media files saved? [08:05:25] nowhere [08:05:31] they are stored in a directory [08:05:45] hmm [08:05:47] just over NFS [08:06:05] is it possible for ipb_address and ipb_user to get out of sync, if _address contains a username? which one is used for actually checking if you can edit? [08:06:41] eh, don't answer that, i'm going to bed [08:31:08] hi, where can i find the mapping from the numerical namespace id to the NS_xxx used in languages/MessagesXx.php ? And why are there two different IDs for one single namespace? [08:36:33] there is only one string for namespace, other are aliases [08:37:53] I think he means NS_MEDIA ~ -2, NS_SPECIAL ~ -1, NS_TALK ~ 1, etc [08:38:21] but there isn't two [08:38:43] doesn't answer his first question ^_^ [08:39:01] found that already.. (Defines.php) [08:39:19] but they aren't all in there, so I wonder what are you doing [08:39:33] oh, where can if find the rest? [08:40:17] i am writing an API in Python because i want to use WP for NLP [08:40:37] custom namespaces can be defined almost anywhere [08:41:42] and right now i am dealing with namespaces, namespace aliases, numerical IDs, namespaces in page titles ("Image:") found in the XML dump, but not in the database... [08:42:30] and so on.. so i just like to figure out what i am dealing with. it seems like everything related to namespaces eventually maps to an unique integer [08:42:35] ugh [08:43:10] export puts namespaces with it [08:43:28] babilen: yes, it maps to a unique integer. but any installation is free to define more namepaces. they are listed in the xml dump, at the very beginning. [08:43:55] if you are looking at traw wikitext without the xml prefix, there's no way to know them. [08:43:55] but not in the database dump, because they are not stored there [08:44:20] Duesentrieb: also, export doesn't list aliases.. so it's kinda a broken [08:44:33] Nikerabbit: btw, any idea what brions change to dumpBackup wrt "uploads" is about? [08:44:38] i just saw the commit yesterday [08:44:49] Nikerabbit: huh, true, i guess [08:44:54] it really should [08:45:06] it seems to me to be the right decision to use the numerical ids as namespace and provide a bunch of dictionaries to translate between the different strings found in the titles to the numeric id... [08:45:24] babilen: which is exactly how it works internlly, yes [08:45:33] babilen: but that dictionary is per-wiki [08:45:45] are you parsing the wikitext? [08:46:23] right now i am working with the XML dump files, so it is OK if all namespaces are defined in the beginning. What gives me a bit of a headache are the namespaceAliases (we talked about that yesterday) [08:47:12] i plan to start using the database in maybe two weeks, and will find the article namespace in the page table, correct? [08:48:04] babilen: the numeric namespace, yes [08:48:22] babilen: but if you only have the database, there's no way to know what namespace names are defined for use in links [08:48:24] Nikerabbit: i am handling a subset of the defined markup with regular expressions right now. will try to write a tokenizer or something using flex/yacc later.. [08:48:45] !parser | babilen [08:48:45] --mwbot-- babilen: For alternative parsers, see . A new formalization and parser for mediawiki syntax is being discussed on wikitext-l, see [08:49:06] Nikerabbit: that stuff about aliases should *really* be fixed. [08:49:18] Duesentrieb: what do i do then (in case i just have the database?) do i have to require my users to have the xml dumps as well? [08:49:47] babilen: you have to have the namespacename -> id mapping asan additional configuration resource. [08:50:25] babilen: you should load the canionical namespaces always, and maybe also know of their translations, but users are required to explicitly define any additional namespaces. there's no way around that right now. [08:50:44] *Duesentrieb uses per-wiki property files for this in his java app. [08:50:51] the parsers listed on that webpage are not really parsers, but programs to transform mediawiki markup to html/latex... that was my impression at least [08:51:08] babilen: some so, some so. [08:51:24] babilen: the thing mediawiki proper contains is not a real parser, but only a to-html-munger [08:51:35] babilen: there's at least one yacc based version. [08:51:39] and the new antlr onw [08:52:08] Duesentrieb: so i have to write and distribute something like languages/MessagesXx.php as well, because there will be no other way to get that information from the database [08:52:12] babilen: ah, no, not yacc, but flex/bison [08:52:16] http://svn.wikimedia.org/viewvc/mediawiki/trunk/flexbisonparse/ [08:52:59] babilen: yes you can do that, but as i said, users would still be required to define any custom namespaces used in the wiki by hand [08:53:13] most wikipedias for example have a custom "Portal" namespace [08:53:24] Duesentrieb: i will deal with better markup handling until later. (1-2 months) [08:54:38] Duesentrieb: i do not care, i want to use wikipedia dumps for natural language processing. so it is fine to use the "standard" WP namespaces... and if some users want to do something else, they can edit languages/langCode.py by themselves ;-) [08:54:51] babilen: all of the mentioned parsers should have at least some namespace handling btw - they need to know all names of the image namespace in order to parse, because image links *have* to be treated differently [08:55:08] (they may contain nested links in the caption. even nested images) [08:55:43] (which is one reason why there can't be a simple grammar fopr wikitext) [08:56:32] yes, nested links are great! fell in love with them two months ago... [08:57:39] don't forget categories ^_^ [08:58:06] Splarka: i already have them. great way to automatically induce an ontology from wikipedia... [09:08:04] babilen: http://brightbyte.de/page/WikiWord [09:08:48] babilen: http://www.bibsonomy.org/user/brightbyte/wikipedia+thesaurus [09:09:15] babilen: http://www.bibsonomy.org/user/brightbyte/wikipedia+ontology [09:14:05] http://www.ukp.tu-darmstadt.de/software/WikipediaAPI [09:14:09] there'S more :) [09:30:18] <_wooz> lo [09:52:24] Duesentrieb: i know about the API's you mentioned. especially the API form darmstadt is not usable, because it is closed source and requires you to install *their* wikipedia db dumps... [09:53:03] i need functionality not available in JWPL and there is no way to fork the project... [09:53:22] (or contribute to it) [09:53:37] babilen: hm... said *said* they would realeas all code (though not under a Free license). I didn't check if they did. [09:54:06] babilen: i'm writing my own api for my thesis. it's due at the end of april, so i have to have it finished by then - and i'll release under GPL :) [09:54:29] Duesentrieb: what are you writing about? [09:54:56] babilen: extracting a multilingual thesaurus from wikipedia. read the first link (WikiWord). [09:55:41] Duesentrieb: i am currently working on a paper for this years LREC conference on multilingual named entity recognition and disambiguation... [09:56:05] Duesentrieb: what is your major? [09:56:12] informatics [09:56:27] i study computational linguistics/comp.sci [09:56:31] disambiguation will be part of the experiments i'll be runing [09:56:35] though it's not a focus point. [09:57:27] yes, the department i do it at is for computational linguistics (though i came to that field rather late in my studies, and i avoid linguistics as such) [09:57:42] i do disambiguation, semantic relatedness, retrieval(indexing). [10:26:52] 03(NEW) Allow image description pages from Commons to translate content - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=13291 15enhancement; normal; MediaWiki: Images; (brianna.laugher) [10:47:16] 03rotem * r31704 10/trunk/phase3/ (6 files in 6 dirs): Adding the fix for lists in RTL wikis to more skins, and fixing the image toc. [10:48:09] 03brion * r31705 10/trunk/wap/hawhaw/hawhaw.inc: Support iPhone OS SDK simulator app as if it were an iPhone/iPod Touch [10:48:46] tsk, why is brion on svn at 3am... [10:50:13] 03rotem * r31706 10/trunk/phase3/languages/messages/MessagesHe.php: Fixes. [10:58:44] iPhone is the shit [10:59:05] 03brion * r31707 10/trunk/wap/hawhaw/hawhaw.inc: [10:59:05] Hack for iPhone SDK simulator... its Mobile Safari doesn't seem to like application/vnd.wap.xhtml+xml, so send it text/html mime type instead. [10:59:05] This may or may not mean the iPhone OS 2.0 update will break on real iPhones too, so don't forget this hack. :) [11:37:42] 03(mod) Create the Hungarian Wikinews, Erzya & Extremaduran & Gan Wikipedia, and Japanese Wikiversity - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=13264 +comment (10joseluislikebetter) [11:43:44] 03(mod) special: namespace issue - special pages (including login) broken - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=13289 normal->high; summary (10FT2.wiki) [11:48:28] Hi, I want smaller 'edit-buttons' on my wiki, is that possible? [12:43:55] 03(mod) Pressing Enter in edit summary should not save the page - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=13065 (103tx38k) [13:15:29] 03(mod) Pressing Enter in edit summary should not save the page - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=13065 (103tx38k) [13:30:02] hi, i haveinstall the wikimedia script. stupid questions, how a user can add an article? [13:33:14] only find a button to edit the main page or the portal, now button for add a page [13:36:58] all away? :( [13:41:47] Hi, I want smaller 'edit-buttons' on my wiki, is that possible? [13:48:17] sure [13:49:38] how, dungodung? ;-) [13:49:47] custom css [13:54:43] we need a syntax for multi-line list items [13:54:46] something like this: [13:54:48] *{ Line one [13:54:50] line two [13:54:52] line three } [13:55:14] hasn't there been like hundreds of conversations about that? [13:56:43] probably [13:56:47] but nothing seems to have happened [13:57:05] true enough. I agree, the curly bracket container makes sense [13:57:14] to programmers anyway. ;-) [13:57:37] alternatively, just make consecutive lists be part of the same one [13:57:44] so you can have as many blank lines as you want [13:57:56] Can the distance between the text above a title be changed ? [13:58:06] GNU_1: for all titles, or just one? [13:58:45] one, i.e: [13:58:45] tekst [13:58:45] =title= [13:58:54] write some blank lines [13:58:59] I did [13:59:11] and it didn't work? [13:59:20] ups
's I need to add [14:00:00] Now it's better, I think:D] [14:02:48] how to delete logs of a edited page ? [14:03:02] you don't [14:03:38] But It's on my site [14:03:46] there's a maintenence script to do it deleteOldRevisions.php [14:03:56] ok [14:41:17] hey, how do i hide the edit history and previous versions of pages from unregistered users? [14:41:36] is that possible? :( [14:42:02] i don't think so no. [14:42:41] depending on your license scheme, it may even be a legel necessity to have all revisions visible [14:42:52] and anyway, it's a wiki, no? [14:42:58] not really, as i've written everything on the wiki myself [14:43:06] if you want a cms, install a cms [14:43:12] cms? [14:43:21] content management system. [14:43:43] "web authoring tool" [14:44:08] i.e. a system to procude and publish content. as opposed to a wiki, which is desigend to collaboratively edit content [14:44:34] mediawiki works perfectly well for my needs [14:44:44] if it worked you wouldn't be here asking that ;) [14:45:15] it does work, for everything except that :P [14:45:53] i may play around with alternative systems this easter or summer [14:46:01] but i don't have the time at the moment to sort out anything new [14:51:24] i don't think we have a "view history" permission. if would probably be simple enough to add that, but it would require a patch. [14:51:38] and it's quite contrary to the idea of a wiki, so i'm not sure it would go through [14:52:15] you can easily enough hide the link to the history from the skin, etc (though that also requires some hacking) [14:52:23] Maybe could be done with an extension? [14:52:24] but that does not *prevent* access [14:53:12] MinuteElectron: preventing access to the history? well, probably, using some pretty heavy tool like PageFromTitle or some such [14:53:50] the obvious way would be to have a permission for that [14:53:57] though i really donÄt understand why pople want that. [14:54:01] if( $wgRequest->getTExt('action')=='history' ) die; [14:54:17] MinuteElectron: bwahaha [15:06:00] hi is there any one can tell me how can i take my irc password back? [15:07:50] whois davidmccabe [15:08:50] alecliu: http://freenode.net/faq.shtml maybe [15:09:21] alecliu: ask in #freenode - basically, if you supplied your email address when registering your nick, you should be able to get someone to reset your passwor [15:10:10] Duesentrieb: thank you [15:10:27] Duesentrieb: i will take a shoot [15:15:34] is there any one here have been read the code of Liquid Threads Extension [15:16:20] "$wgHooks['TitleGetRestrictions'][] = 'Thread::getRestrictionsForTitle';" [15:16:31] what is it use for, the code [15:16:55] i could not find anything about the hook 'TitleGetRestrictions' [15:20:20] alecliu: i donÄt see that hook in the code. maybe something similar to UserCan? [15:22:57] i can not find the hook in the core code either. but it did show up in the extension of liquid threads [15:23:29] it is confusing me [15:54:08] 03(mod) Namespace Request for trwikisource - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=13176 (10N/A) [15:54:19] what do i need to do in order to get fulltext search capabilities in my non-public mediawiki? [15:54:24] it's running on mysql4 [15:54:30] latest mediawiki [15:54:44] apparently it's not enabled by default [15:55:00] or maybe the index takes quite some time to catch up with a rather small import [16:01:42] okay, the index table does not contain the keywords i've searched for. is there a command to refresh the index? [16:02:59] maintenance/rebuildtextindex.php [16:05:44] i see [16:05:56] this is called periodically by the wiki? [16:06:02] called never [16:06:06] it's a manual update [16:06:11] check your jobqueue [16:06:34] jobqueue is mediawiki specific or do i put this into my crontab equivalent? [16:06:36] yawn: note that the index does not contain any words shorter than four characters. [16:06:54] lol, good point. [16:06:58] yawn: it also does not index numbers. and it doesn't handle most non-english characters well [16:07:18] rebuilding the search index is unrelated to the jobqueue, afaik [16:07:23] Duesentrieb: better than nothing though [16:07:40] yawn: your search index should be updated automatically per default. [16:07:48] unless you use something fancy, like lucense search [16:08:01] i have heard people complain that it doesn't work before - no idea why. [16:08:10] Duesentrieb: nothing fancy, the complexity of the whole mediawiki alone is enough for the moment [16:08:12] *TimLaqua thinks the jobqueue is a mystical land where magical faeries update indexes and stuch [16:08:24] TimLaqua: you're right actually [16:08:46] TimLaqua: the problems starts when the faeries are on strike [16:09:32] TimLaqua: it's maily for cases when a single edit causes a lot of pages to be reparsed. [16:09:48] afaik it doesn't have anything to do with the search index as such. [16:10:52] i figured part of the import process might involve the jobqueue [16:11:03] hmm. the manual seems outdated as i get a syntax error when starting the script using cmdline php [16:11:15] ah [16:11:23] sry, different php versions installed [16:13:05] works [16:13:20] Duesentrieb: is there an easy fix for "foreign characters", e.g. umlauts? [16:13:55] yawn: check your mysql manual. mediawiki uses a myisam fulltext index. [16:14:11] you canprobably get soemthing to work with the right server config. [16:14:43] maybe you can also fidlde eith the encoding setting used in the search index table - but beware that mediawiki uses utf-8, and mysql only nominally supports that.- [16:15:32] Duesentrieb: jeez. this whole encoding mess is a ... mess. had quite an adventure with mssql & ucs2 recently. [16:15:34] bah [16:15:52] yep :( [16:15:55] Duesentrieb: i'll check on that later. right now it's 50% ok [16:16:48] is someone on channel interested in skel for editing mediawiki pages with ruby? i don't know any python so i couldn't use the bot framework and the api involved too much work. [16:18:02] *Hojjat wonders why people would like to use precious stones to edit pages?! [16:18:36] Hojjat: perl knowledge is too rusty, python is non-existant. i had to use something ... :) [16:19:03] yawn: I've never worked with perl at all! I know a little python :) [16:20:20] Hojjat: perl is very useful! for any problem x that has occured to p>=10 people there is a 99% chance there is a library that does 100% of the job [16:20:21] true fact! [16:20:22] :) [16:21:28] *yawn wonders why on earth swedish is the freakin default ordering for mysql ... [16:21:50] it was written by swedes [16:21:54]