[00:41:51] what's about the max upload size before you should have your bot upload in chunks using the upload stash? Also, what's about the max size the chunks should be? thanks. [01:46:16] Hi leucosticte. [01:46:28] Carmela: s'up [01:46:35] I've been following your mediawiki.org edits. [01:46:54] The Inclumedia shit really does need to be moved to a personal wiki, but the documentation edits are useful enough that I don't really mind. [02:24:07] Carmela: By "Inclumedia shit" do you mean MirrorTools? [02:38:55] <^demon|away> leucosticte: No, I think that's a valid extension page and so forth. Inclumedia, MediaInclu and their various subpages belong on their own wiki I think is what Carmela is getting at. [05:54:15] https://www.mediawiki.org/wiki/Extension_talk:ConfirmAccount would use some more attention [12:18:26] Oi [13:37:34] Hi [13:39:17] I want to export my ZIM wiki to MediaWiki 1.19.18. I see it's possible to export from Zim to HTML. How can I import in MediaWiki 1.19.18 ? What-is the best way ? There is an extension ? [14:19:28] Halo [14:27:19] Wow, nobody else? https://www.mediawiki.org/wiki/Category:CustomDifferenceEngine_extensions [14:27:45] z1m is an optimist... converting HTML into wikitext... [15:40:58] any idea why i'm suddenly getting a PHP Fatal Error on a call to undefined function count() in ../includes/db/Database.php on line 4149? [15:41:10] count() hasn't been deprecated as far as I know [15:42:44] what version of MW? [15:43:01] 1.23.3 [15:53:50] wmat: Can you reliably reproduce that? [15:53:52] Did you update PHP [15:54:44] Reedy: I did not. And I just saw that pop up in my log file whilst investigating soemthing else. So I'm not sure what's caused it yet. [16:02:29] hello all [16:02:45] i finally could omit index.php from my wiki link [16:03:18] sasan: Yay! [16:03:41] now in each page, as i have toc, if i click on one sub section, it show me this link: http://parsintelligent.com/wiki/%D8%B3%DB%8C%D8%B3%D8%AA%D9%85_HDL-Bus#.D9.86.D8.B1.D9.85_.D8.A7.D9.81.D8.B2.D8.A7.D9.87.D8.A7.DB.8C_.D8.A8.D8.B1.D9.86.D8.A7.D9.85.D9.87_.D8.B1.DB.8C.D8.B2.DB.8C_.D8.B4.D8.A8.DA.A9.D9.87_HDL-Bus [16:03:49] which is not sutable [16:03:59] what should i do? [16:04:33] sasan: Does the page title (and section title) have a lot of unicode characters? [16:04:52] Ah, yes. [16:04:59] Not sure there's anything for it [16:05:06] Get a better browser? :D [16:05:28] My browser automatically "fixes" the page name at least [16:06:39] it is feature of browser? [16:06:42] are you sure [16:06:43] ? [16:07:11] Rewriting urlencoded unicode characters? Pretty sure. [16:07:27] Maybe also dependent on unicode support in whatever system font you're using. [16:08:13] i am using mozila [16:08:17] and test with ie [16:08:19] same [16:12:34] no other idea?? [16:12:50] Not currently, no [16:13:12] sasan: You could come up with some rewrite rules for urlencoded unicode, maybe, but I can't guarantee that will work for all clients [16:13:21] Also it would be painful [16:13:55] he he, so you suggest to forget?:d [16:14:01] Basically yes. [16:14:08] The characters aren't doing any harm [16:14:19] sasan: If you need to link places, ur1.ca is fantastic at making the URLs less shitty [16:17:03] how? [16:18:11] sasan: You take the longnasty link, paste it into ur1.ca, and they give you a shorter URL [16:18:27] It means sending your clients through ur1.ca for a redirect, but it's still useful IMO., [16:18:53] ur1.ca?? [16:18:57] Yup. [16:19:01] what is it? [16:19:07] It's a URL shortener. [16:19:28] how can i work wiyh it? [16:19:34] By going to the website [16:19:41] And Ring TFM [16:19:58] which website? [16:21:58] no it is not good:( [16:22:30] sasan: Why not? [16:22:54] i need such link: [16:23:14] parsintelligent.com/wiki/subsection [16:23:43] Well [16:24:00] That's not really in the cards AFAICt [16:32:38] sasan: if you do not need to link to sections, this should be doable. but section ids cannot at the moment contain unicode characters. [16:33:04] sasan: you can manually assign shorter/nicer ids to sections though. [16:33:56] DanielK_WMDE: Yes but that's apparently not good enough (or ur1.ca would have worked...right?) [16:34:03] like this: == Gödöllön pöllö töllöttää, möllöttää, köllöttää ja ööliä löllöttää == [16:34:10] you can then use #foo to link to that section [16:34:13] Whoa, that's neat. [16:34:31] marktraceur: that's not a mediawiki feature, that works in any html. [16:34:42] DanielK_WMDE: But DerpMediaWiki won't use it in the TOC which is annoying [16:34:46] mw just allows the id [16:35:05] (sure, it's allowed, I just hadn't thought of using it that way) [16:35:14] marktraceur: yea... considering the hack that is section numbering and linking, not to speak of toc generation, that's not surprising... [16:35:45] marktraceur: url shorteners do not work on fractions (stuff after the #) [16:36:12] Fragments, innit? [16:36:13] Anyway [16:36:17] TIL. [16:36:23] yes, fragments, sorry [16:36:29] sections in our context [16:37:13] hm, actually... shorteners *might* work on fragments [16:37:30] not sure whether http redirects can contain fragemnts in the Location header [16:37:33] understand my question? [16:38:15] https://stackoverflow.com/questions/2286402/url-fragment-and-302-redirects [16:38:21] interesting [16:38:35] sasan: not exactly, no [16:38:42] you need shorter/nicer links [16:38:52] but i don't know the exact use case or requirements [16:38:58] if i want to link subsection shown in toc [16:39:03] it show e sth like this [16:39:11] http://parsintelligent.com/wiki/%D8%AE%D8%A8%D8%B1%D9%86%D8%A7%D9%85%D9%87_%D9%87%D8%A7%DB%8C_%DA%A9%D9%84%DB%8C#.D8.A7.D8.B7.D9.84.D8.A7.D8.B9.DB.8C.D9.87_.D8.B4.D9.85.D8.A7.D8.B1.D9.87_13:_.D9.85.D8.B9.D8.B1.D9.81.DB.8C_.D9.86.D8.B1.D9.85_.D8.A7.D9.81.D8.B2.D8.A7.D8.B1_webserver [16:39:15] so it is not nice [16:39:26] DanielK_WMDE: Yeah, http://ur1.ca/ib4zd worked for me [16:39:33] It may be client-by-client though. [16:40:05] understand now? [16:40:40] sasan: "link it" how/where? "shown" when/where? [16:40:58] marktraceur: yea, i guess it would be [16:41:02] it is a link to one of my subsection [16:41:09] it is not pretty [16:42:21] sasan: do you want it to be pretty in the browser's url bar? do you need this for a few sections, or all sections? [16:42:41] all sections [16:42:59] generally, browsers have only recently started to support non-ascii characters in urls, especially in fragment ids. [16:43:18] sasan: and you want them to be pretty in the browser's address bar? or when putting links on other wiki pages? or what? [16:43:36] in all, [16:43:47] basically, mediawiki encodes all non-ascii characters, because not all browser support non-ascii characters there [16:43:50] espically when i want to send it to another quy [16:44:02] when you want to send it, use a shortener [16:44:13] when you want to link, use an explicit link text [16:44:29] in the address bar... sorry, i don't think there is currently any way to do that [16:44:37] and if there was, it wouldn't work in some browsers [16:44:55] ok, thanks;) [16:46:29] sasan: oh, and as i said, for sections you often link to, you can assign an id manually, to make nicer links. [16:46:43] how? [16:46:47] let me know an id [16:46:49] also has the advantages that these links don't break if you change the name of the section [16:47:05] like this: == Gödöllön pöllö töllöttää, möllöttää, köllöttää ja ööliä löllöttää == [16:47:24] you mean i use as its topic? [16:47:40] i mean you use some html tag anywhere, and give it an id [16:47:48] you can then link to that id as a fragment id [16:47:58] this is basic html, mediawiki just passes it though [16:48:10] could you show me somewhwre [16:48:14] you add this to see [16:50:10] i have test it [16:50:15] same answer [16:51:41] sasan: the id you assign manually is not used in the table of contents. you have to type it into the url manually [16:51:55] if you use id=foo, and you add #foo to your url, you should go to the right sectionj [16:51:57] how? [16:52:04] could you remote and show me? [16:52:32] by teamviewer [16:52:35] sorry, no html 101 today [16:52:35] could? [16:52:38] please [16:52:43] please [16:52:47] 'just 5 min [16:52:52] sasan: He said no. [16:53:06] you can> [16:53:10] ? [16:53:16] No. [16:53:37] google "link to id attribute" [16:54:52] * DanielK_WMDE wonders what the hell that finish sentence even means. [16:55:37] DanielK_WMDE: Which? [16:56:22] the one i used as an example heading :P [16:56:49] Finnish. You know. [16:57:22] Ah. [17:34:02] anybody ever have a problem with AbuseFilter stating that an edit matches a filter but yet it did not apply the filter's actions to that edit? [17:34:44] I don't see any reason for it to have failed so I am left thinking I've hit a bug [17:37:47] Hi all. https://he.wikipedia.org/wiki/%D7%A2%D7%95%D7%9C%D7%9E%D7%95%D7%AA - inserting «Olamot Con» in the add languages page gives me this - http://www.shlomifish.org/Files/files/images/he.wikip-error-dollar1.png [17:52:26] I just tested by repeating the edit and AbuseFilter let me do it. [17:52:37] it claims the filter applies, but does not apply it. [17:56:43] ok NOW it's working. [17:57:07] apparently the text "new section" is added into a change's summary by MediaWiki *after* AbuseFilter fires [17:57:18] so you cannot test against that text being in the summary [17:57:36] it'll pass on past edits, but not ones in progress. [18:05:50] hey cool the (details) link on AbuseFilter changes dies with a php exception of some sort :> [18:06:04] time to debug that as well I guess [18:06:57] Fatal error: Call to undefined method TableDiffFormatterFullContext::_start_diff() in /home/doomwiki/mediawiki-1.23.2/extensions/AbuseFilter/Views/AbuseFilterViewDiff.php on line 30 [18:07:20] awesome. [18:07:41] MatmaRex: congrats [18:08:28] QuasAtWork: Old verison of Abusefilter? [18:08:33] master is $this->startDiff(); [18:10:18] was the version recommended for use with 1.23.2 release [18:10:49] I'll try hot patching it [18:11:33] QuasAtWork: if you're using Git, checkout REL1_23 in the AbuseFilter directory [18:11:38] there are more problems, ack [18:12:17] doesn't look like my sys admin used git to do the install [18:12:22] he did recently install it though [18:14:12] QuasAtWork: grab the tarball from the extensions download page [18:16:03] current stable, or latest dev version? [18:19:14] stable looks good. [18:23:00] fixed up; thanks. [18:52:08] hello [18:52:49] I have installed the mediawiki successfully and gone through the gerrit tutorials [18:53:45] What is the next step I should follow [18:54:28] alisha: What are you trying to do? [18:54:29] :) [18:54:33] alisha: find an "easy" bug you want to work on. [18:54:47] (assuming you are trying to get up to speed with mediawiki development) [18:54:57] alisha: or, implement some feature you always wanted. [18:56:30] alisha: https://bugzilla.wikimedia.org/buglist.cgi?bug_status=NEW&bug_status=REOPENED&columnlist=product%2Ccomponent%2Cchangeddate%2Cbug_severity%2Cpriority%2Cshort_desc&keywords=easy&keywords_type=allwords&list_id=349631&order=changeddate%20DESC%2Ccomponent&product=MediaWiki [18:56:42] that's the list of bugs/feature requests tagged as "easy". [19:00:54] Thanks danielK_WMDE [19:01:07] alisha: You can also check this tutorial here https://www.mediawiki.org/wiki/Annoying_little_bugs [19:01:48] alisha: and this one : https://www.mediawiki.org/wiki/How_to_become_a_MediaWiki_hacker [19:02:49] alisha: btw: not knowing where to start is normal. don't hasitate to ask, here or on the mailing list [19:02:50] Can u suggest me any bug related to parser or file conversion [19:03:09] and it's always good to idle here, and peek at the conversations going on. [19:03:13] a good way to learn about stuff [19:03:43] alisha: parser and file conversion are pretty hard topics... [19:04:03] start with something simpler, even trivial, like a style change. [19:04:18] add me as a reviewer, if you like [19:04:25] anyway, enough for today. [19:04:27] have fun! [19:04:28] Ok but I am interested in it because I have worked upon it using c++, Flex and bison [19:05:28] DanielK_WMDEHow can I add you as a reviewer? [19:06:01] sure [19:06:17] alisha: On the Gerrit change, just use the text field that says "Add a reviewer" [19:06:41] alisha: if you have worked with Bison, you do *not* want to work on the mediawiki "parser". it's not a parser. there is no grammar. there is no ast. [19:07:00] it's a horrible mess of regular expression that what around the wikitext until it resembles html [19:07:15] good people have lost their minds trying to fix it. don't go there... [19:07:16] Ok [19:07:21] ;) [19:07:36] https://www.mediawiki.org/wiki/Alternative_parsers [19:07:43] alisha: You could check out Parsoid, though. It was going to use some combination of Flex and Yacc at one point I think. [19:07:45] none of them is really compatible [19:07:58] Right now it's just a multi-pass parser in node.js [19:08:17] Uses peg.js which is fun times [19:08:18] It uses pegjs, doesn't it? [19:09:12] RoanKattouw: jinx [19:09:27] one token of look-ahead === misery [19:09:46] unlimited look-ahead with linear runtime === some sanity [19:10:24] also, lots of context-sensitive stuff that you can't do in any context-free grammar [19:10:46] OK [19:11:42] So according to you I should go for some other topics before these [19:12:15] Which are simpler than this [19:25:36] Good Evening, Please could someone provide some advice. I work at a hosting company and I have been given an issue of which our eu has installed MediaWiki on there shared hosting and they are having an issue when they login there get a problem loading page. I have tried uninstalling and reinstalling the application however this has not resolved the issue and I could not find anything within the manual http://www.mediawiki.org/wiki/Manual:Errors_an [19:25:36] d_symptoms [19:26:37] When I try and log in on my browser iceweasel I get Server not found, cant find the server at index.php [19:27:47] mm [19:28:05] could be a problem with cookies?S [19:28:12] could what? [19:30:35] I have reset my browser and am getting the same thing, it's a peculiar one. [19:31:22] jjacklin: I'm guessing it's not a browser issue [19:31:36] Is it open to the internet? [19:32:32] It sorta sounds like bad rewrites [19:36:39] heya Reedy yeah I believe it is open to the internet [19:36:44] link? :) [19:36:50] Have you checked the apache/webserver logs? [19:37:11] www.rekindler.net is the eus website [19:37:35] jjacklin: Yeah, it's bad rewrite rules it seems [19:37:36] http://index.php/?title=Main_Page [19:38:14] cool, do you know where I need to go to resolve it? do I need to edit a file? [19:38:52] It might just be a .htaccess [19:39:03] I'm guessing if it's shared hosting, they can't alter their webserver config? [19:40:06] no they can, I mean I have ssh'd onto the server of which it is hosted on and navigated to their hosting [19:40:42] but thats as far as I got and read through the manual could find what I was looking for and thought I would ask you guys once I saw the irc help channel [19:40:46] cant* [19:41:15] What webserver is it? [19:41:57] SMP Debian 3.2.60-1+deb7u3 x86_64 [19:42:06] I mean the actual webserver [19:42:13] Apache? [19:42:50] yeah I think its Apache [19:43:06] so it's either a .htaccess in the root of the folder with mediawiki [19:43:13] or else in /etc/apache2/sites-enabled [19:50:52] cool thank you Reedy lol I have been quiet because I have been looking. No joy yet although I do not fully understand how it is setup. Typically it would be in the root files but just looked in /etc/apache2/sites enabled and nothing [19:51:12] Did you look for a .htaccess? [19:51:23] just looking for that now [19:52:25] just did a search for it and it brought back quitea few locations although I would imagine ./cache/ .htaccess sounds the best out of them [19:52:53] Probably not [19:53:01] That should just be there to stop casual browsing [19:59:07] Thank you for your help with this Reedy, what I have decided to do is escalate this to the team above me with the details of which you have helped me with and the direction you have pointed me in! Really appreciate it. They put the system in place so I will let them look for the file as they may find it in an instant. [22:56:55] wmat: around? [23:45:10] jackmcbarn: yes [23:45:32] wmat: i saw your LQT post. is LQT something you're stuck with from when you got the wiki, or is it new? [23:45:59] jackmcbarn: it was added, but users are liking it [23:46:10] you'd have been better off adding Flow [23:46:12] LQT is dead [23:46:28] * wmat looks up Flow [23:47:16] requires MW 1.24+ [23:47:33] aren't you on git master? [23:47:43] at the moment, 1.23.5 [23:47:53] i do intend on moving to master though [23:48:10] haven't gotten to it yet [23:48:24] oh, i see, that's just what the extension page says [23:48:29] use the 1.23 branch and it [23:48:33] it'll work with 1.23 [23:49:34] cool, i'll give it a shot [23:49:49] will old LQT posts still be available? [23:50:03] i'm not sure about that [23:50:11] you should join #wikimedia-corefeatures and ask there [23:50:19] (make sure to mention that this is for your own wiki) [23:51:05] it's not a big deal, i'd expect nobody would miss the old posts anyway [23:51:31] btw, my Scribunto problems went away [23:51:36] i know for sure that it won't "just work", and i also know for sure that once flow is mature, then sites like mediawiki.org will have LQT migrated to it, but i'm not sure where in the process we are [23:51:38] how? [23:51:42] i'm not entirely sure yet [23:52:09] i did a runJobs.php, then update.php and they were working again [23:52:45] i'm still looking into it [23:55:20] also, i asked WikiApiary to scan for wikis with Oracle databases and there are exactly zero [23:55:50] this helped convince the decision makers to let me use MySql [23:56:03] :D [23:56:24] perhaps Oracle should be removed as a supported database [23:56:28] there are probably private oracle wikis I'm guessing [23:57:07] i know of one, but it's uses almost no extensions and is at MW 1.18 [23:58:29] anyway, i'm happy i don't have to use it [23:58:38] :)