[00:10:59] Subfader: looks like namespaces=0|14 don't work for opensearch at all [00:11:48] and opensearach is 3rd party? [00:17:51] i'm using 14 only since cats is what people search for me on y wiki, so you helped me nontheless. thanks a lot! [00:20:00] Subfader: 0e054bbf1a955b02afd4feaa04063c841e19a569 attempted to fix this, but had to be reverted [00:20:14] Subfader: only one namespace can be searched (the first) [00:21:08] !bug 30323 [00:21:08] https://bugzilla.wikimedia.org/buglist.cgi?quicksearch=30323 [00:21:21] couldn't each NS cause a new fetch? [00:21:47] eh, results would need to be combined. [00:22:14] "But Google can do it!" Users. Always. [00:29:43] Subfader: there is a fix in change 141660 [01:00:04] saper: Can you point me to the diff? I'm not confirm with all the different MW ressources [01:05:22] nevermind, got it https://gerrit.wikimedia.org/r/#/c/141660/ [01:06:29] Subfader: yeah, but the patch got broken in the meantime [01:06:41] patch 9 was good, the last one (patch 11) is messed up [01:16:17] Ah great. I had no function search() in PrefixSearch.php so I just replaced the whole file. Works from what I see. Using all NamespacesToBeSearchedDefault. [01:16:28] Thanks again! :Y: [03:26:54] So, I have written some time ago some C++ code which can parse MediaWiki dumps (both SQL and XML) [03:27:31] I suspect it may be useful to other people; is there a place on mw.org/Meta where I'd put info about it? [03:28:14] https://www.mediawiki.org/wiki/Manual:Restoring_a_wiki_from_backup ? [03:28:23] or is it just a parser...? [03:29:15] It's just a parser I wrote when I needed to do some analysis on dumps [03:29:37] https://github.com/vasilvv/mwdumptools [03:30:22] https://www.mediawiki.org/wiki/Alternative_parsers I guess then [03:31:11] Uh, it's not a markup parser, it is a dump parser [03:31:55] some of the things on that list look like dump parsers... :/ [03:32:26] maybe just create a new page for it? :) [06:45:34] hello all [06:47:38] i enable Mod_rewirte in my host and add RewriteEngine On RewriteCond %{REQUEST_FILENAME} !-f RewriteCond %{REQUEST_FILENAME} !-d RewriteRule . /index.php [L] to .htaccess and add $wgArticlePath = "/$1"; $wgUsePathInfo = true; to local setting [06:47:44] but i can not have short url [06:47:47] it show error [06:47:49] 404 [06:47:54] could anybody help? [06:56:09] could anybody help me with it [06:56:11] ?? [07:23:29] no body could help me?? [07:29:38] i enable Mod_rewirte in my host and add RewriteEngine On RewriteCond %{REQUEST_FILENAME} !-f RewriteCond %{REQUEST_FILENAME} !-d RewriteRule . /index.php [L] to .htaccess and add $wgArticlePath = "/$1"; $wgUsePathInfo = true; to local setting [07:29:50] but i can not delte index.php form my url [07:34:01] online? [07:34:04] could help me? [07:38:32] Hi folks, there was an extension a few years ago (which possibly still exists) that allowed you to write PHP-code directly into an article and run it. Does anyone know the name of it? [08:04:11] i enable Mod_rewirte in my host and add RewriteEngine On RewriteCond %{REQUEST_FILENAME} !-f RewriteCond %{REQUEST_FILENAME} !-d RewriteRule . /index.php [L] to .htaccess and add $wgArticlePath = "/$1"; $wgUsePathInfo = true; to local setting [08:04:23] but i can not delte index.php form my url [08:12:35] Is the noprint advice in https://www.mediawiki.org/wiki/Manual:Skinning#Print_mode still valid? [08:14:34] I think so, PDF printing inclusion via special template does not work anymore [08:19:55] could you help me? [09:02:28] why no body asnwer me??? [09:31:26] Please asnwer meeee [09:46:03] sasan: can you repeat your question please [09:52:37] : i enable Mod_rewirte in my host and add RewriteEngine On RewriteCond %{REQUEST_FILENAME} !-f RewriteCond %{REQUEST_FILENAME} !-d RewriteRule . /index.php [L] to .htaccess and add $wgArticlePath = "/$1"; $wgUsePathInfo = true; to local setting [09:52:45] but i can not delte index.php form my url [09:57:24] understand what i mean? [09:58:23] No not exactly [09:58:37] actually i am new to it and just learning [09:59:39] i want to have short URL, deleting index[php [09:59:47] index.php from my link [09:59:57] ok [10:00:27] now could help me? [10:01:05] I can't be sure but I am searching for it [10:01:33] If I found something useful to you I will definitely advice you [10:01:57] Thanks [10:41:06] please see this website [10:41:07] https://www.mediawiki.org/wiki/Manual:Pywikibot/Scripts [10:41:26] i want to add parentlink lik ethat [10:41:33] could anybody help me???? [10:51:41] alisha : please see this website [10:51:45] https://www.mediawiki.org/wiki/Manual:Pywikibot/Scripts [10:51:50] i want to add parentlink lik ethat [11:06:48] understand what i mean? [11:45:26] hi, I'm having this problem and can't see page content. preg_match(): Compilation failed: group name must start with a non-digit at offset 8 in /var/www/html/wiki/includes/MagicWord.php [11:54:25] i want to have parentlink like this website https://www.mediawiki.org/wiki/Manual:Pywikibot/Scripts [11:54:33] could anybody help?? [11:58:09] jgeek, https://www.mediawiki.org/wiki/Thread:Project:Support_desk/MediaWiki_don%27t_work_with_PCRE_8.34_%282%29 [11:58:35] you could help me? [11:59:02] sasan: Why do you ping me specifically? And what have you tried so far to find out how to do it? [11:59:22] sorry [12:24:49] sasan http://www.catb.org/~esr/faqs/smart-questions.html [12:26:03] what do you mean by that? [13:10:14] how to add embed video to mediawiki ? I have mediagoblin embed code [13:13:07] jgeek: Depends. There are extensions that allow embedding videos: http://www.mediawiki.org/wiki/Category:Video_player_extensions [15:04:10] hey all.. my mediawiki has been the most trusted part of my setup for a few years now [15:04:27] but now.. today when tried to bring up my wiki page.. it was loading a white screen [15:04:47] and when I turned on php display errors in apache, I see this one error repeated over and over again [15:04:50] Notice: Use of undefined constant c - assumed 'c' in /var/www/jf/mediawiki-1.18.0/index.php on line 1 Notice: Use of undefined constant c - assumed 'c' in /var/www/jf/mediawiki-1.18.0/index.php on line 1 [15:05:39] I almost never log into this box, so I have no real idea how this happened, unless there was an intruder... but do you have any idea how I can correct this? [15:13:42] bluethundr: Check to see if index.php has been changed - use ls -l or something to get the timestamp [15:24:14] marktraceur: nope! no suspicious change dates for index.php [15:24:16] so odd [15:24:48] Weird. [15:25:18] bluethundr: Any other files in the mediawiki directory changed? Did PHP get updated? [15:27:24] hmmm.. I think that puppet might have pushed out a php.ini that was bad for this host [15:27:45] that's probably where my issue lies [15:28:00] fortunately our backup system is pretty decent [15:30:52] Sweet! [15:45:54] yup! it was a funky php.ini ... TGFB .. thank god for backups [15:45:57] ;) [15:54:32] hi, can someone guide me into submitting a patch [15:56:24] ac16_: Sure, you should probably join #wikimedia-dev though [15:56:40] will do. thanks [15:57:20] I have set up my system, what is the first step to start contributing? [16:44:11] Hi, I'm trying to extract image url from file name when it has unicode characters. I have this "François-Antoine Kirmann, chef d'escadron des mamelouks de la Garde impériale (1808-1811).jpg" and the url part is /wikipedia/commons/3/32/ but no matter what I try I cannot encode the file name in a way that its MD5 start with 32 [16:44:14] is anyone working on the performance portal project? [16:45:37] ac16_: What do you need to know about it? [16:46:09] Guest82901: You can get the thumbnail URL from the API without much trouble, is there some reason you aren't? [16:46:32] I would like to know a good starting point towards the project, I have already set up my system and created the ids [16:46:34] yes, I am processing the whole dump [16:47:25] Obviously calling API for every single file is not an option [16:47:59] Guest82901: "obviously" but OK [16:48:22] Guest82901: Maybe I should start by asking what you're trying to accomplish [16:50:18] marktracer: long story, but it comes to this that I need to extract thumbnail url from image names [16:50:27] I have time [16:50:35] (I'm trying to work out your issue in the background) [16:51:11] marktracer: so far I managed to do it for ASCII file names (MD5 works fine) but it doesn't work for files with unicode [16:51:26] for example I tried: François-Antoine_Kirmann%2C_chef_d%27escadron_des_mamelouks_de_la_Garde_impériale_%281808-1811%29.jpg [16:51:33] or Fran%C3%A7ois-Antoine_Kirmann%2C_chef_d%27escadron_des_mamelouks_de_la_Garde_imp%C3%A9riale_%281808-1811%29.jpg [16:51:48] but non of them have a MD5 starting with 32 [16:52:05] I agree, I'm hitting the same problem [16:53:59] marktraceur: thanks. I tried MD5 from different frameworks too both with ASCII & UTF8 encodings. No success [16:54:25] Yeah, this is odd [16:54:32] Guest82901: What programming language are you using? PHP? [16:54:51] I'm wondering if production has a weird encoding by default [16:54:59] I tried Java, .NET and PHP [16:55:10] Hm. [16:57:52] Maybe the file got moved [16:58:28] Nope. [16:58:33] By any change are you aware of the smallest word with unicode character in commons? A reverse-MD5 table may help [16:58:40] * any chance [16:58:51] Not sure, no [16:59:03] I'm wondering if there's something messy that happened with UploadWizard [17:00:00] oh. that's awful [17:00:55] Is it suppose to use UTF8 or ASCII for upload? [17:01:01] Guest82901: Can you try it on https://commons.wikimedia.org/wiki/File:Penne_pasta_on_top_of_meat_sauce_with_italian_sausage.jpg ? [17:01:12] No UTF8 characters [17:01:37] Oh, no, I can get it with the underscores [17:03:25] My code return df for MD5 and http://upload.wikimedia.org/wikipedia/commons/d/df/Penne_pasta_on_top_of_meat_sauce_with_italian_sausage.jpg as url [17:03:30] which seems to be correct [17:03:50] gi11es: Do you happen to know where exactly we decide what the md5sum of the filename is? I'm rusty on my filebackend [17:10:19] marktraceur: thank you for your help, I have to leave now but I'll check the channel's log later [17:10:41] KK [17:10:43] Good luck! [17:39:04] what is the helpdesk email address please? [17:39:16] hey guys.. new issue... I've upgraded from mediawiki 1.18 to 1.23.5 and even tho I have this value set, the wiki logo does not appear [17:39:20] http://pastie.org/9625400 [17:39:31] the place where you'd expect to see the logo are just blank [17:39:33] any ideas? [17:39:51] I need to make a suggesstion about the definition of cybersecurityin wikipedia [17:41:32] bluethundr: Can you make sure that the file is still there? [17:48:14] Hi marktraceur, sure I did that in the paste. I list out the file to show that it's there [17:48:37] Oh, sure [17:49:00] bluethundr: Maybe $wgStylePath changed [17:49:09] bluethundr: Try replacing it with $wgScriptPath [17:49:25] 'cause it looks like your image isn't in skins/ [17:49:53] ok let me check it out.. thanks! [17:50:18] No problem :) [17:52:02] hey marktraceur .. ok that did it.. the logo's there now. Thanks! [17:52:14] Woohoo! [17:55:56] A lot of helpful people here. :) [17:56:35] We live to serve, etc. [17:56:50] marktraceur :) [17:59:01] Does Wikimedia support the »rel="next"« attribut? [17:59:13] Graf_Westerholt: In what context would we? [17:59:25] Media Viewer, maybe. I guess that's on me. [17:59:37] Search results too [18:00:00] Looks like history pages have it at least [18:00:21] marktraceur, for example here it would be usefull: http://en.wikibooks.org/wiki/Prolog/What_is_Prolog [18:00:44] Ohhh dear. [18:00:49] That's...probably not. [18:01:03] I don't think there are *any* magic words for internal links. [18:01:22] Sorry, marktraceur, I do not understand you. :D [18:01:36] Graf_Westerholt: So, that's in content [18:01:49] Which means the *content* would need to specify rel=next to the parser somehow [18:02:01] But I don't think there's any way to do that with the [[internal link]] syntax [18:02:25] marktraceur, sorry, I do not know much about Wikimedia, so I do not know what you are talking about. ;) [18:02:50] Graf_Westerholt: https://www.mediawiki.org/wiki/Help:Links#Internal_links [18:02:55] !wikipmediawiki [18:02:55] Confused about the differences between MediaWiki, Wikimedia, Wikipedia and wiki? See https://www.mediawiki.org/wiki/Wikipmediawiki [18:03:02] just a heads up.. the peeps in this channel rule :) [18:03:29] * Graf_Westerholt reads [18:13:34] marktraceur, I am not confused about Wikimedia. I know Wikimedia is a Software used by many projects, like Wikipedia, Wikibooks, Wikivoyage … ;) [18:14:07] No [18:14:15] Graf_Westerholt: Wikimedia is a movement. MediaWiki is software. :) [18:14:29] ok :D [18:14:42] Hence the link! [18:14:59] :) [18:16:42] Is there anything besides $wgHiddenPrefs that would cause Real name to *not* display on the preferences page? [18:16:55] Rosencrantz: CSS? [18:17:28] Rosencrantz: Where are you seeing that? [18:18:07] On an internal instance I run at work. [18:18:14] marktraceur, so there is no "easy" way to use rel="next"? It would be useful for a friend’s project. I suggested to use this but he does not know how to use it in MediaWiki. [18:18:39] Graf_Westerholt: I don't think so...I mean, you could write an extension to do it [18:18:49] Basically just adding a tag for internal link that's also a rel=next [18:18:53] So there is even no extension? [18:18:54] I think it has something to do with authentication, but I can't figure out what. When it's using LDAP it works. When it's using SAML it disappears. [18:19:15] marktraceur, I wonder why MediaWiki does not support this. [18:19:16] And the SAML extension I'm using sets the data, so it's there…. it's just not… displaying [18:19:26] Graf_Westerholt: Probably just an oversight. [18:19:44] I hope it will be supported in the future. ;) [18:20:02] It would be nice! [18:20:10] Not terribly difficult to add an extension tag, either [18:20:19] Graf_Westerholt: https://www.mediawiki.org/wiki/Manual:Tag_extensions [18:21:10] See, it is also usefull for Wikibooks. [18:21:22] It would be, I suppose [18:22:01] And WikiSource. I guess they don't have rel=next either. [18:22:51] afk [18:23:10] FWIW, the browser I now use for most of my Web reading things has rel="next" and the like on the hot keys by default. Giving an easy way to supply those for Wikibooks and the like would be quite a improvement. [18:44:22] ! [18:44:22] Hello, I'm wm-bot. The database for this channel is published at http://bots.wmflabs.org/~wm-bot/db/%23mediawiki.htm More about WM-Bot: https://meta.wikimedia.org/wiki/wm-bot [18:44:32] sucheta: ? [19:03:37] Is there any way to get all the users with a certain right? [19:04:47] I want to send a notification to every user with a right (for example, with the delete right) [19:04:48] Polsaker: via the API? https://www.mediawiki.org/w/api.php?action=query&list=allusers&augroup=sysop [19:04:56] oh [19:04:57] rights? [19:04:58] legoktm, from an extension [19:05:08] hmm [19:05:11] Yeah, rights [19:05:46] you can do User::getGroupsWithPermission( 'delete' ); [19:05:54] and then query for users in those groups [19:06:17] marktraceur, I am looking at the Tag_Extension. I am not sure if that works because the rel="next" has to be in the HTML-Head. [19:06:43] legoktm, might need to do more work if you want compatibility with things like CentralAuth [19:07:11] uh [19:07:18] I need it to work with CentralAuth [19:34:18] Graf_Westerholt: Does it? I think you can put rel="next" on any tag... [19:34:56] Graf_Westerholt: Or do you mean the spec link? That would need to be in the head, but you can detect whether there are any rel="next" links and add it conditionally. [19:47:23] marktraceur, I mean » Ah. [19:48:14] Graf_Westerholt: Also possible. [19:48:21] marktraceur :) [19:48:47] Not sure if it's necessary, though, if you just put the attribute on the tag. [19:49:27] It has to be in the head with link. [19:51:09] aaand... is there a way to get all the users in a group without calling the api (from inside the extension)? [19:53:19] Graf_Westerholt: Why does it have to be? [19:53:29] Polsaker: DB calls, I guess [19:53:35] marktraceur, it only works this way. [19:54:21] Graf_Westerholt: Is that what the spec says, or is this a bug in your browser's implementation of the spec? [19:54:22] ugh [19:54:31] Polsaker: Oh, come on, SQL is fun! [19:54:33] marktraceur, specs say so. [19:54:36] * marktraceur puts on a party hat [19:54:42] That is what they said... [19:55:11] Graf_Westerholt: Interesting. OK, file a bug against mediawiki core saying that the spec wants a , and link to the spec section that says so please :) [19:55:23] :D [19:56:51] Whoa, we already do a lot of cool stuff in the head. [19:56:57] [19:57:00] Whaaaaat [19:57:04] marktraceur http://www.w3.org/TR/REC-html40/struct/links.html#h-12.1.2 [19:57:40] Graf_Westerholt: That doesn't specify a requirement that it be in the head... [19:58:47] A comment extension adds comments by editing the talk page using doEdit(). But the passed text is trimmed. How can I force a line break after each entry? "\n" and [real line break in php file] are cut off. [19:59:19] Subfader: Why do you need a line break after an entry? [19:59:59] So humans can still read the talk page when moderating [20:00:19] Subfader: Maybe check if there's text in the page already, and if so, prepend two line breaks instead. [20:00:55] Also, don't we already have a comments extension thing? [20:01:08] Like, five versions of it? [20:01:33] All \n are trimmed. [20:01:50] Yes and none rocks :) I hack ArticleComments atm [20:01:53] Subfader: That's...weird. Even in the middle? [20:02:24] no. only in front and at the end. [20:02:30] OK [20:02:34] So that's a start [20:02:41] HTML comments do but look ugly [20:02:51] Subfader: But you're putting lines *between* sections, right? [20:02:59] Oh, but doing it by adding a new section. [20:03:17] So just take the current text, append two newlines and your text, then perform a normal edit with that. [20:03:32] Instead of a section-add edit. [20:03:37] marktraceur, see the code of http://www.w3.org/TR/REC-html40/struct/links.html#h-12.1.2 and check the head for the link next ;) [20:03:57] Graf_Westerholt: Examples are not specs [20:04:01] It's composed like $commentText = "foo bar baz\n\n"; doEdit($commentText); [20:04:16] Graf_Westerholt: As far as I know, rel= can be placed on any or and "next" can be used on any rel attribute [20:04:28] marktraceur, but for the "buttons" it has to be in the head with link. [20:04:38] marktraceur, but not for "buttons". [20:04:41] Subfader: MediaWiki by design trims the contents of page text before saving [20:04:46] Graf_Westerholt: Again, that sounds to me like a bug in your implementation [20:04:55] marktraceur, no. [20:05:28] marktraceur, I tested in three browsers. No one shows a button if not put link rel next in the head. [20:05:44] Vulpix: Yes, and I need to hack it ;) [20:05:45] But with link rel next the show. [20:05:52] Graf_Westerholt: So at https://en.wikipedia.org/w/index.php?title=User_talk:Jimbo_Wales&action=history you don't see the next button? [20:05:58] Wait, what browser are you using? [20:06:13] marktraceur, right, no next button. [20:06:44] It is Mozilla SeaMonkey, Opera 12.16 and Konqueror. [20:07:07] Graf_Westerholt: I've found https://addons.mozilla.org/en-US/firefox/addon/space-next/?src=search which seems to support A tag definitely needs to be in the head [20:18:20] But can be in the body, and it can have rel="next" [20:18:56] marktraceur: indeed, but that kind of magic only works when it's present in the head [20:19:11] and don't ask me why [20:19:21] marktraceur, with a it is valid. But with Gotta be a SeaMonkey bug, right? [20:19:28] Gotta be [20:19:37] marktraceur, so try Opera 12.16. [20:19:49] the fact is that we have elements [20:20:15] marktraceur, and also try Konqueror with konq-plugins. ;) [20:20:43] You can also try Lynx. ;) [20:20:44] Vulpix: True. [20:20:53] marktraceur, four browser with the same bug? [20:21:15] Graf_Westerholt: Yup. [20:21:20] Because they all mimic each other :) [20:21:21] And w3.org has the link next in the head, too. [20:22:47] And hilariously don't mark the [20:23:00] Got it. MW trims the \n at the very end of the page text, which of course is where every new entry ends. Replacing the merged ones now with $wgHooks['ArticleSave'] [20:24:05] marktraceur, so what I want to do is having a link rel next. ;) [20:25:23] Graf_Westerholt: I filed a bug anyway https://bugzilla.mozilla.org/show_bug.cgi?id=1078697 [20:25:33] We'll see what they think [20:25:43] It may be that they considered it and figured it would be too confusing [20:25:49] Mm, it is not quite a bug I think. But ok. [20:26:00] Why confusing? [20:27:01] It adds another element for doing the same thing, when the is already there [20:27:43] It is not a bug of SeaMonkey. [20:29:01] The spec says: "Links specified by LINK are not rendered with the document's contents, although user agents may render them in other ways (e.g., as navigation tools)." [20:29:26] No, of course not [20:29:34] But that doesn't preclude the use of tags [20:29:39] We're talking in circles here. [20:30:09] https://developer.mozilla.org/en-US/docs/Web/HTML/Link_types [20:30:46] > Not allowed in these elements: None [20:31:04] allowed in , [20:31:20] what about the noble gases [20:31:31] MC8: What [20:32:18] it was supposed to be a hilarious joke about elements but I'm not sure what I was trying to get at [20:32:24] Vulpix: Oh, man, , wow. [20:32:40] MC8: Oh, gotcha [20:32:46] * marktraceur laughs awkwardly [20:33:22] I sort of want to build a sane nav toolbar for Firefox now. [20:33:30] Curse you, long list of projects [20:33:46] marktraceur, what nav toolbar? [20:34:06] Graf_Westerholt: Like the one in Seamonkey [20:34:12] marktraceur, this exists. [20:34:15] Something that gives you next/prev etc. [20:34:18] I saw one, but it was crappy [20:34:49] marktraceur, this? http://www.bolwin.com/software/snb.shtml It is not working since FF 32 or so. [20:35:43] If I want to improve a change made in a commit (that wasn't approved), should I follow these steps https://www.mediawiki.org/wiki/Gerrit/Tutorial#Amending_a_change or should I make a whole new different commit? [20:35:53] Graf_Westerholt: Yeah, that's it [20:36:23] marktraceur, ok. Maybe you can ask to repair and improve it. ;) [20:36:43] * Graf_Westerholt has to go to bed. It is late here and I have to get up early. ;) [20:36:50] tuxilina: yes [20:37:17] yes to follow the steps or yes to make a new one? [20:37:28] tuxilina: yes means "follow the gerrit tutorial" instead of making a new commit [20:37:32] sorry [20:37:34] See you tomorrow! [20:37:54] thanks [20:51:18] tuxilina: in short, "git add... " again and "git commit --amend" [21:10:04] saper: Thank you, it worked [23:45:42] How do I download the Vector skin via Git, for MW 1.24? [23:46:02] (And why isn't that information easily available on mediawiki.org?) [23:46:41] (...or rather, could someone please make that information easily available on mediawiki.org?) [23:47:07] Yaron: the 1.24 branches haven't been created yet. [23:47:23] Can't I just get the latest code for it? [23:48:09] Probably [23:48:24] legoktm: (or anyone else) what's the call to do that? [23:48:37] just clone mediawiki/skins/Vector.git ? [23:49:12] legoktm: I called this: "git clone http://git.wikimedia.org/tree/mediawiki%2Fskins/Vector.git" [23:49:28] ...and it returned the following: "warning: remote HEAD refers to nonexistent ref, unable to checkout." [23:49:32] uhh [23:49:42] try $ git clone https://github.com/wikimedia/mediawiki-skins-Vector [23:49:43] ? [23:50:29] legoktm: aha! Thanks; that seems to have worked. [23:50:58] I think it's quite bad that there's no documentation for this anywhere; but that's not your fault (I assume). [23:51:53] * legoktm spins https://en.wikipedia.org/wiki/Wikipedia:BLAMEWHEEL [23:52:10] It's Jimmy's fault. [23:52:14] !bug 1 [23:52:15] https://bugzilla.wikimedia.org/buglist.cgi?quicksearch=1 [23:52:20] :)