[07:32:09] test [07:32:24] error, test failed [07:32:29] it worked [07:33:40] :) [07:50:53] where do we set $wgAllowExternalImages to true?? [07:59:30] impacific: LocalSettings.php [08:02:07] thanks jimt :) [08:03:44] Is there a way to make TemplateData work with Lua modules? Or is the pattern to wrap each Lua function in a template? [08:05:39] you should wrap lua modules in templates [08:05:46] and then use templatedata with the template [08:06:00] Sigh. I was afraid that was the answer. Thanks. [08:06:49] you don't think that's a good idea? [08:07:29] I would like Lua to replace the templates, not just provide underlying plumbing. [08:08:31] It seems to make it more brittle to put some of the implementation each place. Someone will change/break one and not the other. [08:09:12] maybe. I think it's like parserfunctions, you don't put #ifeq in the article directly, you wrap that in a template, and use the template in the article [08:12:05] Is there a way to specify TemplateData that is a list of options for a parameter? I have a template that only would make sense for one of 6 selections. An option list in the VisualEditor would eliminate users typing one of the 6 wrong when using the template. [08:13:02] I don't know, better place to ask might be #mediawiki-visualeditor, but I'm guessing most people in there who would know are asleep [08:14:09] OK. Thanks. [11:28:50] We have a problem with edit block spam filter [11:29:03] It's triggered when an editor adds a spam link [11:29:16] but not in the case the page was in draft: and moved to mainspace [11:29:41] it's a way to overcome the filter [11:30:01] move page to draft, add spam links, move back [11:30:19] is this something for bugzilla? [11:34:10] magioladitis: just ask the admins to add the draft namespace [11:34:32] What filter are you expecting to be triggered? [11:52:16] Nemo_bis: Sorry. I lost connection [11:53:42] 13.34 < Nemo_bis> magioladitis: just ask the admins to add the draft namespace [11:53:45] 13.34 < Nemo_bis> What filter are you expecting to be triggered? [11:54:10] 06:31, 5 September 2014: Magioladitis (talk | contribs | block) triggered filter 554, performing the action "edit" on 25 (Adele album). Actions taken: Warn; Filter description: top100 blog charts (details | examine) [11:54:20] Nemo_bis: filter 554 [11:54:54] https://en.wikipedia.org/wiki/Special:AbuseFilter/554 [11:55:59] maybe space 0 to 0 || Draft: ? [11:56:04] magioladitis: just ask on https://en.wikipedia.org/wiki/Wikipedia:Edit_filter/Requested to add 118 [11:56:10] ok [11:56:12] thanks [11:58:07] Nemo_bis: " [13:13:23] is there an easy way to make page categorization mandatory? [13:15:12] wmat: should be possibel with an abusefilter rule. don't know if that counts as "easy"... [13:16:19] DanielK_WMDE: ah, I hadn't thought of that, thanks, i'll check it out [13:16:41] my initial thought was to add something on save to check for a category [13:17:01] that's essentially what the abusefilter rule would do [13:18:40] Noting that you can add them via templates [13:18:48] And I'm not sure looking in page_text will help there [13:18:53] yeah, good point [13:26:49] hi [13:26:56] my mediawiki says "A database query error has occurred. This may indicate a bug in the software." [13:27:07] now, I know a possible solution is to run "php maintenance/update.php" [13:27:10] but I have no SSH access [13:27:24] can I create a php file that calls update.php, and will this be enough, or...? [13:28:26] what I mean is, if I create a web-accessible php file that executes that update.php file, will it just fix the problem or will I miss prompts or important information that I would normally see if I'd have had SSH access [13:28:27] https://www.mediawiki.org/wiki/Manual:Maintenance_scripts#No_shell_access [13:29:10] Svetlana: at the top of https://www.mediawiki.org/wiki/Extension:Maintenance it says "This extension does not support running the upgrade script!" [13:29:17] doe you know whether https://www.mediawiki.org/wiki/Extension:MaintenanceShell does support this? [13:29:19] *do [13:29:27] try the other one then [13:29:34] trial and error, ok [13:32:32] are you sure you lack ssh access? cpanel has an option to execute commands on server [13:35:17] Heya. I want to recieve the source (no html, just the source) of a mediawiki article. Is there an URL parameter that I can append so that mediawiki delivers me just the source of the artikle, no HTML, no desing elements, no buttons, ... [13:46:27] use API: https://en.wikipedia.org/w/api.php?action=query&prop=revisions&format=json&rvprop=content&rvlimit=10&titles=Main [13:47:48] https://en.wikipedia.org/w/api.php?action=query&prop=revisions&format=json&rvprop=content&rvlimit=10&titles=Main%20Page, even [13:53:00] hi, is there a way to add custom classes to headings => eG.

my funky title

[13:53:02] ? [13:56:07] sorry, I suck :-) [13:56:07] https://en.wikipedia.org/w/api.php?action=query&prop=revisions&format=json&rvprop=content&rvlimit=10&titles=Main%20Page&format=xml [14:02:17] btw what is the easiest way to get page id in mediawiki? constructing api url is tiresome (copy/paste you know) [14:02:26] (JSON is a default format, but I figured you'd want an output that looks prettier in a browser) [14:02:28] Svetlana: I only need the actual title content (no revisions), so https://en.wikipedia.org/w/api.php?action=query&prop=revisions&rvprop=content&rvlimit=1&titles=Main%20Page&format=xml is better. But is there a way to just get the article content, without the XML around? [14:02:28] what do you mean, 'title content'? [14:02:29] no. you have to pick, either json or xml, and parse that :) unless I find a woraround, 1 minute [14:02:30] nope, none [14:02:30] you have to parse the extra stuff because it may give you an error such as that tge page doesn't exist [14:02:30] and you have to be ready to that :) [14:02:30] Svetlana: Okay thanks, so I think I'll stick to parsing the XML :) [14:02:30] what language are you writing this in and roughly what for? :-) [14:02:31] Svetlana: I try to write an AutoHotkey script, recieving its configuration written down in a mediawiki article. Later on the script shall make some screenshots and upload the results to a mediawiki installation. I'm not sure if AutoHotkey is the right choice, but I don't want to write a C# program for that, that would be too much [14:02:32] question: in the User list, if it says a person is an "administrator", does this put the person in the group sysop, or would that be "bureaucrat"? [14:02:32] hm... I guess if "bureaucrat" is a group by itself (according to https://www.mediawiki.org/wiki/Manual:User_rights) then sysop must be administrator [14:02:33] it's just not clear to me when I create $wgGroupPermissions['maintenanceshell']['maintenanceshell'] = true; how I would add myself to it [14:02:33] yes, sysop = admin (and I personally like the former more, as I expect these folks to assist with technical development of the wiki) [14:02:34] hm [14:02:34] I know I could do $wgGroupPermissions['sysop']['maintenanceshell'] = true; [14:02:34] but https://www.mediawiki.org/wiki/Extension:MaintenanceShell also says it's possible to "Create a new "maintenanceshell" user group for Special:MaintenanceShell:" [14:02:34] yes, usually that. Or if you're an admin you can possibly add users to this group at the user rights page. [14:02:34] yes, I'm an admin (sysop), I guess I'll use the 'sysop' line [14:02:35] if you visit [[Special:UserRights]], you may or may not be able to assign this user right to individual accounts including yourself [14:02:35] I'm not a bureaucrat [14:02:35] I'm not sure how they set it up -- some user rights can be set by sysops and some can't -- and the line you pasted just includes that 'maintenanceshell' right into a sysop's rights set [14:02:36] sysops can alter some user rights but not all [14:02:36] maybe on Wikipedia, as an admin (= sysop, apparently), when accessing [[Special:UserRights]], I get "You do not have permission to add or remove user rights." [14:02:37] then you're not a sysop at english wikipedia I guess [14:02:37] for my own installation, I mean [14:02:37] should've used a ; after Wikipedia [14:02:37] ah I see [14:02:37] then you don't have any user rights set up which sysops can change - go ahead with the line you pasted [14:02:37] or make yourself a 'crat :-) [14:02:37] yes, I'm uploading the MaintenanceShell/ directory into extensions/ [14:02:38] will then add the require_once and permission line, and then check [[Special:MaintenanceShell]] to see if I can excute the update.php script; hopefully it'll fix the error [14:02:39] https://www.mediawiki.org/wiki/Manual:FAQ#Alter_the_database has instructions on making yourself a 'crat [14:02:40] thanks; I know how to do it, already made myself user->admin a couple of months ago [14:02:40] ran update.php, unfortunately the error is still there [14:02:40] right [14:02:41] script says "Done 0 files in 0.0 seconds", so I guess something else is wrong [14:02:41] go ahead and check your logs -- database logs and webserver logs -- I hope you have access to that :) [14:02:41] no, it's probably fine, the problem may be somewhere else and we just should get more info like that [14:06:15] gleki: some of languages have a nice library so that you don't have to type the api url. [14:06:31] gleki: does https://en.wikipedia.org/w/api.php?action=query&prop=revisions&format=json&rvprop=ids&rvlimit=1&titles=Main%20Page&format=xml look relevant? [14:07:01] gleki: http://metacpan.org/module/MediaWiki::API as one of many examples. [14:07:13] aleray: to all headings or only some? [14:07:33] Svetlana: Andy comments/ideas on my project left? [14:09:12] StefanMUC: I don't know what autohotkey and I don't know c#, but using the api to read the configuration sounds reasonable :) https://www.mediawiki.org/wiki/API:Client_code#.NET may also be removely interesting [14:11:44] Svetlana: Okay, thanks [14:13:13] Svetlana, only some [14:13:20] Svetlana: Hmm, the page tells me that XML won't be supported in future. Do you know why? [14:13:29] I found that I can use html directly. Maybe this is ok like this [14:14:00] StefanMUC, won't be supported by a .net lib, or not supported by mediawiki itself? [14:14:21] Svetlana: Won't be supported by mediawiki: We are planning to migrate to JSON-only API, removing all other formats like XML, YAML, and PHP. [14:14:40] aleray: if it shows in TOC, probably yes :) I indeed don't know how to add classes to '==' sort of markup [14:14:41] StefanMUC: news to me. I see. [14:15:14] Svetlana: Okay. It wouldn't make sense to use XML API if it isn't supported/offered in future... [14:17:13] Svetlana: Hmm, https://www.mediawiki.org/wiki/Requests_for_comment/API_roadmap#Remove_obsolete_output_formats says that XML will remain [14:17:46] I'm not sure. It's being discussed at https://www.mediawiki.org/wiki/Requests_for_comment/API_roadmap and I don't think there is clear consencus. Some parts of it say "X will happen" for people to discuss it, but it doesn't mean X will happen. (Note the "only leave JSON" heading a little below) [14:23:16] Svetlana: My browser can't find "only leave JSON" on this page, but okay, thanks for the reply. so I will look if I use JSON or XML. [14:27:40] whatever these .net libraries like, I guess - you wouldn't even notice, as it's encapsulated [14:27:53] Svetlana, thanks [14:28:00] .NET isn't so json friendly standalone [14:28:23] dunno, there's a handful libraries linked on that page [14:29:16] I'm telling you :) [14:29:28] Svetlana: Yes, but I want to use AutoHotkey ;) [14:32:10] are .net and autohotkey mutually exclusive? [14:34:48] Svetlana: .NET is a framework for programming languages, AutoHotkey is a scripting framework. That's like comparing apples and oranges. [14:35:12] scripting framework or scripting language [14:36:08] ah, AutoHotkey isn't c#? [14:37:41] life involving manually typing URLs and parsing json by hand is a hard life - I could probably use something else to sync a mediawiki page to file on disk, and read from that file using ahk [14:40:56] Svetlana: No, AutoHotkey is a software / script parser, c# is a programming language :) [14:41:13] I'm at a loss here but if there's something specific just ask [14:41:31] Svetlana: I found an AutoHotkey script capable of parsing JSON and XML, so I'll look up what's better :) [14:41:33] I dunno whether this thing has a mediawiki api lib which eases your pain or not [14:41:40] ah good [14:41:57] Svetlana: As far as I found there's no mediawiki library for AutoHotkey [14:42:00] maybe try to find a lib which avoids the need to manually type long urls into your code [14:42:22] and ask #ahk for advice if they have such channel here [14:42:49] Svetlana: I'll write a function for this, so the long url is only at one place in code ;) [14:42:57] ya [14:43:07] dont forget the unit tests [14:43:33] Hm, don't know if I will write tests... [14:44:16] I'll have to look up how I can upload pictures to mediawiki. Maybe I'll need help by some Greasemonkey scripts ;) [14:49:23] meow [14:53:04] Okay, I'll go. Thanks for the help :) [15:15:15] Is there a standard way to create/return a variable number of special pages with an extension? I just wrote something up that I think might be along the right track, but there's no need to waste time on it if there's a standard way to do it. [15:16:26] moliver: Variable number based on site config, or per request? [15:16:31] The latter sounds silly. :) [15:16:43] Maybe you could explain what you're trying to accomplish? [15:17:38] Heh. Yeah. So, I'm still working on this patch notes thing and, basically, we store our notes in individual folders for each version. So version 1 has a folder with a file in it that has all the patch notes for that version, the same version two, and so on. [15:17:45] I want a specialpage for each version. [15:18:01] moliver: So do Special:PatchNotes/version1 and so on [15:18:10] You get a $subpage param in each execute() call [15:18:19] Well damn. [15:18:25] Thanks. [15:18:28] Yup! [15:18:53] That sounds much easier than what I was doing which was trying to dynamically generate AutoloadClasses lines on the fly. [15:19:08] Yeah [15:20:22] So, I can have multiple execute lines? Or am I misunderstanding? [15:20:30] So, I can have multiple execute lines? Or am I misunderstanding?*vcalls [15:20:32] Yes, you are. :) [15:20:37] Sweet. [15:20:43] I love misunderstanding things. >.. [15:21:09] moliver: Your execute function should accept a parameter $subpage, which is a string with everything after "Special:PatchNotes/" [15:21:19] If it's null, you're at the base page, just list all possible versions [15:21:26] If it's a version name, load that version's notes [15:21:39] Oh. [15:21:40] OH! [15:21:42] Okay. [15:21:42] If it's something invalid, return a redirect to the base page or an error message or whatev [15:21:52] return 'whatev'; [15:22:24] So, basically, sniff for the directories and then do a foreach where $subpage = $ver or... [15:23:02] Sure [15:23:06] I guess I'm confused, still. I can't pass multiple values to $subpage unless I should be making $subpage an array? [15:23:08] And you could probably cache that list. [15:23:15] No no no [15:23:20] <--- Novice [15:23:28] moliver: You should never be calling execute() yourself [15:23:37] You should just expect it to be a string whenever it gets passed in. [15:24:49] Hum. [15:25:19] https://dpaste.de/2nGx [15:25:31] Right. [15:25:32] Right. [15:25:38] I'm not calling it, I'm just defining it. [15:25:50] That makes sense. >.> [15:27:06] moliver: You don't sound 100% sure. ): [15:34:28] Lol. Nah, I think I get that part. I'm still a bit confused on how execute works. So, it will call execute for each subpage or no? And if so, does the foreach need to be around the main body of the execute function i.e. https://dpaste.de/y5uz#L [15:35:03] Really nasty formatting there since I was trying to edit it in the paster. [15:36:27] moliver: execute gets called whenever someone goes to your special page. If they go to a subpage, that's a separate call to execute, and the subpage param will have a string in it. [15:37:17] Hmm. Okay. Let me mull over that. They're kicking me out for lunch now anyhow. [15:37:22] Thanks a lot. [15:38:59] :) [15:39:02] My pleasure [17:09:13] guys [17:09:19] i am having issues with wikis [17:09:25] most of the times CSS not loading [17:09:34] on wikipedia, and commons [17:17:15] ? [17:20:01] Saqib: try opening your browser's developer tools (Hit F12) and select the "console" view. Try to find what URLs are having problems (they return error status or just time out) [17:20:07] Saqib: known issue [17:20:16] i am in safari. [17:21:08] but this is really annoying me. i couldnt able to work [17:21:19] refreshing the pages 4 /5 times to view the page [17:21:26] they are working on it [17:21:28] consuming too much time [17:21:35] ok.. [17:21:51] also VFC is giving too much error [17:21:57] "Note: You cannot execute new requests without reloading this page. " [18:08:33] slow internet connection? [18:11:09] SudoKing, ? [18:11:22] @Saqib [18:13:03] Anyone know why $output->addWikiText( $pageList . '\n' . $wikitext ); is giving me CenterEdge Update Feed - advantage\n== Update #8 9/9/2014 == instead of CenterEdge Update Feed - advantage, a new line, and == Update #8 9/9/2014 ==? [18:14:02] Nevermind, got it. :D [18:25:25] Double quotes :) [18:42:13] How to remove a page (which I myself created) at Wikiversity? [18:44:04] porton, I don't think any Wikiversity allows standard users to delete pages. [18:44:05] porton: maybe they have a {{delete}} template you can put on that page, and an admin would take care of it. Be sure to provide a reason ;) [18:44:34] As far as MediaWiki is concerned, by default only admins can delete pages. [18:44:53] The right can be given out to any group by the server administrators. [20:11:49] I'm getting a PHP Fatal error: Call to protected MWHttpRequest::__construct() from a new CurlHttpRequest call… does anyone know why? [20:16:53] Hey, guys, I think I'm really confused as to how $subpage works. [20:18:04] I'm sending people to Special:PageName/subpage and my execute function has if(isset($subpage)){Do things}else{Do other things} but, they're still getting Do other things instead of Do things. [20:19:01] Rosencrantz, I think you're supposed to use the static functions in Http, or MWHttpRequest::factory [20:19:10] Ahhh… ok [20:19:27] moliver, pastebin your code please [20:19:49] I'll do you one better. I can give you a Git. [20:20:31] It's pretty messy and I'm repeating myself at least once, which is bad, but meh. I can clean it up when it's working. [20:20:31] https://github.com/olivermt91/PatchNotes [20:23:03] moliver, have you programmed in PHP before? [20:23:38] I'm not completely new to the concept, but I'm not a vet either. [20:23:50] Be gentle. [20:23:55] moliver, $subpage is the parameter given to SpecialPage::execute [20:24:15] You've called it $par instead, which is okay. But it can't be both (just from the function definition anyway) [20:24:28] Oh frick. [20:24:30] That's right. [20:24:55] I remember reading the documentation now where it says Special:PageName/$par [20:24:58] Thanks. [20:26:22] :) [20:39:40] Krenair: That was definitely it. All is in working order now. Just a little cleanup left. [20:42:39] I guess I actually do have a follow up question. [20:42:50] Is there any way to make a special page indexable? [20:43:00] I.e. searchable on the wiki?/ [20:48:27] i don't think so, since special pages aren't actually pages (and thus, have no indexable text in the db) [20:49:16] ^d: ^ can special pages be indexed? [20:49:44] <^d> They shouldn't be indexed, no. [20:50:32] I guess it's important to note that we're using CirrusSearch which parses pages and indexes the results. [20:50:39] Answer's probably still no, though. >.. [20:50:46] <^d> // Never. Ever. Index. Negative. Namespaces. [20:50:46] <^d> if ( $title->getNamespace() < 0 ) { [20:50:46] <^d> return array( null, $redirects ); [20:50:48] <^d> } [20:50:51] <^d> ^ That's in CirrusSearch :) [20:51:42] Welp... [20:51:49] That's going to be problematic. =/ [20:52:26] is there a particular reason for needing to search special pages? there's a list at Special:SpecialPages ... [20:52:49] I just finished an extension that grabs our company patch notes and puts them into special pages. >.. [20:53:22] Then when I went to search for something I knew was in one of the patch notes, I got nada. [20:53:25] =p [20:53:40] d: [21:54:20] Is anybody here experienced with Linux? I'm absoluetly freaking out - It seems I deleted my database tables! [21:55:15] I tried to change the permission in /var to have root and www-data become the owners. This, however, caused my datatbase tables to semeingly vanish - Neither Mediawiki or PhpMyAdmin can find them! [21:56:27] of everything in /var? [21:56:48] Reedy: Yes, that was in error - I meant to do var/www. [21:57:02] what os? [21:57:20] Err, how would I check that precisley? [21:57:34] uname -a might give something near [21:58:05] On ubuntu at least, everything under /var/lib/mysql should be owned by mysql:mysql [21:58:09] I believe I found it: Ubuntu 14.04.1 [21:59:51] sudo chown -R mysql:mysql /var/lib/mysql [21:59:57] sudo /etc/init.d/mysql restart [22:00:44] Reedy: Got it, thank-you so much! [22:03:25] Reedy: It's no longer working again - After sudo /etc/init.d/mysql restart the database can't be located. [22:03:50] Were there any errors from mysql? [22:03:50] It worked just before that step. [22:04:27] Red text "fail" appears under [OK] [22:05:16] that sounds like mysql is having trouble starting [22:05:23] It seems to only stop mysql and notstart it again. [22:05:52] At a guess, your error logs will have bad permissions too [22:06:07] chown -R mysq:adm /var/log/mysql.* [22:06:14] sudo chown -R mysq:adm /var/log/mysql.* [22:06:24] sudo /etc/init.t/mysql start [22:07:09] Not found. [22:07:20] typo [22:07:22] sudo /etc/init.d/mysql start [22:08:18] Failed again. [22:08:42] Oh no - var/lib/mysql is missing. [22:08:55] It was there before I tried to restat mysql. [22:08:57] missing? [22:09:01] :/ [22:09:03] *restart [22:09:08] As it, it's suddenly gone. [22:10:11] Luckily, I think I may have a bac up from the start of the hour. [22:10:17] *back up [22:12:38] Reedy: I was able to retore my server to the start of this hour, but still with the original error. I'm able to only mysql ownership of the folder itself - Not it's contents. [22:12:58] *only give (Sorry) [22:13:13] sudo chmod -R g+s /var/lib/mysql seems to do nothing. [22:15:07] "Setting directories g+s makes all new files created in said directory have their group set to the directory's group." [22:15:11] I don't think that's what you want [22:16:16] Ah, than how should the subdirectories/file be? Right now they're at 700 and owned by the errornous user. [22:17:08] I just can't help but wonder why it worked before I restarted. [22:17:35] so i looked at mine, which were mysql:mysql [22:18:07] 700 looks right too [22:18:40] I guess mysql had some file handles on it already [22:18:48] fixing the permissions allowed it to write... [22:18:53] As to why mysql wouldn't restart... [22:20:34] Reedy: What are the file permissions for your mysql folder? [22:20:52] drwx------ 20 mysql mysql 4096 Aug 29 17:57 . [22:21:21] That's what I have too. [22:21:48] It still refuses to restart, and nothing database oriented works. [22:22:51] I suggest you fix the permissions on the mysql logs [22:22:57] And then try and start mysql [22:22:59] then look at the logs [22:23:18] Though, usually it dumps the relevant errors to the console when you try and start it [22:27:40] Reedy: Are there any other folders associated with mysql? I'm finding no relevent errors - Everything within the mysql folder is owned by mysql. [22:28:10] There's /var/log/mysql [22:30:43] Reedy: Err, would you like to see the error log? There's not much I can make out of it. [22:31:03] You could pastebin it [22:32:25] https://dpaste.de/JjMz - I think the file mentioned in var/run may be relevent too? [22:32:56] var/run/mysqld is owned by root/mysql, though. [22:33:13] and the files under it? [22:33:27] my /var/run/mysqld is drwxr-xr-x 2 mysql root 80 Aug 29 17:57 . [22:34:35] Reedy: It seems that mysqld i empty. [22:34:55] -rw-rw---- 1 mysql mysql 4 Aug 29 17:57 mysqld.pid [22:34:55] srwxrwxrwx 1 mysql mysql 0 Aug 29 17:57 mysqld.sock [22:35:01] if mysql isn't running, the pid shouldn't exist... [22:35:03] * Reedy stops mysql [22:35:20] Yeah, the folder is empty then [22:36:49] I really have no clue what's wrong then. [22:37:36] InnoDB: Unable to lock ./ibdata1, error: 11 [22:37:42] That seems to suggest mysql is already running [22:38:15] ps aux | grep mysql [22:39:17] https://dpaste.de/EGiJ < There's the error I get on my wiki, if it's relevent. [22:39:41] That file doesn't exist. [22:40:18] right, because mysql isn't running [22:40:40] can you pastebin what it outputs when you do sudo /etc/init.d/mysql start ? [22:42:06] https://dpaste.de/cr97 < That happens after about a 15 second wait. [22:43:42] What does t his say? ps aux | grep mysql [22:44:25] https://dpaste.de/32wj [22:45:51] I'll note - Everything titled 'mysql' is highlighted red. [22:49:53] Reedy: Anything unusual? [22:51:49] I'm utterly stressed over this - I don't know what to do. [22:54:52] Reedy: Can I salvage the contents of my database and start over with it? [22:57:39] I'm going to be horriffic trouble if this isnt sorted. [23:02:57] Unicorn_: Can you connect to the mysql server manually? (use mysql -u root -p to log in via the command line as root) [23:03:08] Or if you don't have the root password, use the mediawiki database user [23:04:02] marktraceur: Nope: ERROR 2002 (HY000): Can't connect to local MySQL server through socket '/var/run/mysqld/mysqld.sock' (2) [23:04:26] Unicorn_: Have you tried sude /etc/init.d/mysql stop [23:04:27] sudo* [23:05:29] marktraceur: Yes, it's already stopped it seems. [23:05:39] Starting it gives me a [fail] message. [23:05:47] Except for the ps ax output that tells you mysqld is still running. [23:06:19] marktraceur: What? [23:07:05] Stopping it again seems to do nothing. :( [23:10:28] marktraceur: Is it possible to completely remove mysql from my server and somehow import the raw database data..? I'm out of options, this is direly serious. [23:12:52] Anybody? An entire wiki with 1000s of contributions - All gone because I made a mistake! That can't happen! [23:14:16] the snippet says mysqld is still running(?) [23:17:12] danhedron: Everything else implie it's not - The Wiki can't access the database, shell tells me it won't start - Something's wrong. [23:17:30] right but mysqld is still running [23:17:41] the process is still there (acording to that paste) [23:17:56] Unicorn_: Backups ;) [23:17:58] But yeah [23:18:07] danhedron: So if it's running, what's wrong? [23:18:20] You can remove and reinstall mysql, it will leave your data [23:18:35] Reedy: Really?! [23:18:58] But it may not fix your problem [23:19:07] Anything's worth the try. [23:19:51] mysql 5719 0.1 3.8 615808 78396 ? Ssl 18:00 0:02 /usr/sbin/mysqld [23:19:58] That looks like it's running without the correct config etc [23:20:18] I'm using mariadb (fork of mysql), and when running it says something like [23:20:19] mysql 20811 0.3 11.2 620020 114984 pts/0 Sl 23:35 0:00 /usr/sbin/mysqld --basedir=/usr --datadir=/var/lib/mysql --plugin-dir=/usr/lib/mysql/plugin --user=mysql --log-error=/var/log/mysql/error.log --pid-file=/var/run/mysqld/mysqld.pid --socket=/var/run/mysqld/mysqld.sock --port=3306 [23:20:30] You could try killing that process, then sudo /etc/init.d/mysql start [23:20:58] sudo kill -9 5719; sudo /etc/init.d/mysql start [23:22:37] Reedy: Words cannot express my deep graditude. I was about to have an emotional breakdown. [23:23:24] Now's probabley the time for a backup. [23:23:44] :) [23:26:01] Reedy: Suddenly, it's not working again - As if it's only capable of operating for several moments before stopping. [23:26:13] is mysql still running? [23:26:32] Yes. [23:26:55] what's the error? [23:27:15] (Cannot contact the database server: Can't connect to local MySQL server through socket '/var/run/mysqld/mysqld.sock' (2) (localhost)) [23:29:25] Results of ps aux | grep mysql: https://dpaste.de/7H0c [23:29:47] sockets can be a PITA [23:30:14] I can rightfully agree. [23:30:57] What on planet earth could possibly be wrong?! [23:32:12] Why it just breaks... [23:32:13] All I did was *accidentally* change the permission of the var folde ONLY becuase I needed to allow PHP to write/read files in the cache folder of my Mediawiki installation - And now look at it! [23:32:15] is wtf-ing [23:33:38] Uninstalling MySQL also doesn't work. [23:34:20] you might need to purge and reinstall, maybe [23:34:47] Okay, um, is it just possible to clear the entire server and re import my database data? That and the mediawiki installation are all that matters. [23:35:08] All I have is the raw data. [23:39:00] Reedy: I was able to get the database working temprarily and exproted everything via PHPMyAdmin. [23:39:48] *exported [23:40:04] It looks like I'll have to clear the server. [23:40:22] yum, phpmyadmin [23:40:32] It might be easier... [23:40:43] I'd make sure mysql is stopped, and take a tar of /var/lib/mysql too just in case [23:41:21] I've already backed all that up, but on my dektop. [23:41:53] I might as well back up the content of the netire server just to be safe too. [23:42:02] *entire