[00:21:34] hi, could there be any hope that skins like Erudite or DarkVector will be updated to the non-autodiscovery mode? Can't the mediawiki developers update these skins when the system changes? I feel like there are not many nice skins, and I'm not sure other than the 4 shipped with mw don't use autodiscovery. [00:23:42] myier: Perhaps, not familiar with those skins myself but if they're in the WMF git repositories there's a good chance I might have a look at them after I've cleared my coursework backlog if I remember [00:25:04] great! [00:25:26] Lcawte: is it complicated to migrate them? Is there a procedure written somewhere? [00:27:24] It depends on their current states... [00:29:08] If I understand right, if a skin works with 1.23 it will work with 1.24 too? [00:31:11] I'm not sure I'm afraid, I normally only play around with master copies and I don't maintain any production worthy bits of code yet. [00:38:34] ok thanks! [01:21:19] Uh I have a question how do I make a bureaucrat protection level on my site? [01:47:19] Repton: https://www.mediawiki.org/wiki/Manual:$wgRestrictionLevels [01:51:19] Thanks [01:52:54] f [01:52:54] f [01:52:54] f [01:52:54] f [01:52:55] f [01:52:55] f [01:52:55] f [01:52:56] f [01:52:56] ff [01:52:56] ff [01:52:57] ff [01:52:57] f [01:52:58] f [01:52:58] ff [01:53:23] firefox? [01:53:31] Please don't do that. [01:53:42] o_0 [01:55:03] Eurodyne: who you talking to? [01:55:26] Oxyaena [01:55:50] * Eurodyne thanks JD|cloud [01:55:56] he quit already... [01:56:01] this is to prevent his return [01:56:19] oh okey [02:26:11] Hi [02:26:18] Anyone here use Phabrictor? [02:26:20] https://en.wikisource.org/wiki/User_talk:ShakespeareFan00 [02:26:29] Comments at the end? [02:26:37] Is there a way to withdraw tasks? [02:32:31] Qcoder00: What? [02:32:39] Resolve it as invalid or whatever [02:32:48] Ah OK? [02:32:53] I wasn;t sure I could do that... [02:33:26] Qcoder00: i disagree with GOIII [02:33:46] you're perfectly within your rights to open a feature request to do something a template currently does [02:33:47] I'm not going to argue with him, because I sort of agree [02:34:02] There SHOULD have been a local disscussion on this [02:34:16] You are of course welcome to respond to GOIII in a personal capacity... [02:34:36] imo, that should be done as a Lua module actually [02:34:37] In an official capacity, the Task concerned isn't valid as it's "Too Soon" [02:34:53] So I'm withdrawing the Phab request [02:37:16] https://phabricator.wikimedia.org/T86984 [02:37:27] This SHOULD have had a local disscussion [02:43:32] Qcoder00: seriously. you just withdrew tons of good feature requests [02:43:41] you do not need community consensus to get a software change made [02:44:01] If you feel they should be re-opened on their merits start the relevant local disscussions :) [02:44:11] there shouldn't be a local discussion [02:44:20] local communities have no authority whatsoever on software changes [02:44:25] GOII has a reasonable point , given some stuff that's broken [02:44:26] (see: MediaViewer) [02:44:43] That's I think also part of the argumennt GOIII has to be fair... [02:45:09] GOIII is not putting me under pressure on this... He's raised a very valid concern... [02:45:29] If you disagree, start the local discussion and demonstarte the consensus :) [02:46:12] what consensus? [02:46:25] if a software change is useful (according to the devs), it gets done [02:46:27] if not, it doesn't [02:46:50] Whatcha talkin bout Willis? [02:47:16] T13|mobile: that you don't need to start a community discussion for every gerrit change [02:48:07] Oh, most don't need it. I thought you where talking about that withdrawn phab ticket I commented on. [02:48:21] T13|mobile: there were about 5 more withdrawn phab tickets [02:48:36] The one about table css [02:49:19] That should probably all be done through the addition of css classes locally on wiki in skin.css or common.csd [02:49:28] well, that one, yes [02:49:33] the other tickets, not so much [02:50:06] What are the other ones? [02:50:57] T87179,T73414,T86767,T87017 [02:51:22] https://phabricator.wikimedia.org/T87017 - This one was also withdrawn for technical reasons... [02:51:42] In that I am not sure there IS An easy cross browser soloution for it... [02:52:02] look at the 50 changes at https://gerrit.wikimedia.org/r/#/q/owner:%22Jackmcbarn+%253Cjackmcbarn%2540gmail.com%253E%22+status:merged,n,z [02:52:11] 3 of them had community discussion. 47 didn't [02:53:29] As I said if you feel pasionatley about this, I have no objections to them being re-opened ONCE the relevant disscussions have had the semblence off having occured.... [02:53:37] T87017 doesn't need discussion [02:53:52] I'd have not withdrawn it. [02:54:12] T13|Mobile : There is nothing to stop it being re-opned... [02:54:24] Provided theres a good rationale :) [02:54:37] but that would be your call not mine [02:55:27] If T86767 is a known issue and has another ticket, merge them [02:55:38] Otherwise leave it open. [02:57:17] I can't merge , because I don't know if it's a Known issue that's reached Phab [02:57:40] T87179 is unclear. I'm not sure what you're asking for. Might want to clarify even if you don't reopen. No discussion is needed on any of these as I see it yet. [02:57:56] It was GOIII that said it was a known issue on either his or ShakespeareFan00's talk page [02:58:13] I am not personally aware of a bug that could be merged [02:58:15] If it's not a known issue on phab, then it's not a known issue. [03:00:32] T73414 also doesn't need discussion, although it does need clarification. [03:01:29] Feel free to comment [03:01:33] Or re-open [03:02:00] But I agree with GOIII some of these probably need to be taken up on the relevant project wikis first... [03:02:13] If you disagree then I suggest starting those disscussions :) [03:04:32] Let's have the discussion on the relevant wiki instead of here on IRC given there's a backlog of other things, and the requests where somewhat project specifc [03:05:58] jackmcbarn: Of course if you were voluntering to look at some of these, I have objections to feedback on (temporarily?) closed tasks [03:12:11] BTW I've just noted something [03:12:24] I've you edit a whole page you get a template list [03:12:35] You don't seemingly get this on a section edit [03:12:51] I'll ask about that .... [06:26:05] ify qwerty1 [07:07:43] morning sunshines [07:08:00] somebody used to the ldap authentification with mediawiki? [07:16:47] bnntd: https://wikitech.wikimedia.org/ uses it, i believe. [07:18:25] I'm trying to configure it at the moment.. Struggling now already for hours -.- [08:04:27] has somebody an idea what could be meant with "Binding configuration"? [mediawiki:extension LDAP Authentication] [08:04:55] I don't get these config steps [08:48:01] hello world [08:48:22] I'm having dificulties in finding the syntax for cyting [08:48:25] can you please help? [11:19:42] hi, it is possible to protect a page that only bots & sysops can edit it (with titleblacklist?) [11:22:50] * Steinsplitter waves to legoktm [13:26:41] Hi [13:27:03] It is the responsibility of the community to create an alternative project without having centralized information on a server [13:27:18] automatic load balancing, better locality, no storage limits and we Dont need Donations!!! [13:27:38] Several projects are already launched, but there is no big solution which would permit users to use and to contribute to a peer to peer webhosting [14:57:22] jackmcbarn [14:57:25] You around? [14:57:48] Alternatively is there someone here that understands Lua code ? [15:02:13] Qcoder00: don't ask to ask, just ask! [15:04:28] myself: https://en.wikisource.org/wiki/Template:Hymn/verses I think this would work better as LUA code [15:04:46] There is also the monster called ollist which is mess [15:05:43] *sets up popcorn stand for nice folks* :) [15:06:13] varnent: You can call it "A kernel of decency" [15:06:17] The backgorund to https://en.wikisource.org/wiki/Template:Ollist [15:06:26] marktraceur: I like it! [15:06:35] is that currently mediawiki lists are very simple [15:06:37] marktraceur: lol - a whole new feature for Hackathons :) [15:06:49] you use a # for numbered lists [15:07:00] This works well on Wikipedia... [15:07:15] However, this hits a snnag on Wikisource where you have paged content [15:08:14] You can still use # but it will start from 1 on each new Page: which is more cumbersome for proofreading [15:08:30] In addition (unless you know otherwise) all lists using # start at 1 [15:08:43] Whicg is why

You cant have a list starting at a different number... [15:08:57] bawollf: I was coming to that [15:09:17] Technically using start on ol is not now recomended practice [15:09:39] Is that html5 ruining everything? [15:09:52] And I feel that there should as far as possible to be a view to the elimination of inline HTML in wikitext [15:10:04] * bawolff just joined channel. May be missing context [15:10:06] CSS3 would let eveything be a DIV [15:10:25] https://en.wikisource.org/wiki/Template:Hymn/verses I think this would work better as LUA code [15:10:35] There is also the monster called ollist which is mess [15:10:37] bawollf: hi, protecting (edit=only bot & sysop) is with titleblackist not possible? [15:10:53] olliist is a bodge... [15:11:00] Which shouldn't need to exist... [15:11:18] but then everything would be css. We could simplify that by coming up with an sgml-like syntax to express such stylistic info! ;) [15:11:45] ROFL [15:12:06] There is also the issue of inlined CSS across Mediawiki [15:12:08] Steinsplitter: hmm let me check. Off the top of my head i think you can do sysop but not bot [15:12:48] sysop and autocinfimed for what i can see and tested (on test.wikipedia) [15:13:17] I've had people comment that HTML5 and CSS3 allow for the display: attribute to be tweaked... [15:13:43] Thusly the Oldstyle HTML TABLE syntax is now archaic [15:14:44] https://en.wikisource.org/wiki/User_talk:ShakespeareFan00#Task_86984 [15:16:04] Steinsplitter: looks like you cant do it [15:16:19] bawollf: oh. how? [15:16:36] http://www.w3schools.com/cssref/playit.asp?filename=playcss_display&preval=list-item also [15:17:13] Steinsplitter: no, i mean you cannot do it [15:17:18] Thusly I think it times the list handling was using DIV based syntax and display: in CSS styling rather than older syntax :) [15:17:32] ah :( thanks aniway [15:17:33] This would make it easier to do what ollist was trying to do [15:18:16] The useful thing about DIV's is that you can put other things in them that you can't necessarily do with a LI [15:19:07] Id point out that probably goes against using semantic html which is current fad/best practise [15:19:31] Hmm [15:19:43] CSS3 also has counters [15:20:01] Html tables arent meant to be superceeded by css for tabular data [15:20:21] unless you mean purely presentational tables [15:20:34] I think that's what G03 is getting at [15:20:41] which have been considered bad since about css1 [15:21:10] He also had a rant about TABLE being used as LAYOUT element rather than a content structure element... [15:22:38] if # and * style mediawiki lists where

and there was a css class element defining display:list-item for that class it could make some things easier... [15:23:39] maybe class=mwlistemcounter for # (so a CSS3 counter could be used maybe? and melistitem for * lists [15:24:10] This moves the problem from the HTML side of things to being a style issue which may be easier to resolve [15:24:25] The same goes for layout presentational table vs actuall data [15:24:32] Im kind of confused what the actual problem you are facing [15:24:36] is [15:25:20] The actual issue is that ollist is a monster of a template that should not need to exist if wiki-syntax had some tweaks [15:25:36] * and # represent lists. They should be

Yes but currently you can't stat a * or # list midway [15:26:04] You have to start at 1 [15:26:15] Hi all. Recently installed wikimedia to my server and I have been having some troubles with user accounts and such. I have made no edits to anything yet but still unable to do much. I have the main account which i cant log into.... I just get [15:26:15] "Login error [15:26:16] There seems to be a problem with your login session; this action has been cancelled as a precaution against session hijacking. Go back to the previous page, reload that page and then try again." [15:26:16] Then I made another account to test things. It autoconfirmed that user, but I cant change any settings for the user because when i try, it says i have to log in again and all i get when i try that is [15:26:16] "Login error [15:26:16] Project Tamriel Wiki uses cookies to log in users. You have cookies disabled. Please enable them and try again." [15:26:57] Currently you also can't split * or # based list items over a page break [15:27:08] (which granted is a Wikisource specfic issue) [15:27:14] * bawolff just looked at ollist. No idea what it even does [15:27:51] * and # for the common case, and

What it's supposed to do is allow the construction of

If it's purpose isn't clear it's a bad template [15:28:46] I just havent read wikitemplates in a while [15:28:59] literally worst macro language ever [15:29:28] Why do you think the really complex templates migrate to Lua [15:29:31] ? XD [15:30:01] But other than wikitext templates sucking, whats the issue here [15:30:31] ollist works fine for its purpose i assume [15:30:32] # and * lists need extensions to work better with Wikisource [15:30:59] bawollf : Well , it was deprecated because I could never get to work OVER a pagination consistently [15:31:14] As you know Wikisource has Page: s which get transcluded [15:32:03] # and * can be set up to work on a single Page.... but they can't at present be configured to work consistwntly when the list concerned is over multiple Page: s [15:32:42] meaning that

end up being used in situtations where with some tweaks they wouldn't need to be [15:34:15] https://en.wikisource.org/w/index.php?title=Template:Ollist&action=history If you check the early history of the template, there was a serious attempt to come up with a way of getting multi-page lists working which did not succed [15:35:30] This makes it harder to proofread, which actively discourages people from contributing to that project because it's a hassle [15:35:40] (and a hassle which shouldn't need to exist) [15:36:05] I just checked. Its possible to have a * or # list that's continued across multiple transclusions [15:36:17] Wheres the hints? [15:36:24] because when I've tried it... [15:36:42] It didn't behave consistently in both the trancluded version and the Page: [15:36:58] provided there are basically no newlines between the transclusions [15:37:12] Hmm [15:37:28] That solves one the issue of split list-ITEM issue possibly [15:38:07] How do you suggest solving the match item numbering to that on the Page: when editing the Page: issue? [15:38:29] when I've tried to do this continuation instead of say a list numbered 5,6,7 I get 1,2,3 [15:38:50] in the Page: version, but obviously 5,6,7 when it's transcluded [15:39:06] OK, when i first try and login it tells me one thing. Then I try log in again it tells me another, then i try log in a third time and it tells me im already logged in [15:39:27] (There's also a long standing issue about where exactly LST puts linefeeds...) [15:40:03] SGMonkey: check php session storage settings. Maybe [15:40:09] At one point it says im logged in. I click to edit a page and it says in the top right you need to log in [15:40:25] bawolff: How do i do that? [15:40:50] Just a moment [15:40:53] bawollf: In that in some sitaution when doing LST of Page: s you have to put in {{nop}} and in others take them out :( [15:40:56] Anyway [15:41:03] we were talking about LIsts [15:42:00] There is also (an aside) The issue of documents where you have numbered paragraphs... These technicaly are NOT a list... [15:43:34] !session | SGMonkey [15:43:34] SGMonkey: make sure session.save_path is set correctly in your php.ini. Verify that it exists, is readable and writable to PHP, and check that session data is actually being written there. On a server farm, make sure it's a shared directory, accessible from all servers. [15:44:00] SGMonkey: Although to be honest, that doesn't exactly sound like what you're describing, but its the closest I can think of [15:44:33] your php.ini file can usually be found in /etc/php5 or something like that [15:47:35] Qcoder00: e.g. https://en.wikisource.org/wiki/User:Bawolff/sandbox [15:48:16] https://en.wikisource.org/wiki/User:Bawolff/sandbox/2 Yep [15:48:27] Those should ideally be displayed as 4 5 6 [15:48:33] etc.. [15:48:49] Ah you want the numbering to be consistent on the non translcuded version too [15:48:53] YES [15:49:23] ok, i see why you're running into problems [15:49:40] Which would be something like a #@6 in the header or at the start of the list telling it where to start :) [15:50:14] Unless of course there numbered list items which WOULD start with a @ [15:50:49] Actually you'd need 3 new #tags [15:50:59] Maybe you'd be better off just typing in 1. 2. [15:51:13] True [15:51:19] As you basically just don't want auto-numbering behaviour [15:51:36] but that means you loose the list styling... [15:52:41] As I said it's 3 new tags #@ to mark where a #list starts.. #< marks the start of a split list item... #> marks the end of a split list item... [15:52:48] The best I can think of is something like, have templates ollist-start, ollist-item, and ollist-end [15:53:17] Well that would in effect be templatizing the HTML [15:53:28] Which I've explictly been told is a BAD thing [15:53:43] And then do {{ollist-start|7}}{{ollist-item|first}}{{ollist-item|second}}{{ollist-item|third}}{{ollist-end}} [15:53:55] This is all templatizing html [15:54:09] to some degree or another [15:54:13] Which is more typing then the same syntax using lI [15:54:35] its all templates, and eventually it all becomes html :) [15:55:15] I also don't like using templates when it's something that shhould reasoably be fixed in the core parser [15:56:03] Honestly, you haven't made that convincing a case for this feature [15:56:04] There's also the issue of where proofreadPage thinks there should be linefeeds... [15:56:20] Its very specific to the wikisource use case [15:56:39] bawolff: OK, well I'll just have to stop recomending MediaWiki to people on the grounds it's broken [15:56:40] And I can't really think of an elegant syntax that would be intuitive [15:56:56] and likley to remain so :( [15:57:43] Well nobody other then wikisource would need it, and wikisource isn't moving any time soon, so... :P [16:13:35] bawolff, shouldnt these all be setup properly to begin with? [16:13:42] It was a fresh install [16:14:06] SGMonkey: Well if we're going with "supposed to", MediaWiki is supposed to just work :) [16:14:40] Should i try reinstalling? Maybe it was a corrupt download or some shizz [16:15:34] SGMonkey: It sounds more likely to be a problem with some other layer than the mediawiki layer (I think, hard to say for sure from just what you've said) [16:16:19] OK, so when i try and log in it just says [16:16:19] "Login error [16:16:19] Project Tamriel Wiki uses cookies to log in users. You have cookies disabled. Please enable them and try again. " [16:17:10] And it does this on every browser, PC and location. I installed it when i was away in a different part of the country, same diff [16:24:14] So what is session.save_path in your php.ini config? [16:26:04] <^d> I hate that error. [16:26:10] <^d> Especially since I've never disabled cookies. [16:26:56] ^d: It beckons back to the good ole' days when all you had to do to be a privacy nut was disable cookies :) [16:54:14] Steinsplitter: hi, you can use the AbuseFilter for that. [17:12:45] OK, installing it from scratch (Nothing was on the wiki anyway, plus got stable internet here) Its telling me that "MediaWiki 1.24 requires at least PHP version 5.3.2, you are using PHP 5.2.6." Even though in my panel it is set to 5.4. Can i just use a .htaccess to set it correctly there? [17:13:26] SGMonkey: that depends on how php and apache are configured. ask your hoster [17:13:51] (the answer *should* be "yes", but who knows...) [17:14:17] haha, fair enough [17:14:23] I'll give it a whirl [17:14:54] Yeah that worked [17:14:56] AddType x-httpd-php55 .php [17:15:14] Should i have any other info in my htaccess? [17:16:32] If you want to set up short urls, you'd add that to .htaccess, but if you don't need that, no other info is neccessary [17:16:39] I dont know why people still suggest using filezilla [17:16:43] Short urls would be nice [17:17:30] The last ftp client I used was the command line one.. [17:17:35] !shorturl [17:17:35] To create simple URLs (such as the /wiki/PAGENAME style URLs on Wikimedia sites), follow the instructions at or try the new beta tool at . There are instructions for most different webserver setups. If you have problems getting the rewrite rules to work, see !rewriteproblem [17:18:15] short urls can sometimes be a pain, I'd reccomend making sure your wiki is fully set up and working first, and then go back and do short urls second, once you know that the main part of the wiki works [17:18:18] Was just on that page! [17:18:37] Yeah i was thinking that bawolff. I wasnt gonna go messing with anything until it works atleast [17:22:41] Warning: Could not find APC, XCache or WinCache. [17:22:41] Object caching is not enabled. [17:22:51] Git version control software not found. [17:23:30] http://pastebin.com/wrqVKBtp [17:23:43] Do i need to be concerned about these things? [17:25:34] Lack of object caching will make things slower, particularly if lots of people are using the wiki [17:25:41] Ignore the git version control software thing [17:26:10] I personally think that the git version control warning should be removed, it mostly serves to confuse people [17:26:22] I thought as much [17:26:26] However, object caching can always be added later if you need it [17:26:34] Any ideas how i can add it? [17:27:22] If you're on a shared host, you'd have to contact your host [17:27:40] On linux you normally have to add a package named php5-apc or something [17:28:13] "On linux you normally have to add a package named php5-apc or something" to the htaccess? [17:28:19] no [17:28:25] Im pretty sure im on linux hosting [17:28:32] In fact i know i am [17:30:24] SGMonkey: its a separate program you need to install (like with apt-get). If you're on shared host (e.g. Don't have root access) you probably can't do it [17:31:29] Root access? [17:32:01] afaik i have root access but how do i know [17:32:46] SGMonkey: Is your hosting a "VPS"? [17:33:05] No idea. Im using TSO [17:35:26] bawolff: It is linux platform [17:35:38] probably [17:36:08] hmm [17:36:23] How to actually install apc might differ based on how php is set up. But probably the instructions at https://www.digitalocean.com/community/tutorials/how-to-install-alternative-php-cache-apc-on-a-cloud-server-running-ubuntu-12-04 would work [17:36:32] (assuming either debian or ubuntu [17:36:44] wasn't apc dropped in some version? [17:38:21] Would those commands be entered via Shell Access? [17:38:38] * bawolff doesn't know. Possibly [17:39:11] Who are TSO? [17:39:13] if your php is up to date you should have OPcache installed anyway with it, which means you don't need APC [17:39:54] which is more important than the object cache side of it [17:41:04] afaik php is up to date. It says its using version 5.5 [17:41:05] i see now that SGMonkey said his php is version 5.4 though, so won't have OPcache [17:41:15] Oh, okay! [17:41:36] MySQL (or compatible) [17:41:36] PostgreSQL [17:41:36] SQLite [17:41:37] Which of these are best? [17:41:44] MySql [17:42:39] k thought so thanks [17:43:53] Database character set: [17:43:53] Binary [17:43:53] UTF-8 [17:43:59] Which is best? [17:44:10] UTF-8, unless you need chinese [17:45:03] Pretty sure i dont need chinese. Afaik we don't have any chinese members. Mainly english and german [17:45:25] It's not really Chinese members, it would be Chinese content [17:45:51] yeah i figured. But we arent gonna have chinese content if we dont have chinese members [17:46:30] Hey, can someone tell me why i cant see any text here http://wiki.supaham.com/XChat? Even though its there when i click edit [17:47:21] There looks to be a lot of text missing from your gui [17:47:25] I recently moved my database and mediawiki to a different server, but as far as i can tell, it isnt an issue of connecting to the database [17:47:30] my gui? [17:48:10] If it was a db connection error, it'd be blowing up [17:48:17] "XChat is a 1337 plugin for cross server communication, backend and front end." <--- lol [17:48:33] I'd expect http://wiki.supaham.com/XChat?action=purge (as anon) to have text about purging [17:48:35] ) [17:48:58] bawolff, where do you see that? [17:49:04] Reedy, not sure what you mean [17:49:08] on view source [17:49:13] I just find the phrasing funny [17:49:35] ah, i wanted to create the page quick, i haven't finished the actual project yet :P [17:49:58] http://wiki.supaham.com/Special:Version is also borked [17:50:01] I copied the mediawiki directory exactly as it was on the old server [17:50:08] looks pretty broken. as far as I can tell, the parser bails out [17:50:13] !debug [17:50:13] For information on debugging (including viewing errors), see http://www.mediawiki.org/wiki/Manual:How_to_debug . A list of related configuration variables is at https://www.mediawiki.org/wiki/Manual:Configuration_settings#Debug.2Flogging [17:50:25] Reedy: thanks ready, i was just about to do that :) [17:51:28] mmm, my error.log says this: http://pastebin.com/rXXYw66g [17:51:51] Screwed localisation cache? [17:52:33] but the pp report still shows up on http://wiki.supaham.com/api.php?action=parse&text={{PAGENAME}} oddly enough [17:52:40] What is the function of the extention "cite"? [17:52:47] Extension* [17:52:47] Citations [17:52:53] ahh doh! [17:54:11] Is there any harm in using all the bundled extensions? [17:54:16] although his magic words seem to still be there - http://wiki.supaham.com/api.php?action=query&meta=siteinfo&siprop=general|namespaces|namespacealiases|statistics|magicwords [17:54:17] i should probably update mediawiki to latest before complaining [17:54:21] SGMonkey: no harm [17:54:49] SupaHam: you're using 1.22, that's not that old [17:55:21] Settings for object caching: [17:55:21] No caching (no functionality is removed, but speed may be impacted on larger wiki sites) [17:55:22] Use Memcached (requires additional setup and configuration) [17:55:22] What about these? [17:55:25] >group name. Is that referring to linux group name as in the owner of the foldeR? [17:55:38] SupaHam: Where do you see it? [17:55:46] SGMonkey: you don't have memcached, so select no [17:55:47] i enabled debugging, http://wiki.supaham.com/Main_Page [17:55:58] bawolff: makes sense XD [17:56:06] SupaHam: Oh, i bet its the PCRE issue [17:56:31] Perl Compatible Regular Expressions? [17:56:40] there was something about pcre on fedora being compiled without unicode I think (?) [17:57:16] oh god, i hope i didnt break something in the system when loading up my backup D: [17:57:38] also im on ubuntu [17:57:51] I have to upload the local settings.php right? [17:57:59] Im sure i do but i just wanna check :P [17:58:09] SGMonkey: yes [17:58:23] haha, k, sorry, just double checking [17:59:29] Awesome. The wiki is faster than it was before. And it doesnt complain about cookies and other bullshit [17:59:38] Cheers for all the help bawolff [18:01:34] anyone know if i can use the PersonalUrls hook to display a picture as the link instead of text with a picture next? [18:03:01] SupaHam: try doing something to force a cache flush of the localization cache [18:03:23] e.g. Disable all your extensions and re-enable them [18:03:23] im not sure how to do that [18:03:28] ah okay [18:04:38] bawolff, if by that you mean comment out the require_once calls in my LocalSettings.php, that didn't fix it [18:10:46] SupaHam: found the issue. This is appearently https://phabricator.wikimedia.org/T60640 [18:11:30] So either need to use a different version of PCRE, or upgrade mediawiki [18:12:05] how do i find my current version of mediawiki? i just upgraded and want to confirm [18:12:29] Its 1.22.0 [18:13:07] SupaHam: Normally you'd look at the page Special:Version, but since that's not working for you, go to view source and look for the line [18:22:59] is it bad to untar a new version into my old mediawiki dir? [18:25:57] nvm [18:28:11] What to do if apparently the github clone has not been updated, even though a commit has been merged in gerrit? [18:28:33] UltrasonicNXT: github is just a mirror, there's no problem [18:28:37] SupaHam: just for the record, no, it's perfectly okay (unless you modified any of the existing files you're going to overwrite) [18:28:42] It'll sync eventually [18:29:00] yeah, give it an hour, then file a bug [18:29:06] ok sure [18:30:14] I'm getting an error in UploadWizard, text "NS_ERROR_FACTORY_NOT_REGISTERED: Factory not registered" [18:30:32] Web searches bring up vagrant errors, I tried restarting the VM but not fixed still [18:30:53] Trying to update, we'll see if it helps [18:31:29] UltrasonicNXT: I'm fairly sure the Foundation offers HTTP/HTTPS cloning without logging in, not really sure why Brickimedia is using the gH clones instead... [18:32:36] github is probably more reliable :/ [18:33:04] i frequently get timeouts when pulling from / pushing to gerrit from europe [18:38:14] Damn it. Why must there be Vagrant problems while bd808 is on a plane [18:49:47] Could we ever have an SQL wrapper in MW so extensions add their tables and stuff through PHP, rather than a raw SQL file, like how we don't do manual inserts/updates/deletes etc? Just something that came across my mind the other day [18:50:14] (then the wrapper converts them into different SQL types etc.) [18:55:40] UltrasonicNXT: there's an RFC or two for that [18:56:12] Glaisher: Can you add my bot account and my account to the simple.wikipedia AWB/CheckPage please? I'm on the CheckPages on enwp and commons as well (just added to commons couple days ago). Technical_13 and T13bot [18:56:54] I've got 235 script talk pages that I need to post a message to about deprecated elements. [19:00:46] MatmaRex: ah cool [19:51:22] !pastebin [19:51:22] To avoid overflowing the channel with inane amounts of text, use https://phabricator.wikimedia.org/paste/create/ or other awesome pastebin sites to share code, errors, and other large texts. [19:52:23] marktraceur: https://phabricator.wikimedia.org/T87248 [19:52:50] Why am I involved [19:57:39] marktraceur: because when I asked about it, you replied :P [19:57:49] is there a good training video for media wiki? [19:57:54] formatting etc. [19:58:44] UltrasonicNXT: I'm looking at some logs now [19:58:55] thaaankyou :) [19:59:06] (I literally have no clue how to approach the problem) [21:20:15] hi [21:20:38] I have questions about your services [21:20:45] ask away [21:20:47] !ask | Dat_ [21:20:47] Dat_: Please feel free to ask your question: if anybody who knows the answer is around, they will surely reply. Don't ask for help or for attention before actually asking your question, that's just a waste of time – both yours and everybody else's. :) [21:22:01] Well i would like to get 100 slots with the promotional offer but in the event that ill need more half way through the year of the service is there a reasonable way to increase the server slot size while keeping the promotional value [21:22:19] Dat_: You must be confused, we don't offer servers. [21:22:37] Dat_: This is the channel for a free wiki web application called MediaWiki. [21:22:50] omg [21:23:06] omg indeed. [21:23:08] i was looking for contact support for the Ts3 services [21:23:20] Ts3? [21:23:21] Teamspeak to MediaWiki [21:23:30] Not sure how you wound up here :) [21:26:44] Does anyone have any experience with setting up geographically redundant mediawiki instances (possibly via anycast)? [21:27:48] I ran across Manual:Cloudflare, but that seems pretty specific to Cloudflare, not setting up your own [21:31:58] Rosencrantz, that page sounds sensible in general. the details depend on your caching setup, really [21:33:53] srsly, caches that don't support htcp suck! [21:36:27] Ok, I'm not sure how I follow on where the caching bit comes in. Would all of the instances have to use the same cache or something? [21:36:56] depeeends on your setuuup ;) [21:37:41] Well right now I'm using memcached for caching, and that's it. Not sure squid would do any good since if I read that right, it only caches anonymous (and we don't allow any anonymous) [21:38:15] it still helps you with static resources [21:38:35] (unless those are under http auth or something) [21:39:10] why do you need a geo cdn for a private wiki, then? [21:40:19] So if something goes down, the wiki is still up. [21:40:54] mmm, that doesn't sound very combinable with logged in only access [21:43:10] i'd rather concentrate on making the wiki resilient [21:51:52] I don't follow on the combinable with logged-in on bit… and isn't that a part of making it resilient? Just in a different way than say just a local cluster? [21:54:19] *that refers to the global load balancing, not the logged-in thing [22:06:48] Rosencrantz, i mean that cdn will not help you since it will just pass all the requests for actual p[age content to appservers [22:07:23] thus, it won't be adding much reliability in your case [22:08:13] Ahhh ok [22:08:14] thanks! [22:08:16] (I've got to go)