[00:53:22] why am I getting ERR_TOO_MANY_REDIRECTS trying to enable DPL on a newish wiki [00:55:09] enabling tries to perpetually load Template:Extension_DPL [01:54:04] Hi c. [01:54:19] It sounds like you're hitting a redirect loop. [03:45:08] need help here ... https://www.mediawiki.org/w/index.php?title=Help:History&curid=189572&action=history [10:15:28] Hi, can someone recommend an easy to use bot for my own wiki. I want to replace some text in the articles. [10:20:52] ralphlaar: would https://www.mediawiki.org/wiki/Extension:Replace_Text work? [10:43:18] samwilson it seems to work for few pages, but not for a lager number of pages. It's a 1.31 wiki. Maybe it's to old to work with Replace Text [10:44:08] it simply doesn't execute the job, only for few pages. So out of 5000 page maybe 30. Don't know why [10:44:15] I think it works with 1.31; have you got the right version? And it should be able to do hundreds of pages at least, although I think it can do any number (in batches) [10:44:20] command line also tried [10:44:39] have you run `php mainentance/runJobs.php`? [10:44:48] maybe your job queue isn't being processed quickly [10:45:02] no didn't do that [10:45:22] when should I do that samwilson? [10:45:48] First go to special page and start the replacing there and then run the script? [10:46:14] or run the script from command line replaceAll.php and then the script [10:49:44] [pwd] i have weird problem with logging. I use "pwd login" and it shows I'm logged in with my user name, but when I try to run script it shows I'm logged as "anon" and the operation faild due to lack of edit rights. I want to use pagefromfile and I have bot configured. Any ideas how to fix it? [10:51:22] getuserinfo shows: {'id': 0, 'name': '10.15.60.1', 'anon': '', 'groups': ['*'], 'rights': ['read', 'createpage', 'createtalk', 'writeapi', 'viewmywatchlist', 'editmywatchlist', 'viewmyprivateinfo', 'editmyprivateinfo', 'editmyoptions']} [11:02:46] pefi: how do you log in and how do you run the script? [11:03:20] log in with python3 pwb.py login [11:03:36] script run with: ython3 pwb.py pagefromfile -showdiff -file:redirect.txt -notitle [11:04:32] there is error in pywikibot/site.py but I'm not sure where to report it (missing bracket in self.user call) [11:06:20] The script shows: User "" does not have required user right "edit" [11:06:25] instead of [11:06:36] User "None" does not have required user right "edit" [11:07:40] the missing bracket is in decorator function, in raise UserRightsError line [11:12:25] yep, exactly [11:16:35] Ok, I think my question boild down to how to log in before using pagefromfile from command line. i did a test and logging form python scipt works fine [11:35:32] The pywikibot.lwp cookie has expiry time = current time. Is that OK? [11:37:06] samwilson thank you. It works now [13:07:59] Hi all - I'm trying to create a template, that when called upon, 'echos' the year + 400 years. So 2020, becomes 2420. (See: https://dpaste.org/bht4 for code). It's worked before, but for some reason doesn't work anymore. [13:09:46] The switch seems a bit pointless if you're always adding 400 ;P [13:10:56] NixonInExile: WFM on my dev wiki [13:11:41] It's not my original code tbh, I can scrape some stuff together and pick up a basic idea of whats happening. [13:12:05] Would you recommend a how-to guide? [13:12:15] What's it actually showing/outputting? [13:12:54] it just echo's the code [13:13:03] I can PM you links. [13:13:36] Would it work just as well, if it was {{#expr: {{CURRENTYEAR}} + {{400}}}} ? [13:14:29] You shouldn't need {{}} around the 400, ti'll try and use a template [13:15:04] "{{#expr: {{CURRENTYEAR}} + 400}}" works fine (without the "") [13:17:09] Urm, that didn't work. I'm guessing expr = express? [13:17:31] expression [13:18:00] https://imgur.com/a/sDPDNkO [13:19:19] https://imgur.com/a/wFBRrTz is what i'm getting. [13:19:35] Is ParserFunctions installed on your wiki? [13:19:38] Makes me think it's an issue with my installation [13:19:41] Can check via Special:Version [13:21:22] ahhh ha [13:23:35] Okay, I'll download this and upload it to the server! Cheers Reedy for your help :) [13:44:34] WHOOOHOO That worked. [13:44:43] Oh, net split. [14:05:03] I forget, what's the parameter name for when I'm doing an API query and I want to index the pages/users/etc? [14:05:21] index? [14:05:37] I tried that, it said unknown [14:06:11] "warnings": { [14:06:11] "main": { [14:06:12] "*": "Unrecognized parameter: index." [14:06:12] } [14:06:12] }, [14:06:13] "batchcomplete": "", [14:06:14] "query": { [14:06:16] "users": [...] } [14:06:34] I want a separate index of userids from users. [14:07:44] indexpageids ? [14:07:47] I dunno if that works for users [14:08:36] That's the one I was thinking of. [14:09:19] It does not seem to work. Oh well... I'll do it the hard way. Thanks Reedy [14:09:58] :( [14:40:44] Am I correct in reading https://www.mediawiki.org/wiki/Manual:Varnish_caching where it says "A more complex caching strategy may use multiple web servers behind the same Varnish caches (all of which can be made to appear to be a single host)..." to mean that multiple Varnish instances can be made to appear as a single server? If so, I'm not seeing [14:40:44] in the Varnish docs how to do that. [14:41:37] justinl: So that is what Wikimedia does, so yes [14:41:43] The problem I'm trying to solve is with $wgSquidServers, rather than having to provide a list of Varnish server IP addresses, I'd like to specify one unchanging host/cluster name for the MW HTTP PURGE requests. [14:41:45] I think it uses LVS maybe [14:42:01] Yeah, that. It's not a "varnish" feature per se [14:42:18] Ok, so I'm not crazy, it's not built in to Varnish. /whew [14:42:19] justinl: So what WMF does for that, is use HTCP over multicast [14:42:38] So MW sends a udp packet to a multicast address [14:42:56] HTCP is new to me so that's some research I'll have to dive into. [14:43:21] Cool, thanks! This has been a thorn in my side for years! [14:43:28] And then vhtcpd converts that into a normal purge on each host (We use htcp largely because squid used to support that natively, but varnish doesn't) [14:43:32] justinl: https://wikitech.wikimedia.org/wiki/Multicast_HTCP_purging [14:45:50] Out of curiosity, given your squid vs varnish comment, is there any reason to go with something other than HTCP if you're not already using it but are using Varnish, or is it still the best approach regardless of Varnish's lack of native support for it? [14:46:32] I suspect in part it's what MW has functionality for.. [14:47:49] yeah, we use it because we already implemented it [14:48:13] If you wanted to redo from scratch, something like zeroMQ which has reliable multicast, would be better, strictly speaking [14:48:44] or maybe kafka, i don't know how all those event things actually work, or which one is appropriate [14:49:09] But multicast UDP has the issue that it is an unreliable protocol, so if network gets overloaded or whatever, purges just silently get missed [14:49:28] Heh, I had my first exposure to Kafka last year at a job I was at part of the year. Pretty cool, but pretty heavy for this case, I think. Plus ZMQ is already installed everywhere thanks to running SaltStack on it. [14:49:53] The Zookeeper stuff was kinda annoying though. [14:50:20] So if I can leverage ZMQ, I may go for that approach, but it's nice having options. [14:54:21] I'd love to one day work on cache purging via kafka and stream processing for WMF. [14:54:34] def way too heavy for lower traffic wikis [14:55:11] but for WMF where many edits to the same page can happen quickly, it would be nice to purge (and reparse) less often if edits are happening quickly [14:55:30] this needs some buffering and windowing and de-duping, which stream processing systems can do well [15:13:40] Wanted to follow up on a previous question about running Semantic MediaWiki and after my recent upgrades, trying to run the rebuildData script on two of our largest wikis that keep hitting "Error: 1213 Deadlock found when trying to get lock; try restarting transaction" errors, forcing me to restart the script, though it picks up where it left off. [15:13:41] However, there is so much data that the biggest wiki literally takes a week or so for the rebuild to complete, even without the deadlock errors. [15:14:09] Anyone know what can cause this and how to look into possibly correcting whatever database/data issues are causing these errors? [15:15:46] FWIW the function its usually running when it dies is Function: SMW\SQLStore\PropertyTableIdReferenceDisposer::cleanUpReferencesById [18:00:35] Hey all, I still have an issue with my upgraded wikis regarding searching of namespaces. For example, we can't search on the Template namespace, results during typeahead don't show anything under there. The URLs for the opensearch api calls in the browser tools have namespace=0 on our non-SemanticMW wiki and namespace=0|102|108 in the SMW wikis [18:00:36] (102 and 108 being namespaces added by SMW). We thought it might be the Popups extension, but after disabling Popups, PageImages, and TextExtracts, the problem remained. [18:01:06] Any further thoughts on where to look? Could this be some config oversight again or perhaps an undocumented change in search functionality? [18:01:22] IIRC you checked the config for that one? [18:02:42] Yes but I'm not sure if any of our configs are the culprits. We do define a couple of custom namespaces using $wgExtraNamespaces, could it be in there somewhere? [18:02:55] As long as you're adding, not overriding... [18:03:06] ie [] = numbers, not = [] [18:03:52] We're basically doing $wgExtraNamespaces = array( 200 => "Company", 201 => "Company_talk", 202 => "API", 203 => "API_talk" ); [18:04:04] That won't help [18:04:07] You're removing the defaults [18:04:15] We've had it for ages [18:04:15] Uh [18:04:18] And this used to work [18:04:21] No, not extra namespaces [18:05:15] $wgNamespacesToBeSearchedDefault wasn't it? [18:05:35] That value is still set to its default in DefaultSettings.php [18:06:15] We do also set $wgNamespacesWithSubpages and $wgNamespaceProtection[NS_API] (where we previously defined NS_API as namespace 202) [18:06:28] $wgNamespacesWithSubpages += array( NS_MAIN => true, NS_PROJECT => true, 200 => true, 201 => true, NS_TEMPLATE => true, 202 => true, 203 => true,); [18:06:55] Those really shouldn't affect it [18:07:21] didn't think so but I there's so much I don't know, never can be sure [18:07:36] I can't remember where we got with the debugging last time [18:07:49] Beyond it looking like possibly SMW related [18:08:07] It's not SMW because we have a wiki that doesn't have SMW and it has the same problem [18:08:43] And the feature works elsewhere on the same version [18:08:53] Which narrows it down to config/infrastructure [18:12:47] $wgNamespacesToBeSearchedDefault = [ NS_MAIN => true,]; [18:12:54] that is what's in DefaultSettings.php [18:15:32] I can't actually see where SMW is adding to that variable [18:15:47] SemanticMediaWiki/src/NamespaceManager.php [18:15:53] https://github.com/search?q=org%3ASemanticMediaWiki+NamespacesToBeSearchedDefault [18:16:30] GH search sucking [18:16:30] https://github.com/SemanticMediaWiki/SemanticMediaWiki/blob/master/src/NamespaceManager.php#L243-L246 [18:16:47] yep there you go, those are NS 102 and 108 [18:17:09] but our non-SMW wiki has the same problem and its api call only has namespace=0 while template is namespace 10 [18:21:27] As before, works fine on my dev wiki when passing only =0 [18:22:13] on 1.34? [18:22:21] master [18:22:43] And it works here on https://wikifarm.wmflabs.org/cpt/index.php/Special:Version [18:22:45] Which has SMW etc [18:23:46] could it be some kind of data or search index issue? I'm not familiar with how opensearch works. [18:24:28] possible [18:24:34] I presume you're not using elasticsearch etc? [18:24:44] correct, just the mediawiki default [18:27:54] what's the contents of the searchindex table like? [18:28:26] no idea, let me go look, my dev wikis have the same behavior, so I'll check there [18:28:34] in the non-smw one to keep it as simple as possible [18:29:46] test [18:29:56] Is there a way to put formatted code in here? [18:30:06] If it's a lot of text, use a pastebin [18:30:29] It's not, just mysql output. But basically there are about 311k rows in the non-SMW wiki's searchindex table [18:31:17] anything in particular you want me to query? [18:32:25] Mostly that it had contents and wasn't empty [18:34:06] yeah for that 311k rows there are about 219k pages in the wiki and there are about 254k distinct si_title values in the table [18:38:17] sorry got disconnected [18:42:24] I guess the next helpful thing would be trying to see what sql queries it's actually running in the search [18:43:07] I think I did that last week, hang on... [18:46:57] hmm, probably need to re-enable wgDebugLog and test again to try to get the query [18:51:04] SELECT page_namespace,page_title FROM `titlekey`,`page` WHERE (tk_page = page_id) AND tk_namespace = '0' AND (tk_key LIKE 'TEMPLATE%' ESCAPE '`' ) ORDER BY tk_key LIMIT 11 [18:51:48] Titlekey? [18:51:58] uh yeah we use that extension [18:52:25] * Reedy looks at brion [18:52:28] maybe not the right thing to have pasted, or could that extension be an issue? We've had it for some time [18:52:56] Did you type TEMPLATE into the search box? [18:53:10] in lowercase [18:53:28] let me try again with some weird text as a comparison [18:53:30] * brion looks up, watches [18:54:23] justinl: OOI, disable the extension and see if it fixes it? [18:54:34] TitleKey might still be a little primitive. But iirc it can only search one namespace at a time efficiently [18:54:43] k one moment [18:55:06] If you type "template:" it should start searching in the template namespace, otherwise it'll stick to main ns? [18:55:26] it doesn't, no results with "template:" [18:57:43] i've disabled TitleKey but same problem [18:58:22] could it be search index side effects from having used TitleKey? [18:58:47] oh wait... hang on... [18:59:31] ok it works if I use "Template:" [18:59:41] capital T [18:59:50] let me re-enable [19:01:01] Ok it does work with TitleKey enabled if I do "Template:" [19:01:15] however, the search page shows no results if I hit enter with that text [19:01:29] but it show show suggestions in the dropdown while that text is in the box [19:02:24] but we actually have a legit wiki page named Template so hitting enter without the colon brings up that page [19:05:58] strange though, one of our other wikis doesn't work with "Template:" with TitleKey enabled [19:06:09] I'm wondering what would happen if I disable and re-enabled it [19:08:04] not even working with TitleKey disabled [19:10:15] oh snap, it works when I change the skin from Vector to Monobook! [19:10:31] the first wiki defaults to monobook, the second I tried defaults to vector [19:10:52] and when I changed the first working wiki from monobook to vector, it stopped working [19:11:46] it's the search bar in the top right with vector that doesn't work, in the left nav bar with monobook it does [19:11:53] o_0 [19:11:58] yeah [19:12:32] I know our editors add a lot of CSS and JS customization, possible they could have added something affecting that? [19:12:44] sitewide CSS and JS, that is [19:12:57] maybe they tweaked the skins in a way that is breaking search? [19:14:22] CSS I wouldn't have thought so... JS is possible depending on what they're doing [19:16:58] hey, so im having a bit of a bother with mw 1.33.1 [19:17:29] does a later version fix this oddity where i get misformed urls on the error log? [19:17:35] urls that appear to be correct [19:18:36] such as this [19:18:45] [core:error] [pid 19560:tid 1576] (20024)The given path is misformatted or contained invalid characters: [client] AH00127: Cannot map GET /mediawiki/index.php/Special:RecentChanges HTTP/1.1 to file, referer: http://localhost/mediawiki/index.php/Main_Page [19:18:58] hold on this is weird, trying to figure out the combinations of TitleKey enabled/disabled and Vector/Monobook [19:19:46] Juesto: there seems to be numerous search results for this if you google... Mostly possibly windows related [19:20:05] Yeah i am using windows aparche httpd [19:20:11] Sorry to bother. [19:20:17] ok with TitleKey enabled neither skin works, let me disable [19:20:41] https://www.mediawiki.org/wiki/Topic:Ptfixrg86x22c0zv [19:20:48] Reedy: lately my experience with Google has been poor but sure, i should've bothered searching [19:20:49] https://www.mediawiki.org/wiki/Topic:Rp3ch884qkelkkgh [19:20:56] https://www.google.com/search?q=mediawiki+The+given+path+is+misformatted+or+contained+invalid+characters%3A+%5Bclient%5D+AH00127&oq=mediawiki+The+given+path+is+misformatted+or+contained+invalid+characters%3A+%5Bclient%5D+AH00127&aqs=chrome..69i57j69i64.2565j0j7&sourceid=chrome&ie=UTF-8 [19:21:21] it does show up fine on the browser, but this is still thrown on the server [19:21:38] ok nvm both skins work with TitleKey disabled, let me try the same thing on another wiki just to be sure [19:21:41] i have apcache and opcache enabled [19:23:22] ok same on the other wiki, so it is TitleKey [19:24:11] also again, i prefer chat, thanks anyway [19:24:19] i'll see what i can do with research [19:24:35] Juesto: it's more, there are answers that may work onwiki [19:24:48] I'm not telling you to ask elsewhere, but at least try the solutions proposed there [19:25:13] I'm aware this is the right channel for mw support [19:25:54] call it unnecessary but i think that sometimes discussing helps [19:26:04] but of course, most dont use windows here [19:29:17] oh do I need to run rebuildTitleKeys.php? [19:29:36] the docs only mention it for installation [19:31:59] fwiw it looks like I installed TitleKey on our wikis back in April 2016 while we were on MW 1.26.2. [19:32:13] I don't think I've ever run that script since then. [19:32:31] I doubt the extension has changed much since then [19:32:49] Could it have a problem with newer versions of MW? I haven't looked at its release notes. [19:33:04] Certainly possible [19:33:11] Wikimedia used it at one point IIRC, but don't know [19:33:25] I don't know if anyone else actively really uses it to maintain it [19:33:53] From what I can see, it should run the rebuild script every time you run update.php [19:36:09] Unless an updatelog row exists... [19:37:15] not sure what that means [19:37:19] https://github.com/wikimedia/mediawiki/blame/master/maintenance/update.php#L181-L195 [19:37:39] The docs don't suggest you need to [19:38:38] Well since it does seem to be interfering, not sure of the best approach. Remove it, removing the functionality it does provide, or not be able to search outside the Main namespace. [19:38:45] Or try a rebuild... [19:39:09] try a rebuild [19:39:16] you can't make it any worse :P [19:39:33] ok let me repro the issue on our smallest wiki so a rebuild would be fast [19:40:56] ashley: ^^ Are you actually using/maintaining TitleKey? :P [19:41:47] Reedy: ShoutWiki uses it (and IIRC there's a pending patch or two @ gerrit and been for quite a while); I'm not sure if it still is necessary in a post-MW 1.32+ world, but at least for 1.32 it seems to be needed [19:42:28] MW suddenly hasn't become case insensitive... So I guess it's needed if that's a feature you want [19:42:42] Reedy: (both merge conflicted) https://gerrit.wikimedia.org/r/#/c/mediawiki/extensions/TitleKey/+/497086/ https://gerrit.wikimedia.org/r/#/c/mediawiki/extensions/TitleKey/+/382968/ [19:43:19] And one is C-1 :P [19:43:38] the rebuild didn't fix the small wiki while titlekey is enabled [19:43:42] Uhhhh [19:43:43] >Apply Mabinogi World patch to make search suggestions work again for non-NS_MAIN namespaces under MW 1.32+ [19:43:54] This sounds very relevant to what justinl is seeing... [19:47:56] Thought someone would know better around here.... :s [19:48:41] Have you even tried the solutions on those bugs? [19:48:47] /topis [19:48:49] /topics [19:49:06] im not sure what i would be doing or getting into [19:49:06] IRC isn't a magic place. If the answer is potentially in the docs, we will point people at it [19:49:15] No value in us repeating stuff [19:49:23] Right [19:49:33] don't want to get spoonfed [19:50:40] ugh got disconnected again [19:56:44] So there is a possible fix but it conflicts with another and there's not currently a roadmap for implementing a fix? [19:58:35] The fix just conflicts because of other changes to similar files [19:59:21] That's what I could see, just wasn't sure if there would be a plan to resolve that in the near future. [19:59:36] Trying to decide if we should just disable the extension for now. [20:00:21] great, i feel overwhelmed, i probably should try to look at it later [20:05:42] ashley: justinl: rebased that patch [20:06:46] Just pulled the change, I'll give it a test on my dev wikis. [20:08:07] Of course... // This doesn't seem to be working in core? Temporary override. [20:08:12] Feels like someone should be writing a bug [20:08:22] ashley: Did one ever get written as far as you know? [20:09:53] cause broken shit aint gonna get fixed if no one knows about it... [20:10:44] doesn't seem to work [20:11:09] Reedy: awesome, thanks for rebasing :) wanna give it a +2 as well? ;) (all I can say is that it's been running on ShoutWiki ever since I submitted that patch to gerrit...actually prolly a day or two before that, even) I...can't say I've looked, so I'm not able to say with a 100% certainty but I don't think so? [20:11:29] justinl: There's an extra line needed in the config according to the README [20:11:45] I'd be keen to get the issue in core actually fixed [20:11:47] I saw that, tried it before the patch and it broke the wiki,I'll try it now. [20:12:13] back in a few, need a break and some fooood [20:18:03] fatal exception: [exception] [b09ac9e032233cd0d82a4c72] /api.php?action=opensearch&format=json&formatversion=2&search=Mesmer&namespace=0&limit=10&suggest=true Error from line 109 of /var/www/sites/gww-en/includes/api/SearchApi.php: Call to undefined method TitleKey::getProfiles() [20:18:20] that's after setting $wgSearchType = 'TitleKey' after the wfLoadExtension call [20:28:02] I don't quite get why you'd be seeing that, but in any case, you shouldn't need to set $wgSearchType (or at least we don't have it set on ShoutWiki despite running that TitleKey patch) [20:28:56] I saw it in the patch information at https://gerrit.wikimedia.org/r/#/c/mediawiki/extensions/TitleKey/+/497086/, figured I'd see if it helped. Nope. [20:32:43] Is there something else I'm not seeing in the README that needs to be done along with the patch that Reedy applied? [20:37:27] my macbook is crashing again... very helpful [20:38:30] I haven't dug into the tree to see why that fails [21:41:00] Is there a way to run JS when Visual Editor displays a parser function? [21:41:26] Inside of the edit. Both on initial load, and after editing it via the popup you get when clicking the edit button over the parser function [21:42:58] I found the first part quickly: I can run code on initialization using: mw.hook( 've.activationComplete' ).add( function() { [21:43:26] But after looking at the docs and source for an hour I'm still not sure what event to use to detect changes [21:46:26] my first idea would be to override ve.ce.MWTransclusionNode.prototype.generateContents [21:48:26] if you really wanted to detect changes, i think you'd have to somehow hook into the transaction system and parse the transactions to see what they are adding/removing. that would probably be a lot of work, and i'm not sure where to start. probably could look at how the undo/redo buttons function [21:48:44] but i'd look into overriding generateContents first [21:54:32] MatmaRex: uh. Is there no event I can just listen to? This seems like a really basic usecase for extending VE [21:55:07] At this point I'm fine with listening to an event which is fired way more often than I need here [22:00:16] JeroenDeDauw: we have events for when a child node is inserted, but they are only fired on parent nodes, so you'd have to register tons of handlers and that seems terrible [22:00:23] what's wrong with generateContents? [22:01:24] you could define your own subclass of ve.ce.MWTransclusionNode and ve.dm.MWTransclusionNode, adjust ve.dm.MWTransclusionNode.static.matchFunction in your subclass, and then you don't have to override the built-in things [22:01:41] (that is similar to how you'd add support for an extension tag) [22:03:18] this might not be well-documented, but the 'wikihiero' extension has a simple implementation you can copy-paste from [22:03:28] i need to go now, sorry. good luck [22:03:31] JeroenDeDauw_: ^ [22:03:51] oh, in case you missed my messages: [22:04:02] JeroenDeDauw: we have events for when a child node is inserted, but they are only fired on parent nodes, so you'd have to register tons of handlers and that seems terrible [22:04:02] what's wrong with generateContents? [22:04:03] you could define your own subclass of ve.ce.MWTransclusionNode and ve.dm.MWTransclusionNode, adjust ve.dm.MWTransclusionNode.static.matchFunction in your subcl [22:04:19] …subclass, and then you don't have to override the built-in things [22:55:50] ok [22:55:57] nothing i can do about the : error [22:56:03] it's specific to windows [23:02:06] Juesto: you mean : in IIS? [23:02:12] you can modify web.config to allow those [23:02:32] Skizzerz: apache httpd, and it's prone to misuse [23:02:41] ah [23:02:49] : shouldn't be blocked in apache though [23:03:07] its not, however i get unnecessary error log spam [23:03:16] with harmless errors that arent affecting the pages themselves [23:03:24] ..on clients [23:03:44] try setting up .htaccess with explicit RewriteRules instead of relying on path info [23:03:46] https://bz.apache.org/bugzilla/show_bug.cgi?id=41441 [23:03:51] !shorturl [23:03:51] 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 [23:03:57] that [23:04:12] i thought special pages can't be moved? [23:04:22] oh, aliases [23:04:23] this isn't moving any pages [23:04:29] nvm [23:04:37] by having a RewriteRule you can have /wiki/Special:Whatever URLs [23:04:41] and apache will know how to map them [23:04:41] it'll still happen, thanks for the workaround through [23:04:45] (Alias should work too) [23:05:30] that's the remedy then hm? [23:05:58] odd, i didnt get such answer on #httpd