[00:00:31] Is there a way to extract the first X characters of a wiki article? [00:03:01] I'm aware of TextExtracts but looking for something that wouldn't require an outside hook into the API; essentially, I'm just trying to get valid meta descriptions written [00:20:42] Hello [02:56:16] hi. what is the recommended way to save a wiki page for offline reading? [02:58:14] i'm not asking about the app, btw. [03:02:55] !collection [03:03:18] rooftopjoe: https://www.mediawiki.org/wiki/Extension:Collection [03:03:49] legoktm: i was trying to use the export facility (e.g., https://www.khronos.org/collada/wiki_collada/index.php?title=Special:Export&action=submit ) [03:03:59] but nothing seems to happen when i add or export [03:04:23] ah, add pages from *category* not pages by name [03:06:48] legoktm: i don't own the wiki/server so i don't have control over the installed extensions [03:07:13] but i seem to have managed [10:42:23] Hi! How do I add Cat-a-lot.js to my own Wiki? It's ok if it only works for me. [10:49:33] MikaelLindmark, Preferences -> Appearance -> custom JS ? [10:53:13] LuisMiguel: I run a MediaWiki 1.23.6, so I have no "custom JS " tab [10:54:13] LuisMiguel, Thanks for the help, anyway! [13:41:17] Hi, I have a question on the Translate plugin. Is someone here who has used that plugin? [13:52:38] phry: yes [13:52:52] cool :) [13:54:02] I noticed that when I link a page from within a translated section like [otherpage], I have to change that link to for example [otherpage/ar] in the arabian translation. [13:54:02] Is there a way to make a link always link to the same language it is linked from, without having to specirfy the language? [13:54:22] I'm thinking [otherpage/{CURRENT_LANGUAGE}] or something [13:58:46] hello, do mediawiki oauth tokens last forever / is there an expiry time? [14:05:01] phry: you must use Special:MyLanguage, see tutorial https://www.mediawiki.org/wiki/Help:Extension:Translate/Page_translation_example [14:05:30] chippy: observationally, there is an expiry [14:08:14] Nemo_bis: that links to the current UI-language, but not to the language of the currently selected site. [14:14:43] alternatively I would be happy with a way to extend the selector to also use the &setlanguage= parameter. [14:21:40] Nemo_bis: thanks [14:21:48] that's good enough for me :-) [14:22:03] I know some ouath providers have very long expiris [14:44:22] Hey, I've got a patch I'm getting ready to submit to gerrit for an extension. It was originally written by an ex-coworker of mine. What's the best way to give them attribution? Just put it in a comment? Or mention it in the commit message? [14:47:03] Putting it in the commit message and/or a comment in the code would be appropriate. [14:48:02] Though I will probably be told I am wrong about the commit message because the code reviewers in Mediawiki's Gerrit system get super anal about commit messages. [14:56:22] phry: "setlanguage" doesn't exist [14:56:47] phry: I'm not sure what you mean by "language of the currently selected site" [14:57:43] Okay, you have two different types of language using the Translate+ULS solution: [14:58:00] If I understand correctly, you worry that people will use the page language selector to change content language when they meant to change itnerface language? [14:58:09] The "Wiki"/User-Interface language. This is set by ?setlang=en [14:58:30] Interface language is not the wiki's language [14:58:34] On the other hand, the "page" language. That is accessed by /en [14:59:03] This is approximately correct, yes :) [14:59:06] So a user can access a page that is translated to french while the UI language is spanish [14:59:15] okay [14:59:31] time to see if I can get Scribunto working on FreeBSD [14:59:41] FatalNIX: thanks for the effort [14:59:49] phry: so, what do you expect the users to do? [14:59:50] It was complaining about not finding Lua for standalone because I have no idea where to put it :) [14:59:52] With your solution, when I am on testpage/en and I follow a link to Special:MyLanguage/testpage2, it will get me tt testpage2/sp [14:59:54] Nemo_bis: ? [15:00:12] because my UI language is spanish. [15:00:21] (es, not sp) [15:00:31] Yeah, sorry, just pulling an example ;) [15:00:33] phry: yes, and why is that incorrect? [15:00:57] I would assume that a user is able to navigate within a language. [15:00:59] FatalNIX: put the lua binary wherever you want it, make sure it's executable, and point $options['luaPath'] to it [15:01:05] phry: why? [15:01:34] interesting [15:01:38] Unless you have translations for all pages in all languages, there is no reason to assume that a user is reading a specific content language out of choice rather than necessity [15:02:00] FatalNIX: correction, point $wgScribuntoEngineConf['luastandalone']['luaPath'] to it [15:02:00] jackmcbarn: I wish the documentation said that, or at least I didn't see it. It just presents "you have this directory you call YourOSHere"... [15:02:15] FatalNIX: https://www.mediawiki.org/wiki/Extension:Scribunto#Additional_binaries says that [15:02:39] OH [15:02:40] We are making a wiki for refugees. Their mobile devices are most likely set to English, although they want to see content in arabic. So they use the selector to go to the page in arabic. [15:02:46] I missed it :> [15:02:59] Every link they follow, they will be set back to english. [15:03:09] phry: at last you explain your use case :) [15:03:29] ;) [15:03:37] phry: the solution is for them to use the ULS language selector to switch the interface language to arabic; why do you think that's not possible? [15:04:05] Is this wiki for refugees a spin-off of https://en.wikibooks.org/wiki/Refugee_Phrasebook ? [15:04:21] No, it's more something just for the current area [15:04:32] ULS is a bit when it comes to non-standard skins ;) [15:04:41] a bit what? [15:04:48] *bitch, sry [15:05:18] So it took a while but we finally established that your issue is one of skin compatibility with ULS. Ok. [15:05:30] Essentially, I currently have the choice beween a language selector that is weirdly hanging around or a language selector that's only available when I am logged in. [15:05:53] Hm. [15:05:55] Sorry, I've been working on this for a few days now. Maybe I shouldn't have ruled out ULS ;) [15:06:01] What skin is this? [15:06:07] Foreground [15:06:09] WHAT [15:06:18] so after fixing all the errors [15:06:19] That's a rather common one, it should be fixed. Please file the bug. [15:06:21] They also have a github issue on this, but nothing seems to be moving [15:06:30] Ah. [15:06:31] my import said that my templates were already installed ..... [15:06:31] lol [15:06:47] https://github.com/thingles/foreground/issues/177 [15:07:45] My tries to modify Foreground were not really overwhelming. [15:09:13] Also, the ULS menu looks far from good on smartphones, even when I get it to open. [15:10:45] So I'd rather prefer the users to use the tag for language changing. Preferrably I'd like it to change the wiki language with ?setlang=


[15:10:45] 	 The other workaround is currently to link all english pages to /en, all arabic to /ar etc.
[15:11:24] 	 phry: what is not good on mobile?
[15:11:51] 	 just a second, I'll make a screenshot
[15:12:31] 	 oooh. I'm doing something weird, because I tried importing wikipedia's templates, and I thought I had all the dependencies installed but the top of my template pages say things like {{[[Template:{{{1}}}|{{{1}}}]]}} 
[15:12:43] 	 phry: you could also test the Commons and Wikidata hack https://commons.wikimedia.org/wiki/MediaWiki:AnonymousI18N.js
[15:13:21] 	 Though I have no idea where the dropdown would go in foreground
[15:15:59] 	 http://imgur.com/veTAirX that's what ULS looks on mobile
[15:16:17] 	 I'll take a look at that nemo, thanks
[15:20:38] 	 phry: what browser is that?
[15:21:31] 	 In CyanogenMod, it's just "Browser". I'd guess some sort of dumbed-down Android Chrome.
[15:22:33] 	 That's the android browser
[15:22:54] 	 Okay ^^
[15:26:57] 	 I've just tested three browsers and I don't see anything especially bad with the ULS trigger on vector and monobook
[15:28:28] 	 Opera Mini just breaks down with both skins, the Android browser is a bit overzealous with zooming and Chrome is to the contrary quite a zoom-slacker. Nothing special.
[15:30:02] 	 yeah I dunno
[15:30:18] 	 So my best bet would be finding another skin that's a bit more responsive, but still works with ULS
[15:30:30] 	 my infoboxes are definately not working, and I think it has to do with the fact that I am seeing {{}}'s on my wiki pages
[15:34:14] 	 phry: see http://imgur.com/a/D7Rxw for Opera Mini, Chrome and android browser in order
[15:35:10] 	 Nemo_bis: That's the  tag, not the ULS menu. Following those links won't change the wiki language, unlike the ULS menu.#
[15:35:57] 	 Which brings back the problem: open the page, it is english, select arabic, follow a link, it's english again.
[15:38:02] 	 You get the ULS menu that will really change the UI language (including LTR vs RTL) only if you click on the language symbol in the upper right corner (see this link: http://i.imgur.com/sK47x72.png )
[15:39:32] 	 phry: I suspect you didn't scroll down.
[15:39:55] 	 Nemo_bis: yuo, you got me. sorry
[15:40:13] 	 *yup
[15:41:53] 	 mh... if I would have to rate your screenshots, I'd say it still renders "experimental" at best :|
[15:41:53] 	 I guess I'll have to work myself into the code..
[15:44:04] 	 phry: maybe, bu definitely better than  :)
[15:44:20] 	 And also better than any other language selector I ever saw on the web. If you find a better one, please tell us.
[15:45:40] 	 I'll try ;)
[17:24:42] 	 Hello trusted online friends!!!
[17:25:53] 	 I am trying to get a sitemap to work with Google Webmaster...I am following the tutorial found here: https://www.mediawiki.org/wiki/Manual:GenerateSitemap.php#Options
[17:26:12] 	 I get an error saying that I do not have libssh2.dll installed
[17:26:19] 	 where can i download that?
[17:26:35] 	 refering to the code segement at bottom 
[17:26:40] 	 called "Example"
[17:27:04] 	 Having an issue with mediawiki and TimedMediaHandler throwing me an error: Strict Standard Declaration of OggHandlerTMH::formatMetadata() should be compatible with MediaHandler::formatMetadata($image, $context = false) in ****/extensions/TimedMediaHandler/handlers/OggHandler/OggHandler.php on line 5
[17:28:06] 	 using php maintenance/generateSitemap.php \ --fspath sitemap \ --server https://wiki.gamepaduniverse.com \ --urlpath https:/wiki.gamepaduniverse.com/sitemap in commandline
[17:35:27] 	 I figured it out trusted friends..nw
[17:44:27] 	 I don't like how English Wikipedia blocks by IP regardless of auth. Seems to me that if someone auths from a blocked IP, they should be allowed through (assuming they created the account from a non-blocked IP and have verified from the non-blocked IP that they are also coming in from the blocked IP). I assume this problem is due to some limitation in mediawiki?
[17:45:41] 	 I believe you can have accounts that ignore IP blocks. Also Blocks can be changed as to not block logged in editors or to block account creation or something
[17:45:58] 	 It's probably set up how it is because of how easy it would be to just log in with an account
[17:46:00] 	 Yeah, but their admins don't grant this unless you're a political dissident or something.
[17:46:13] 	 JordiGH: no, it's not a limitation of mediawiki. somebody has to choose to do (check a box) that every time they issue a block.
[17:46:32] 	 :? Lmao I've seen people using proxies and tor get ip exemption
[17:46:38] 	 It depends on why you need the right
[17:47:12] 	 JordiGH: you can have IP-only blocks (do not affect logged in users), non-IP-only (affect logged in users), and then there's the ipblock-exempt user right which makes the second kind of blocks not affect you.
[17:47:12] 	 Is it really that easy to verify from a non-blocked IP?
[17:47:39] 	 This situation doesn't make sense to me: https://en.wikipedia.org/wiki/User_talk:JordiGH
[17:49:05] 	 JordiGH: i'm not sure why you expect any sense from the english wikipedia. :)
[17:49:28] 	 It's kind of comforting if you can confirm that the admins are kind of nuts.
[17:50:17] 	 JordiGH: depends on the person, but there is a certain amount of ones that are, and they affect the policies disproportionately.
[17:50:59] 	 I mean in this particular case. Is it because they have had cases of people being able to temporarily auth from a non-blocked IP and then proceed to do all sorts of nasty things by authing from a blocked IP?
[17:52:27] 	 JordiGH: you can try applying for a global IP block exemption, but i have no idea what the policies around that are. https://meta.wikimedia.org/wiki/IP_block_exempt
[17:52:31] 	 NDK|Cloud: The limitation I have in mind is that Mediawiki can't currently say, "hey! we detected a suspiciuos log-in attempt to your account from a blocked IP. Was it really you?"
[17:53:53] 	 (the people on Meta, in my experience, are generally much more reasonable :) )
[18:04:48] 	 chippy, Nemo_bis: AFAIK there is no expiry for OAuth access tokens
[18:08:06] 	 tgr: ahh interesting.
[18:10:04] 	 I'm doing a little oauth thing - where most simple get api queries would be done on behalf of the logged in user, but occasionally the system itself may perform one (as a bot), so it could be good to use the same code rather than other/no authentication
[18:42:18] 	 chippy: feel free to ping me if that does not work as expected
[18:44:35] 	 JordiGH: allowing anyone with occasional access to a non-blocked IP to opt out of the IP block would kind of defeat the purpose of having the IP block in the first place
[18:45:13] 	 tgr: Why? Why would a person who can create an account from a non-blocked IP suddenly become evil when that same person goes to a blocked IP?
[18:46:23] 	 the more typical use case is that they are evil all the time but their access to the non-blocked IP is temporary or less convenient
[18:47:36] 	 tgr: Then block the user, not the IP? They would have to go through the annoyance of continually create accounts from the non-blocked IP.
[18:47:47] 	 And hopefully end up giving up.
[18:48:50] 	 JordiGH: I don't think it's possible to block account creation but not edits from an IP currently
[18:49:13] 	 So, perhaps if we address this Mediawiki limitation, the English Wikipedia admins will be more reasonable?
[18:51:59] 	 you should probably ask that on enwiki, people who actually deal with vandalism will have more insight into what might or might not be reasonable
[18:52:58] 	 FWIW I plan to work on the suspicious login thing ( https://phabricator.wikimedia.org/T107707 ) at some indefinite point in the future, but I don't think that's related to your problem at all
[18:57:30] 	 Sounds very related to my point.
[18:57:38] 	 Why do you think it's not?
[19:01:27] 	 JordiGH: his issue has to do with logging into a new device for the first time and providing a warning to the user, not with how blocks are handled
[19:01:57] 	 NDK|Cloud: It could easily be extended to disable logging in from a blocked IP and warning the user.
[19:02:51] 	 And then you wouldn't have to block logged-in users by IP address.
[19:03:04] 	 Since you need access to a trusted IP to log in.
[19:03:48] 	 not really, you could log in on another wiki where you are not blocked, for example
[19:04:04] 	 Then don't propagate permissions across wikis?
[19:04:14] 	 login (and to some extent) account creation is global, most blocks are local
[19:04:19] 	 authorisation isn't propagated, but authentication is
[19:05:51] 	 anyway IMO it would be a good idea to verify first that enwiki admins would accept this solution before putting too much energy into advocating for it
[19:06:24] 	 The enwiki admins so far seem unable to grasp the subtleties of the argument... it' like trying to convince someone whose job depends on not understanding it. :-/
[19:07:01] 	 I wonder if they're just unaware of what kind of granular controls are possible to create.
[20:10:44] 	 Huh. That's an interesting problem
[20:11:03] 	 so I have a private wiki with several articles documenting virtual machines I have running
[20:11:11] 	 every article is titled VM:hostname
[20:11:42] 	 they are all part of the Virtual_Machines category, but when I type Category:Virtual_Machines, I didn't realize that all of them fall under the letter V XD
[20:12:06] 	 [[Category:Foo|Name to use for sorting]]
[20:12:34] 	 Actually... Is it a proper namespace?
[20:13:06] 	 shouldn't you have VM: set up as a namespace?
[20:13:18] 	 I don't remember whether that changes sorting or not
[20:13:26] 	 Yeah, hence me asking
[20:13:29] 	 I couldn't remember either
[20:13:42] 	 Presumably it does, ie images
[20:13:56] 	 yeah... wouldn't want them all appearing under File
[20:14:28] 	 Although media shows up separately in categories
[20:14:45] 	 i.e. https://commons.wikimedia.org/wiki/Category:Gerrit_Alberts
[20:14:56] 	 Oh.. so THATs what sorting is for
[20:15:23] 	 yep
[20:15:36] 	 I shouldn't have ignored sorting in the documentation :)
[20:15:43] 	 I was like, pssh, don't need that lololol
[20:17:40] 	 Since about 1.16, namespaces aren't included in the sortkey anymore
[20:17:48] 	 I think
[20:17:55] 	 bawolff: IF they're registered, presumably
[20:18:07] 	 yeah
[20:18:36] 	 !extranamespaces
[20:18:36] 	 To add a namespace, modify your LocalSettings.php file, and add namespaces via $wgExtraNamespaces. A comprehensive set of instructions for adding a namespace can be found at . $wgExtraNamespaces can also be used to rename the default namespaces.
[20:18:41] 	 FatalNIX: ^^
[20:19:39] 	 note that if you already have pages with that namespace before it was created, they will be hidden
[20:19:48] 	 there's a maintenance script to fix that
[20:25:56] 	 Huh.
[20:26:11] 	 I wonder why adding custom namespaces isn't something you can do without edits to the config file
[20:26:49] 	 Same reason you can't do a tonne of stuff to MW without editing that file :)
[20:27:04] 	 because namespace configuration isn't stored in the db, but in the config file?
[21:57:40] 	 I am having issues with getting my site map to register with google webmaster. On my server it is saved to https:/wiki.gamepaduniverse.com/w/sitemap. A number of .xml files get exported there.  I tried including the entire directory, but get a the following error:General HTTP error: HTTP 403 error (Forbidden) HTTP Error: 403 
[21:58:04] 	 how do I get google webmaster working with my site map
[21:58:28] 	 I followed the Example at the bottom of https://www.mediawiki.org/wiki/Manual:GenerateSitemap.php
[21:59:23] 	 php C:\Inetpub\vhosts\gamepaduniverse.com\wiki.gamepaduniverse.com\w\maintenance/generateSitemap.php --fspath sitemap --server https://wiki.gamepaduniverse.com --urlpath https:/wiki.gamepaduniverse.com/w/sitemap --compress no 
[21:59:38] 	 ^ is what I used to create the sitemaps
[21:59:48] 	 and i didn't notice any errors with that
[22:03:36] 	 any help
[22:03:36] 	 ?
[22:10:52] 	 hola
[22:11:01] 	 good folks of the mediawiki realm
[22:20:02] 	 *sigh*
[22:26:03] 	 ?
[22:27:30] 	 lolololol, what is the issue exactlY?
[22:27:32] 	 exactly*
[22:28:03] 	 General HTTP error: HTTP 403 error (Forbidden) HTTP Error: 403
[22:28:14] 	 on google webmaster
[22:28:23] 	 when trying to submit sitemap
[22:28:55] 	 what link are you sending it to?
[22:29:51] 	 hello
[22:29:53] 	 omg
[22:29:58] 	 "/w/sitemap/"
[22:30:09] 	 can you browse to that directory
[22:30:09] 	 ?
[22:30:12] 	 sorry apparently this chat does not accept backslash as the first character
[22:30:17] 	 yes
[22:30:19] 	 i have
[22:30:24] 	 and everything is in there
[22:30:35] 	 looks like it is broken down by namespace
[22:30:38] 	 https://wiki.gamepaduniverse.com/w/sitemap/ returns 403 to me as well
[22:30:38] 	 for each xml file
[22:31:21] 	 so what should i do?
[22:31:55] 	 did you say this directory shows to you, you don't get a 403?
[22:32:12] 	 i get the same error as you
[22:32:17] 	 i am on my server now
[22:32:20] 	 looking at the directory
[22:32:30] 	 if you need me to check anything to troubleshoot this guy
[22:32:57] 	 sounds like an issue with your server config rather than mediawiki
[22:33:14] 	 question:
[22:33:39] 	 on my robots.txt file I have "Disallow: /w/"
[22:33:41] 	 on there
[22:33:46] 	 does that impact anything dealing with this?
[22:33:59] 	 i also have Allow: /w/sitemap
[22:34:12] 	 robots.txt? no
[22:34:20] 	 it's up to the client to follow that
[22:34:24] 	 this is an issue on your server
[22:34:35] 	 ok, well i mean, what should i do
[22:34:50] 	 the directory is set be writable
[22:34:55] 	 but i guess that isnt the issue
[22:34:57] 	 if we cant read it
[22:35:06] 	 web server needs to be able to read it for starters
[22:35:18] 	 so check that
[22:36:14] 	 its say its is readable and writable
[22:37:42] 	 all the individual files show up alright
[22:37:46] 	 such as https://wiki.gamepaduniverse.com/w/sitemap/sitemap-mediawiki_1-NS_10-0.xml
[22:37:50] 	 its just the directory
[22:38:34] 	 FYI, I am using Windows Server 2012 R2
[22:46:14] 	 lolololol, does google webmaster tools deal with directories..?
[23:05:38] 	 Krenair: I decided just to add each single xml to it
[23:05:44] 	 it seems to be pulling through
[23:05:57] 	 there are some ns that i probably dont want indexed by google
[23:05:59] 	 so that is cool
[23:06:28] 	 I created a cron job too
[23:06:43] 	 so the content in the xml files should updated every day automatically
[23:06:45] 	 that works with me