[00:00:17] hahaha, I could immediately spot Tim's joke there [00:00:23] checked the history - here it is, Tim in 2003 [00:00:30] that's pretty funny [00:00:31] 03(mod) Provide accessibility of Wikipedia' s sister projects through *one* interwiki prefix - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=4285 (10gangleri) [00:00:48] 'Thanks MediaWiki default, for being the best bot possible. You bring defaults to MediaWiki. --Onekopaka 00:51, 8 June 2007 (UTC)' [00:00:56] TimStarling: your humor has a signature :) [00:01:01] TimStarling: I think it has a personality cult somehow [00:03:28] Mediawiki_default got banned on Uncyclopedia [00:03:53] by banning 127.0.0.1 [00:04:20] doesn't work, it doesn't check for bans [00:04:36] hmm, well some local bot did [00:04:40] maybe spam cleanup? [00:04:54] you know, I made cleanupSpam.php specially check for local blocks on request from memory alpha [00:05:11] ;-D [00:05:22] they said they were going to block it if I didn't wise up, so I said "well first I'll have to make it blockable, hang on" [00:05:36] then they backed off [00:06:07] haha [00:06:21] win [00:06:43] TimStarling: It is not nice to pwn everyone [00:07:39] =) [00:08:12] btw tim, since timtest=newpp works on post ("sweet"), I've added some radio boxen to the edit page on test.wp to let you preview with the new or old parser, depending on mood [00:08:14] domas: SILENCE! man the squids! [00:08:20] *domas cries [00:08:39] not that people with local test machines need it, but think of the children... [00:09:03] Splarka: it is pain to import all the templates from real wikis to test wikis [00:09:07] Splarka: and extensions and such [00:09:12] if it gets annoying, I can move it to an optional gadget [00:09:22] I usually miss lots of things on my local test instance because it doesn't do so many things that live site does [00:09:24] domas: no kidding [00:09:52] like... only yesterday I did set up two-db-server stuff here [00:09:59] for my mediawiki master/slave communication testing [00:11:25] domas: I tried to import current revs only, but my drive has little space left ;( [00:18:11] Hmm, I'm getting this warning, but I have no idea what part of my extension is causing it... any suggestions? Warning: call_user_func_array() [function.call-user-func-array]: First argument is expected to be a valid callback, 'Array' was given in /w/includes/Parser.php on line 3220 [00:19:32] that's because you lose [00:19:51] owait, just figured out what caused it, now for how to fix it... [00:19:52] domas: can you really 2x squid efficiency? [00:20:08] Skizzerz: right, stop being a loser ;) [00:20:26] AaronSchulz: yup [00:20:30] thank you all for the help [00:20:38] AaronSchulz: it's storage efficiency [00:20:47] except I suck at the API... which is probably why it arose to begin with :( [00:21:44] DUH!! >_> [00:21:59] I was using a title object instead of a title name when calling the API [00:37:44] help please... i am having a heck of a time... how does one implement the following = "An internal image linking to an internal page" ? [00:37:55] i cant figure it out - http://en.wikipedia.org/wiki/Help:Navigational_image has me all confused [00:38:18] One checks if one has 1.12 [00:38:52] i want to put up a firefox image, and when someone clicks it, i want mediawiki to take the user to an internal mediawiki page = "Firefox+Localfile" page [00:39:06] and not display anything about the Image:Firefox_logo.jpg [00:39:18] make sense to anyone? [00:40:00] 03(NEW) Page count not updated. - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=12618 15enhancement; normal; MediaWiki: Special pages; (bugzilla.wikimedia) [00:40:35] just want the webpage to show an Image = Firefox_logo.jpg and when you click this link it will take the person to the internal page "Firefox+Localfiles" ... not the Image page for Firefox_logo.jpg ... [00:40:35] !imagemap | agntdcooper [00:40:35] --mwbot-- agntdcooper: ImageMap is an extension that allows you to override the default linking behavior of images. See for more information. [00:40:45] aha [00:40:47] excellent [00:40:48] thank you [00:40:49] also [00:40:53] do you have external images enabled? [00:41:23] Splarka: im still kinda new to mediawiki... i would ASSUME so [00:41:48] try pasting into your wikicode: [00:42:06] 14(INVALID) "No such special page" message on special: makesysop with aliases - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=12561 +comment (10stanley) [00:42:08] "this is an external image http://www.google.com/intl/en_ALL/images/logo.gif " [00:42:21] if the link is just rendered as a link, you don't, if it renders as an you do [00:42:59] Splarka: the image showed up, with "this is an external image" text [00:43:02] =) [00:43:11] so that works [00:43:15] then you can do things like this [00:43:29] 03(FIXED) Provide accessibility of Wikipedia' s sister projects through *one* interwiki prefix - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=4285 (10cbrown1023) [00:43:30] [http://www.google.com/ http://www.google.com/intl/en_ALL/images/logo.gif] [00:43:44] but you need the full URL of the image, not the [[Image:Foo.gif]] format [00:43:50] and the full URL of the local link [00:43:56] cool [00:44:00] is it possible to call for {{LOCALTIME}} + 1 hour? [00:44:04] it isn't as nice as imagemap, but it doesn't require an extension [00:44:20] Juandev: do you have ParserFunctions ? #expr can probably do it [00:44:24] Splarka: i am gonna try yr suggestion real quick [00:44:26] thanks! [00:44:47] k [00:44:56] Splarka: kinda: {{#expr: {{LOCALTIME}} + 1}}? [00:46:43] dont, know whats the deffinition for one hour, if exist [00:47:33] http://meta.wikimedia.org/wiki/Help:ParserFunctions [00:47:46] http://meta.wikimedia.org/wiki/Help:Magic_words [00:47:48] *Juandev reads:-) [00:47:56] huh [00:48:22] there might be an example on a big wiki like en.wp [00:52:11] Juandev: http://en.wikipedia.org/wiki/Template_talk:Utc [00:52:22] 03(mod) Recursive checkuser featured needed - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=9858 +comment (10tstarling) [00:53:57] ahh, it just uses #time [00:54:43] I guess that will work for your local time if you adjust it properly [00:55:49] Hello.. it seems that now bot's edit is displayed if the edit is new page creation, how to make this new page edit hidden? [00:56:31] Splarka: ha, this link is better [01:06:03] Splarka: many thx, for nav, we got it [01:06:15] 03(NEW) cleaner URLs for editing, previewing skins, etc. - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=12619 15enhancement; normal; MediaWiki: General/Unknown; (onekopaka) [01:06:16] cool [01:17:57] borgx, that was fixed, the change just needs to go live. [01:18:02] Or it should have been fixed. [01:19:01] !b 12611 | borgx [01:19:01] --mwbot-- borgx: http://bugzilla.wikimedia.org/show_bug.cgi?id=12611 [01:20:02] mwbot: thanks [01:28:56] 03(mod) Double-parse in parser function results - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=5678 summary; +comment (10tstarling) [01:36:40] 03(mod) Extension stylesheets should be loaded before any others - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=10184 +comment (10arnomane) [01:44:43] hi. i just updated from 1.10 to 1.11. but now i have a strange problem. a request to e.g. index.php?title=Hauptseite is answered with a redirect to "Index.php". i don't have a Index.php with uppercase i and i don't see why and where this redirect comes from. [01:47:45] Hi Rdsmith4 [01:48:11] Rdsmith4, http://en.wikipedia.org/wiki/Wikipedia:Requests_for_adminship/Soxred93 <--- is that closed properly? [01:48:14] it seems to happen somewhere in the finalCleanup call [01:48:30] looks like it, yes [01:48:55] sweet [01:49:49] 03(mod) Separate group for ipblock-exempt on en.wikipedia - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=9862 +comment (10lar) [01:49:56] anyone willing to help? this looks like a bug. [01:50:11] a strange one [01:51:37] 03(mod) Recursive checkuser featured needed - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=9858 (10lar) [02:03:49] 03(NEW) utf8 issue with list=recentchanges - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=12620 15enhancement; normal; MediaWiki: API; (ariel) [02:16:13] 04(REOPENED) Recursive checkuser featured needed - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=9858 +comment (10JSchulz_4587) [02:16:59] 03(mod) Separate group for ipblock-exempt on en.wikipedia - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=9862 (10NavouWiki) [02:29:24] 03(mod) Recursive checkuser featured needed - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=9858 (10JSchulz_4587) [02:36:43] 03(mod) Recursive checkuser featured needed - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=9858 (10lar) [02:40:23] 03(mod) cleaner URLs for editing, previewing skins, etc. - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=12619 +comment (10dan_the_man) [02:42:50] 03(mod) Recursive checkuser featured needed - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=9858 (10JSchulz_4587) [02:44:28] 03(mod) Extension stylesheets should be loaded before any others - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=10184 +easy; +comment (10Simetrical+wikibugs) [03:09:34] 03(mod) Javascript variables for section and preview mode - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=8946 (10danny_b) [03:13:18] 03(mod) Allow import to different page name - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=6808 (10danny_b) [03:18:43] heya any mediawiki dev gurus here? I am trying to work out how to parse mediawiki syntax to HTML [03:19:06] sour1, call Parser::parse(). [03:19:32] I've had a look at that, I couldn't get it working properly (goign back about 6 months ago) [03:19:44] spent a good couple hours trying to sort it out [03:25:23] 03erik * r29715 10/trunk/extensions/LiquidThreads/ (Lqt.i18n.php LqtBaseView.php LqtPages.php lqt.css): [03:25:23] First attempt at sorting controls, support three modes: [03:25:23] - last modified first (any change to thread) [03:25:23] - most recent created first [03:25:23] - oldest created first (essentially old talk page style sorting) [03:25:25] default sort can be persistently saved in user prefs. [03:25:43] hello [03:26:12] any taker for a beginner config question? [03:27:10] !ask | manu1226 [03:27:10] --mwbot-- manu1226: Don't say "I have a question", or ask "Is anyone around?" or "Can anyone help?". Just ask the question, and someone will help you if they can. Also, please read < http://workaround.org/moin/GettingHelpOnIrc > for a good explanation of getting help on IRC. [03:28:18] is there a way to force SSL usage when a user login? [03:28:58] I could use an apache rewrite rule so that the login page goes to an https URI, but that won't prevent a logged in user to go back to http by mistake on its own later [03:29:46] what i'm looking for is http:// working only for anonymous users, and https:// mandatory for logged in users [03:30:34] Hmm .. not really a "beginner" config question :) [03:31:08] well, I'm a beginner with mediawiki, not with web apps security :-) [03:31:32] My first (and likely wrong) impulse would be to set up two separate installs of mediawiki that use the same database. [03:31:46] Is there some way to work cookies or session IDs or something into RewriteCond? [03:31:47] Have one accessible through https, the other through http, and disable login and editing on the http one [03:31:56] That sounds . . . silly. [03:32:03] Why don't you just make it HTTPS for everyone? [03:32:06] Yeah, as I said, it's probably a wrong impulse :) [03:32:34] Simetrical: that what I'll do if I can't get http only for anonymous users, but https for everyone is CPU consuming [03:32:54] You could use an extension. [03:32:57] should there be ways to create empty pages? [03:33:19] MZMcBride: don't see why [03:33:25] Pick some point early on, right after $wgUser is initialized, check if the user is logged in, and if so output an HTTP redirect and exit(). [03:33:27] Simetrical: which one? Or do you mean write one? [03:33:36] It would be a fairly trivial extension to write, I think. [03:33:45] well i found two: {{subst:ns:0}} and {{subst:}} [03:33:49] Although, of course, it would result in a 302 redirect on every page view. [03:34:17] MZMcBride, I guess emptiness check happens after pre-save transform. But I don't really care, why does it matter if someone makes an empty page? [03:34:25] s/after/before/ [03:34:29] Simetrical: Well, check the local server. If it's https and the user's logged in, no need to redit. If it's http and the user's not logged in, no need to redirect either [03:34:41] there's a VPT post about it; it confuses people, mostly, because generally it's not possible to do [03:35:09] amidaniel, how are you going to change the link formatting? Change $wgServer after things have been initialized enough to figure out if the viewer is logged in? [03:35:23] MZMcBride, er, then why are they trying to do it? [03:35:37] It should be trivial enough to fix, I imagine, although it depends on the code path. [03:35:42] Could be it would be messy. [03:35:43] no one's trying to do it; someone posted asking how it happened [03:35:49] anyone know of an xwiki style skin? (see here for an example of an xwiki: http://gometric.us) [03:35:52] I don't see why we don't allow it generally. [03:36:23] well generally it's difficult to do; you save page with no content; how do you expect it to create a page? [03:36:24] Chrisgrant, this isn't a great place to ask about skins. We're more dev and Wikipedia types here, than third-party users. [03:36:27] Simetrical: No, I mean just query $_SERVER['HTTPS'] [03:36:38] ok [03:36:42] !skins | Chrisgrant [03:36:42] --mwbot-- Chrisgrant: Overview: . Skin usage: . Gallery of CSS styles: . Writing your own: [03:36:42] it seems mod_rewrite can handle HTTP_COOKIE. Does that help? [03:36:55] amidaniel, hmm, interesting. So start off serving the page according to whether the request was secured, then redirect later if necessary? [03:36:58] That's an interesting tactic. [03:37:00] Simetrical: So, if ($wgUser->isLoggedIn() && !$_SERVER['HTTPS']) header("Location: http:...'); [03:37:15] Simetrical: Yes, if it's do-able [03:37:19] I don't see why not. [03:37:32] You can set $wgServer contingent on $_SERVER. [03:37:43] Sure [03:39:48] anyway, it seems that this https enforcement for logged in users is not something ready out of the box [03:39:55] manu1226, nope. [03:40:32] that's weird nobody with enough time to implemnt it never needed it [03:40:53] It's not a very common request. [03:40:54] manu1226: Most users / developers of mediawiki aren't security freaks :) [03:41:02] Why would users' sessions need to be more secure than anyone else's? [03:41:18] amidaniel, well, or those who are don't mind running the whole site on HTTPS. [03:41:29] Simetrical: True [03:41:34] because it's pointless to authenticate yourself if the authentication can be defeated with just a sniffer [03:41:41] HTTPS is a good idea if you're dealing with stuff like credit card numbers or whatnot, for whatever reason. [03:42:04] manu1226, well, if you think someone's either going to get on a LAN with a user of your wiki, or hack some routers, just to break into your wiki. [03:42:15] Otherwise it's not a big problem. [03:42:33] I use LDAP authentication backend, and I've spent quite some time making the whole stuff secure, it's not to have authentication broken by a plain HTTP login page. That would be too stupid :-) [03:43:10] Simetrical: Well, https has become significantly more important in the wireless age. That's how most passes, etc. are fished out nowadays [03:43:19] Yes, that's true. [03:43:27] 03tstarling * r29716 10/trunk/phase3/includes/Parser.php: [03:43:27] * Fix heading immediately after comment-on-own-line [03:43:27] * Fix ===== on own line [03:43:36] manu1226: Couldn't you configure the LDAP extension to reauthenticate during every edit? [03:43:36] MZMcBride, you know, I bet the article save process doesn't even bother parsing the page. [03:43:43] It seems like too much effort to go to. [03:43:51] Simetrical: last sniffing attacks I have seen were warried by infected PCs that were spying other machines on the same LAN, sometime inserting exploits in ethernet frames sent to others PCs... [03:43:52] Maybe it could do only PST. [03:44:03] should i file a bug? [03:44:08] MZMcBride, if you like. [03:44:11] s/warried/carried/ [03:44:33] manu1226, if they log in using HTTPS, then subsequent requests only have a session token, so at most you can steal a session (which can be invalidated at will by just logging out). [03:44:40] Granted that reveals e-mail address and whatnot. [03:44:54] Simetrical: so no need to hack a rooter, an automatically compromised windows host stting on the network is enough [03:45:01] But I don't even think you can edit using the session, unless you intercept the initial login or the edit itself. [03:45:31] Simetrical: I believe the sessions are tied to the hostname. So you could hijack someone at the same comp as you :) [03:45:37] Or spook the hostname, obviously [03:45:42] amidaniel, but you need the edit token to edit. [03:45:57] That's not put on the network except at login or edit, is it? [03:46:11] Of course I suppose you could still delete, protect, block, etc., if those permissions are available. [03:46:13] session tied to hostname? what about NAT? [03:46:19] 03(NEW) Creating empty pages - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=12621 15enhancement; normal; MediaWiki: General/Unknown; (MZMcBride) [03:46:35] brevity is the soul of wit [03:46:42] manu1226, false positive then at worst. More worrisome is people with dynamic IPs, then, who might get logged out accidentally. [03:47:28] Simetrical: I assumed that the tokens were generated / looked up when the user accessed the edit page. That is, a session could be used to obtain the tokens [03:47:32] May be wrong though [03:47:39] Maybe. [03:47:58] ok, the easiest path is probably the two mediawiki instances sharing the same database. [03:48:11] Could do that. [03:48:23] I'll just have to disable logins on the read only one [03:48:41] The purpose of the tokens, I believe, are to prevent me from giving you a link like /w/index.php?title=Main_Page&action=edit&wpTextbox1=Foobar&wpSummary=vandalisms [03:48:57] and to have the login link replaced to a link to the write enabled instance [03:49:23] which means one more bit of hacking :-/ [03:49:33] Wow, you mean my first impulse was right? :P [03:50:15] well, it's not very appealing either. https only is probably the safest way [03:50:41] Anonymous users have rights too!! [03:52:04] anonymous users can use https [03:52:51] https imply server authentication by client, but the other way around is not mandatory [03:53:56] (anonymous = not logged in) [03:54:15] (even though really, they're less anonymous than registered users, since their IP is displayed publicly) [03:54:21] (at least if they edit) [03:55:43] oh, you mean that makes the two instances way not workable. Yes, I realized that. [03:58:19] 03(mod) It's possible to create new empty pages using {{subst:}}, etc. - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=12621 summary; +comment (10Simetrical+wikibugs) [04:20:35] hi all, i'm installing this extention here: http://www.mediawiki.org/wiki/Extension:HttpAuth . i'm wondering if the edits to LocalSettinsg.php need to be at the top of the file or down teh bottom? the page doesnt say, and i'm not familar with php [04:20:56] as far as I understod, bottom [04:21:04] thanks [04:21:07] Bottom. [04:21:28] but I don't know what I'm taking about, so listen to Simetrical [04:22:03] thanks both. [04:25:43] now to understand where it goes alongside the debianisation :| [04:26:16] hello! [04:26:47] Hi. [04:27:02] whats up ami? [04:28:12] are you using mediawiki? [04:28:27] !ask | ifreeman2 [04:28:27] --mwbot-- ifreeman2: Don't say "I have a question", or ask "Is anyone around?" or "Can anyone help?". Just ask the question, and someone will help you if they can. Also, please read < http://workaround.org/moin/GettingHelpOnIrc > for a good explanation of getting help on IRC. [04:28:58] hello bot [04:30:35] cu [04:31:17] hmm.. shame he had to go [04:32:25] Lol, that was odd :) [04:37:09] Am I the only one who hates wget? [04:38:00] Doesn't use stdin/out reasonably, is way too noisy, and apparently the quiet option overrides even explicit print switches like -S. [04:38:45] tar annoys me as well, but at least that has the excuse that it was deliberately meant to be a *tape* archiver. [04:39:30] So outputting to $TAPE by default is reasonable. [04:39:54] Actually, it seems it even does output to stdout by default on Ubuntu at least, if $TAPE is unset. [04:40:20] And always to stdout by default on RHEL. [04:40:28] is there anyway to make wget spit out to stdout by default instead of littering whatever directory I'm in with index.html.985453 ? [04:40:39] As that's the one thing that really annoys me about wget [04:40:49] amidaniel, alias wget='wget -O - ' [04:40:57] Or whatever the syntax is for aliases. [04:41:09] Oh, -O makes it use stdout? [04:41:22] -O [location] = output to location, - = stdout. [04:41:35] (If you want to output to ./- you can say ./-, I guess. :P) [04:41:38] Aha, good to know :)) [04:41:51] *Simetrical points amidaniel to the "man" command [04:42:05] Yeah yeah yeah .. I ain't got time for that shit [04:42:08] You are my "man" :P [04:46:46] is brion around? [04:46:51] brion-away: [04:46:53] dammit. [04:47:03] I was hoping to have lunch with him tomorrow (I'm in the area) [04:47:31] *amidaniel is sad Werdna doesn't want to come have lunch with him :(( [04:47:39] amidaniel: live in San Francisco? [04:48:58] Nope :) [04:50:45] Werdna! how are you finding SF? [05:24:56] hi all, iin the faq it explains how to remove the 'login' text. i'm wondering how to remove/modify the 'logout' text. i'm using a custom login module, which doesnt seem to logout properly when its click [05:25:00] ed [05:25:42] kgoetz: which logout text? [05:27:57] MZMcBride: eg "My talk My preferences My watchlist My contributions Log out" [05:28:16] those should all be in Special:Allmessages i think [05:28:44] yeah; [[MediaWiki:Mytalk]], etc. [05:30:10] um. [05:30:31] oh dear. did i confuse you? [05:30:47] yes you did. [05:30:55] i'm still not sure how i'm goin to edit/remove it ;) [05:31:21] http://pruebita.com/wiki/MediaWiki:Mytalk [05:31:26] go to that page on your own wiki [05:31:32] has anyone seen any Parser wrappers or something that's a bit more easier to use than the parse class [05:31:33] and then edit it and make it whatever you want [05:31:52] MZMcBride: thanks, i'll go and have a look [05:31:56] $html = parse ($wikitextblob) would be great !! [05:38:14] sour1: $wgOut->parse($wikitextblob); [05:39:52] I have a quick question, when I install MediaWiki I get this error message - # Warning: A value for session.save_path has not been set in PHP.ini. If the default value causes problems with saving session data, set it to a valid path which is read/write/execute for the user your web server is running under. | I un-commented that var in php.ini file last night but MediaWiki still gives me... [05:39:52] ...that error message | I have phpinfo() giving me this session.save_path /tmp /tmp | So it should be working! [05:39:55] My host allows me a php.ini each subdomain/domain. [05:40:44] I have a /tmp folder in the root directory of my webspace with permissions 755 and I have another one I created in the root of my wiki's domain directory which is /tmp with permissions 777 [05:41:05] it's just a warning [05:43:17] TimStarling I'll give that a go, cheers ! i think i was doing something simular but it kept coming up with strange errors [05:43:30] i'll look into it more later tonight [05:43:35] 03shinjiman * r29717 10/trunk/phase3/languages/messages/MessagesYue.php: * Update Cantonese translations [05:45:32] i'm using MW 1.10, and i'm thinking of filing a bug on some descriptive text - is ther ea public repo i can check for changes in? [05:46:36] svn.wikimedia.org [05:47:16] more specifically, you'll probably want: http://svn.wikimedia.org/viewvc/mediawiki/trunk/phase3/languages/messages/MessagesEn.php?revision=29709&view=markup [05:47:33] that'll show all of the most update to date messages [05:52:45] THANKS [05:52:49] er, *thanks [05:55:30] hm. the string doesnt appear at all in that file. perhaps its a part of the debianisation [05:55:54] ew. debian [05:56:08] :( [06:02:02] What size should the logo be of a wiki? [06:02:35] How can I show a remote hosted image. [[Image:URL]] doesn't seem to work. [06:03:22] !externalimages [06:03:22] --mwbot-- To allow images from elsewhere to be included in your wiki, see . To limit this to some specific sources, see . [06:03:45] thanks [06:06:31] 150 x 150 ? [06:09:46] Jacroe: 135px x 155 px [06:10:59] thanks [06:12:01] where'd you get that number mzm? uncyc? [06:12:25] 150x150 is nice and filling ^_^ [06:12:32] the enwp logo is rather skinny [06:12:40] and too close to the top [06:12:49] i used a random number generator :) [06:13:14] if you don't fill the area or you overfill it, it doesn't like it [06:13:18] !logo [06:13:18] --mwbot-- The logo that appears in the top left of each page is determined by the $wgLogo configuration setting in the LocalSettings.php file. To change this you simply need to change the value of $wgLogo to point to the URL of your own logo image. See for more information. [06:13:41] meh. it says it's supposed to be 135 x 135 [06:13:52] though mediawiki.org is right about 50% of the time [06:13:54] who does? [06:14:25] the height is 155px; in monobook/main.css [06:14:36] so bigger than that and it'd look ugly [06:14:44] but the width is based on text size [06:15:27] yes. so poor-vision wiki can have a really wide logo [06:17:18] quarterly branch a little delayed... [06:17:44] September 10 --> now is four months... [06:17:56] quarterly is too often [06:18:04] we haven't finished fixing the bugs in 1.11 yet [06:18:27] perhaps tri- or bi-annually? [06:18:34] or "whenever it's ready"? [06:19:48] I think 6 months would be better [06:21:56] does mediawiki.org have an editprotected system or something similar? [06:22:21] I don't know [06:22:24] I don't edit it very often [06:44:20] 03(mod) Title->getDBkey() often spelled incorrectly - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=12608 normal->15enhancement; normal->low; +comment (10ThomasBleher) [06:53:22] ugh, JS page cache acting up [06:57:54] no kidding [06:58:15] on wikt, common.js wouldn't purge for me, with a hard reload... I had to log in and disable page caching [06:58:21] it is just stuck [06:58:34] try disabling page caching in your prefs [06:58:35] I blanked the whole JS and it still loaded [06:58:45] that finally did it for me [06:59:30] blame domas [06:59:33] 03tstarling * r29718 10/trunk/ (4 files in 3 dirs): [06:59:33] * Fixed loop detection in LST [06:59:33] * Renamed some LST parser tests [06:59:33] * Fixed caption behaviour .It should not generate
 when there are spaces at the start. Used recursiveTagParse() instead of parse(), to avoid generation of a limit report, among other things.
[06:59:33] 	* Refactored template loop check
[06:59:35] 	* Updated a couple of parser tests
[07:00:08] 	bah, no luck
[07:01:24] 	this is kind of annoying
[07:01:29] 	Splarka: ;)
[07:01:47] 	TimStarling: how is the new parser coming along?
[07:01:57] 	rar?
[07:09:56] *Werdna 	waves
[07:12:30] *MZMcBride 	waves back
[07:16:26] 	Werdna! how are you finding SF?
[07:16:35] 	san francisco is fine
[07:16:41] 	very nice weather.. for san francisco, at least.
[07:17:38] 	no, I mean
[07:17:45] 	*how* are you finding it? in all the fog
[07:17:49] 	haha
[07:17:50] 	GPS
[07:18:07] 	not too much fog now though
[07:18:14] 	AFAIK fog is mainly a summer phenomenon.
[08:48:10] 	re
[08:54:48] 	03(mod) Add Create permission to Protectedpages - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=12583  (10huji.huji)
[09:01:29] 	Hello
[09:01:39] 	I got a spam problem on my Wiki.
[09:01:49] 	!spam
[09:01:49] --mwbot--	For more information about combating and handling spam in MediaWiki, see .
[09:01:53] *MZMcBride 	runs
[09:02:10] 	I already changed the permissions so that people needs to be registred / logged in to make changes, but I still got spam on the wiki
[09:02:36] 	!captcha
[09:02:36] --mwbot--	For more information about CAPTCHAs and MediaWiki, see .
[09:04:28] 	cool
[09:04:31] 	thanks
[09:05:19] 	btw: is there a way to force registered users to confirm their email address? I would like to be able to properly credit contributors
[09:06:00] 	require that for them to edit?
[09:06:03] 	with a known email address I can always send them money using PayPal (working on a book / courseware - want to split proceedings with the contributors)
[09:06:58] 	$wgGroupPermissions['*']['edit']              = false;
[09:07:02] 	$wgGroupPermissions['user']['edit']           = false;
[09:07:04] 	yes
[09:07:06] 	$wgGroupPermissions['emailconfirmed']['edit'] = true;
[09:07:10] 	for example
[09:07:17] 	see http://www.mediawiki.org/wiki/Help:User_rights for more
[09:07:24] 	basically, i don't want any un-confirmed users in the system
[09:07:36] 	can they read?
[09:08:05] 	be sure to disable upload/reupload/move etc
[09:13:18] 	03huji * r29719 10/trunk/phase3/ (11 files in 2 dirs): (bug 12608) Unifying the spelling of getDBkey() in the code.
[09:14:03] <_wooz>	lo
[09:15:10] 	03(FIXED) Title->getDBkey() often spelled incorrectly - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=12608  +comment (10huji.huji)
[09:20:17] 	03(mod) Title->getDBkey() often spelled incorrectly - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=12608  (10huji.huji)
[09:23:40] 	03siebrand * r29720 10/trunk/extensions/MetavidWiki/ (80 files in 12 dirs): 
[09:23:40] 	* set eol-style:native for all php/js/css/html/xml files
[09:23:40] 	* fix inconsistent newlines in a few files
[09:23:40] 	* fix typo in message
[09:26:48] 	03huji * r29721 10/trunk/phase3/ (11 files in 5 dirs): (bug 12608) (in continuation of r29719) Unifying the spelling of getDBkey() in the code.
[09:27:33] 	03(mod) Title->getDBkey() often spelled incorrectly - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=12608  (10huji.huji)
[09:55:46] 	03siebrand * r29722 10/trunk/extensions/ (44 files in 27 dirs): 
[09:55:46] 	* set eol-style:native where missing
[09:55:46] 	* fix inconsistent newlines in a few files
[10:01:42] 	03siebrand * r29723 10/trunk/extensions/ProofreadPage/proofread_index.js: * set eol-style:native (missed a new file)
[10:09:21] 	03siebrand * r29724 10/trunk/extensions/ (42 files in 35 dirs): (bug 12608) Unifying the spelling of getDBkey() in the extension code.
[10:12:21] 	03(mod) Title->getDBkey() often spelled incorrectly - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=12608  +comment (10siebrand)
[10:14:12] 	huh, siebrand: so you're fixing mis-casings now? heh
[10:14:27] 	I saw one the other day, but someone said php didn't care, so I put it out of my mind
[10:14:37] *Splarka 	sees the same thoughs in the bug comments
[10:14:43] 	siebrand: I was going to commit the fix for extensions, when I saw you already did that!
[10:15:02] 	well, if it is so important
[10:15:17] 	someone should do a global check for all function and variable names for case ^_^
[10:15:20] 	03(mod) Warning got in includes/Title.php - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=11801  +comment (10ian)
[10:15:36] 	Splarka: LOL
[10:18:00] 	{{nobodycares}}
[10:19:25] 	03(mod) File links on image description page should be in alphanumeric (alphabetic) order - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=12532  (10greg.ebdon)
[10:21:52] 	:-D
[10:22:20] 	say domas, would you happen to know why it is so hard to purge a gen=js even after MediaWiki:Common.js has been edited?
[10:22:40] 	*shrug*
[10:22:47] 	(and some other user or MediaWiki JS reported to be hard to purge too)
[10:23:34] 	on en.wt, c*nnel edited the JS, I refreshed, hard refreshed, cleared cache, purged, and even logged in and redid all that to no avail, I had to go to prefs and [x] disable caching and then re-refresh to see the change 
[10:24:01] 	03midom * r29725 10/trunk/phase3/includes/ (4 files): 
[10:24:01] 	* Shortcutted Title::userCanRead() for public wikis
[10:24:01] 	* Moved redirects code in OutputPage before skins and ajax stuff, to handle likely scenario better
[10:24:01] 	* Stripped off lots of fat (like... 30-50% of execution time) from Special:Randompage - mwahaha... ha... hahaha...
[10:24:15] 	;-D
[10:25:20] 	how can i tell from javascript if a site has api.php ?
[10:25:30] 	oh hey domas, per Special:Random, would it delay execution time to pass on any non-title parameters given to it? like say, &action=render or timtest=newpp ?
[10:26:09] 	Splarka: special:random now cares about nothing
[10:26:10] 	so say, going to index.php?title=Special:Random would 302 to index.php&title=Whatever&otherparameters ?
[10:26:29] 	see, it used to build the page, set headers, etc, then just send a redirect
[10:26:33] 	now it just sends a redirect
[10:26:33] 	;-)
[10:27:58] 	will you be happy and relax and take a break, when every page load is down to 1ns?
[10:28:15] 	;-D
[10:28:26] 	.00000001ns
[10:29:00] 	hehehe
[10:29:05] 	Special:Random makes 5% of backend hits
[10:29:09] 	archivist: keep in mind that the .00000001ns itself is because some other processes keep the CPU busy!
[10:29:09] 	it had to be optimized!!!!! :)
[10:29:47] 	still, I have PF and OH 
[10:29:51] 	domas: if ($wgGroupPermissions['*']['read']) { return true; }    so what if there is another line of premission later which prevents a special group of people from reading?
[10:29:52] 	that don't allow me to sleep well
[10:29:54] *Splarka 	tempts domas with a lack of uncachable cookies, for parameter passing to Random
[10:30:07] 	Hojjat: that should never happen
[10:30:25] 	logically yes but practically... ?
[10:30:32] 	does anybody know much about api.php ?
[10:30:33] 	Splarka: could be possible, in theory
[10:30:47] 	Hojjat: I did that check in other place already
[10:30:51] 	it got certified as safe ;-)
[10:30:53] 	hippietrail: I know some! ;)
[10:31:03] 	domas: good
[10:31:25] 	Hojjat: RawPage is using it for caching
[10:31:27] 	also. Special:Random/:all for all namespaces, and Special:Random/:talk and Special:Random/:subject
[10:31:39] 	Splarka: I'm eliminating that namespace random feature
[10:31:42] 	and Special:Random/:funny Special:Random/:inoffensive Special:Random/:nudity
[10:31:48] 	:-)
[10:31:59] 	well, how can one get to a random template?
[10:32:01] 	:)
[10:32:07] 	heh
[10:32:25] 	ahh well, can still randomly get to pages via &curid=
[10:32:58] 	some of features live there cause nobody uses them
[10:32:59] 	;-D
[10:33:09] 	once they start using, I'm tempted to block 
[10:33:17] 	Hojjat: i want to detect in javascript whether api.php exists before i start sending off queries
[10:33:43] 	Hojjat: ideally i think there should be a wgHasApi variable or somesuch
[10:34:22] 	hippietrail: the "workaround" is to send an XMLHTTP request for api.php and see if the returned value is a 404 or what
[10:35:01] 	hippietrail: but no, we don't have a wgSomething JS variable in the code which shows API is available or not
[10:35:22] 	Hojjat: bandwidth! ick - and since a js extension can't retain such info between page views that's not very good
[10:35:37] 	i'll file a feature request
[10:35:42] *Hojjat 	agrees
[10:37:11] 	hippietrail: is this extension not meant to execute exclusively locally?
[10:37:52] 	Splarka: well i would like the code to work nicely as possible if people were to try it on various wikis
[10:38:01] 	I mean, it seems like you could check $wgEnableAPI and adjust the content of the .js
[10:38:05] 	oh, it is a pure js extension?
[10:38:12] 	nm
[10:38:45] 	you'll still have to do something like...
[10:38:46] 	Splarka: in this case it's a js extension that does several ajax lookups but could optimize them into a single lookup if it knows api.php exists
[10:39:57] 	if(!window.wgEnableAPI && !window.wgAssumeAPI) alert('This older install of MediaWiki does not allow my script to tell if API is enabled, please add var wgAssumeAPI = true; to your site JS')
[10:40:12] 	(or, yknow, not so annoying)
[10:40:42] 	better than an ajax call anyways
[10:41:03] 	wgEnableAPI already exists?
[10:41:32] 	ok, back to work
[10:41:41] 	I kind of promised myself I'm on wikivacation this week
[10:41:48] 	but had few pending commits since yesterday
[10:41:49] 	:)
[10:42:02] 	hippietrail: nope, but if they have older mediawiki, they could just declare it to override the check
[10:42:21] 	and if they do, but have api disabled, any problems are purely their fault ^_^
[10:42:34] 	here
[10:43:58] 	hippietrail: would this be the type of script enabled site-wide or in user JS/gadgets?
[10:44:07] 	i think i'll file an rfe for a wgEnableAPI or somesuch
[10:44:21] 	it's for wiktionary where we so far have not embraced gadgets
[10:44:35] 	well, yes, an rfe would be good
[10:44:50] 	but you mentioned "i would like the code to work nicely as possible if people were to try it on various wikis"
[10:44:58] 	most of the ones they'd try it on wouldn't be running trunk for a while
[10:45:00] 	it's actually for a function which is semantically library code. emulates wfMsg and wfMsgForContent as well as possible
[10:45:03] 	so a backup to that might be good
[10:45:39] 	in that case my backup plan would just be to fall back to the multiple ajax hits code i think
[10:46:04] 	ahh k
[10:46:06] 	messy!
[10:46:32] 	i would prefer if api.php could just call those functions for me!
[10:46:53] 	but a wg var exposed to js seems like generally a useful idea
[10:52:15] 	03(NEW) A JavaScript constant to declare whether api. php is available - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=12622 15enhancement; normal; MediaWiki: API; (hippytrail)
[10:58:50] 	domas|wikivacati: ouch. Gloating in commit messages ;)
[10:59:13] 	mwahahaha
[11:00:41] 	domas|wikivacati: oh well, you're entitled to your holiday fun, I guesss.
[11:00:58] 	Splarka: was fixing a reported bug that was already partially resolved.
[11:01:15] 	siebrand: come on, optimized special:random
[11:01:18] 	what can be better
[11:01:19] 	;-D
[11:02:17] 	sie: but there are other examples of miscase I have seen in other functions, so you should finish ^_^
[11:02:29] 	^in other functions' names
[11:03:08] 	as well as some cases of a call to functionname() having no space and later a call to functionname () having a space...
[11:03:25] 	you'd think a million monkeys typed this stuff ^_^
[11:06:21] 	03(mod) Kazakh message updates - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=12145  +comment (10alefzet)
[11:07:05] 	04(REOPENED) Kazakh message updates - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=12145  (10alefzet)
[11:09:52] 	03(mod) Kazakh i18n updates for extensions - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=7971  +comment (10alefzet)
[11:11:53] 	04(REOPENED) Kazakh i18n updates for extensions - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=7971  (10alefzet)
[11:13:27] 	meh
[11:13:30] 	I have that other thing to fix
[11:13:36] 	which is... Special:Search redirects
[11:13:42] 	:-)
[11:13:56] 	so much for your vacation
[11:14:09] 	03raymond * r29726 10/trunk/phase3/languages/messages/MessagesDe.php: 
[11:14:09] 	* Updates German
[11:14:09] 	Patch by Tim 'avatar' Bartel
[11:14:14] 	.. must... resist...
[11:14:18] 	... temptation...
[11:14:40] 	search is making calamari out of your squid
[11:14:41] 	mmmm
[11:15:27] 	if anyone wants, I have few gigs of profile traces available :)
[11:23:21] 	Splarka: I guess source code prettifiers exist? (have no idea)
[11:23:33] *Splarka 	shrugs ^_^
[11:23:43] 	Splarka: maybe a with a random comment generator plug-in? *huge grin*
[11:25:47] 	03(mod) Add a tag for inline math - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=12223  (10psychomessiah)
[11:40:44] 	03grondin * r29727 10/trunk/extensions/ContactPage/ (ContactPage.i18n.vo.php ContactPage.i18n.vo.php): svn:eol-style = native applied
[11:41:59] 	03thomasv * r29728 10/trunk/extensions/ProofreadPage/ProofreadPage.php: MsfForContent
[11:45:55] 	i wish svn:eol-style=native was the default
[11:58:03] 	03grondin * r29729 10/trunk/extensions/ContactPage/ (ContactPage.i18n.vo.php ContactPage.i18n.vo.php): Set fil to eol-native
[12:01:35] 	03siebrand * r29730 10/trunk/phase3/languages/messages/ (8 files): Localisation updates for core messages from Betawiki (2008-01-14 12:52 CET)
[12:03:51] 	03raymond * r29731 10/trunk/ (4 files in 4 dirs): * Updates German
[12:07:43] 	03siebrand * r29732 10/trunk/extensions/ (17 files in 17 dirs): Localisation updates for extension messages from Betawiki (2008-01-14 12:52 CET)
[12:10:12] 	03thomasv * r29733 10/trunk/extensions/ProofreadPage/proofread_index.js: insert backslask to ensure correct parsing of the template
[12:14:47] 	domas|wikivacati: you broke special random, I think.
[12:14:59] 	domas|wikivacati: http://translatewiki.net/wiki/Special:Specialpages?uselang=nl
[12:15:18] 	top of page shows "<>" linked to Special:Random
[12:16:32] 	heeeee
[12:16:45] 	which call has to be done there...
[12:17:05] 	domas|wikivacati: wfMsg() probably.
[12:17:09] 	naaah
[12:17:15] 	it asks SpecialPage for it's name
[12:17:37] 	domas|wikivacati: oh, getName() IIRC
[12:18:39] 	domas|wikivacati: me wrong. SpecialPage::getDescription()
[12:26:27] 	03(NEW) serve SVG where supported by UA, not PNG - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=12623 normal; normal; Wikimedia: Downloads; (j.chetwynd)
[12:28:57] 	wt
[12:28:57] 	f
[12:29:02] 	it shows Special:Random now
[12:29:07] 	as if it would be fetching -url
[12:30:58] 	03(NEW) Cleanup for Special:Randompage and Special: Randomredirect after r29725 - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=12624 normal; normal; MediaWiki: Special pages; (alex.emsenhuber)
[12:38:45] 	domas|wikivacati: the bug above :)
[12:38:56] 	haha, nice, thanks
[12:39:00] 	reviewed it already
[12:42:34] 	03gri6507 * r29734 10/trunk/extensions/TodoTasks/ (SpecialTaskList.i18n.php SpecialTaskList_body.php): Added a new i18n message for the body of the notification email.
[12:45:59] 	ialex: does this fix special:specialpages problem? :))
[12:46:44] 	domas|wikivacati: yes
[12:47:27] 	03rotem * r29735 10/trunk/ (2 files in 2 dirs): Update for he.
[12:50:33] 	03midom * r29736 10/trunk/phase3/includes/ (4 files): 
[12:50:33] 	Cleanup for Special:Randompage and Special: Randomredirect after r29725 - http://bugzilla.wikimedia.org/show_bug.cgi?id=12624
[12:50:33] 	Thanks to: alex.emsenhuber
[12:50:48] 	03(FIXED) Cleanup for Special:Randompage and Special: Randomredirect after r29725 - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=12624  +comment (10domas.mituzas)
[12:58:01] 	03gri6507 * r29737 10/trunk/extensions/Tooltip/Tooltip.php: Cleaned up a warning about Undefined variable: output
[12:59:20] *AaronSchulz 	waits for domas to sync :)
[13:01:05] 	hmm I see domas has been in the speed again :O
[13:02:34] 	domas: updating my cache for .js pages seems not to work :(
[13:32:07] 	hi.
[13:32:41] 	i have a strange problem since upgrade to 1.11.0. if i open http://do.main/Spezial:Anmelden, i get the login dialog. the url is rewritten by apache to http://do.main/index.php?title=Spezial:Anmelden.
[13:32:56] 	however if i open the latter url directly, i only get a error because mediawiki looks for the Page "Index.php".
[13:33:20] 	03siebrand * r29738 10/trunk/extensions/TodoTasks/SpecialTaskList.i18n.php: Update new message
[13:34:23] 	Oye, first rule of rewrite rules is don't hide files with them!
[13:34:30] 	kju: what is your $wgArticlePath ?
[13:34:32] 	amidaniel: rewrite is not the reason.
[13:35:23] 	i have replace the index.php with a call to phpinfo(). the request is mainly the same with the difference in the variables _SERVER["REQUEST_URI"], _SERVER["SCRIPT_NAME"], _SERVER["PHP_SELF"]
[13:35:41] 	O.o
[13:35:46] 	$wgArticlePath      = "/$1";
[13:35:56] 	kju: the cause _is_ the rewrite rules.  this changed in 1.11
[13:36:15] 	kju: putting articles in / no longer works unless you set a configuration option that might not exist (i don't remember offhand)
[13:36:20] 	have i missed something in the upgrade docs?
[13:36:53] 	flyingparchment: Well, it works. But you have to make explicit exceptions for index.php, etc. AFAIK at least
[13:37:12] 	amidaniel: i have those exception in my rewrite rules.
[13:37:24] 	amidaniel: no.  in 1.11, the logic changed to work around an Apache 2 bug/feature.  if the request uri matches $wgArticlePath, the article name is looked up that way
[13:37:34] 	amidaniel: if the article path is /$1, this is _always_ the case
[13:37:44] 	flyingparchment: so how to fix? changing the URLs is not an option.
[13:38:00] 	flyingparchment: Ah, okay.
[13:38:04] 	and why isn't there a warning in the docs?
[13:38:06] 	kju: moment
[13:38:47] 	http://www.mediawiki.org/wiki/Manual:Short_URL/Page_title_--_Working_method_with_mediawiki-1.11  ?
[13:38:59] 	$wgUsePathTitle is what you're looking for
[13:39:23] 	amidaniel: hmm, regarding to the page mentionend by splarka its $wgUsePathInfo = false;
[13:39:47] 	Erm .. $wgUsePathInfo yes
[13:39:47] 	Splarka: that page is neither linked from the release notes nor from the upgrade docs.
[13:39:59] 	kju: google knew about it though
[13:40:07] 	I googled for:
[13:40:09] 	site:mediawiki.org 1.11 $wgArticlePath "/$1"
[13:40:10] 	kju: That's nothing *officialy* mediawiki
[13:40:13] 	putting articles in / has never been supported, it's a shame people do that (and that mediawiki.org tells them it's possible)
[13:40:21] 	heh
[13:40:24] 	flyingparchment: Agreed :)
[13:41:02] 	Splarka: if you don't know about the reason of the error, those search terms are not that obvious.
[13:41:20] 	true
[13:41:36] 	i tried to debug this myself last night but gave up after a hour.
[13:41:58] 	I could guess though from flyingparchment and amidaniel that there'd be a page about it related to 1.11
[13:42:12] 	but, ok, thanks, works for now.
[13:42:20] 	maybe we will change the urls somewhen in the future but not now.
[13:42:21] 	and that from their assertations that the articlepath search above would be fruitful, etc etc
[13:42:46] 	i think there really should be a warning in the upgrade docs. the guy on the page mentioned by you states he spent hours debugging this as well.
[13:44:00] 	Well, $wgUsePathInfo is in the release notes
[13:44:39] 	not in http://www.mediawiki.org/wiki/Release_notes/1.11
[13:45:10] 	Hmm .. actually it's not.
[13:45:21] 	Was this introduced in 1.11 or 1.10? Seems like it's been around for a while
[13:45:30] 	i upgrade from 1.10.0
[13:45:36] 	and with the old version it worked.
[13:45:50] 	amidaniel: 1.11
[13:46:03] 	Oh .. hmm ... I can find no mention in the release notes
[13:46:26] 	ok, so it isn't my fault.
[13:46:49] 	it's not your fault you believed someone who said $wgArticlePath = "/$1"  is a good idea
[13:46:52] 	(that person lied to you)
[13:47:08] 	Now, now, don't be too harsh flyingparchment :P
[13:47:20] 	i'm not sure if someone told me, but i think is there isn't a warning neither that one shouldn't do this.
[13:47:33] 	amidaniel: if i start being harsh on people who tell other people things like that, maybe they'll stop doing it
[13:47:38] 	ialex: can you review every commit I make and fix errors for me? :)
[13:47:54] 	domas|wikivacati: :D
[13:47:55] 	ialex: I could get commit rights for you arrangedĄ
[13:47:56] 	!
[13:47:56] --mwbot--	I don't know anything about "". You might try: !centralauth !config !csshideheader !errors !extensions !feeds !help !highlight !imagelink !info !networkauth !package !rewriteproblems !rights !shell !smtp !subversion !test !threads !wikibugs 
[13:47:57] 	:)
[13:48:03] 	kju: there was.  people keep removing it from the manual.  it's annoying.
[13:48:03] 	damn, it would be so nice to have personal bug fixer
[13:48:06] 	:)
[13:48:48] 	^^
[13:49:00] 	03thomasv * r29739 10/trunk/extensions/ProofreadPage/proofread_index.js: use \{\{\!\}\} instead of pipe inside parameters
[13:49:01] 	mmm ... wikivicati ... sounds like a scrumptuous pastry of some kind :))
[13:51:08] 	03vasilievvv * r29740 10/trunk/phase3/ (docs/hooks.txt includes/Title.php): * Introduce getUserPermissionsErrorsExpensive hook
[13:51:18] 	flyingparchment: now i guess a warning combined with a hint that those URLs are not supported in the upgrade docs might make sense.
[13:52:18] 	03vasilievvv * r29741 10/trunk/extensions/TitleBlacklist/ (TitleBlacklist.list.php TitleBlacklist.php): 
[13:52:18] 	* Use this (see prev commit) hook in the TitleBlacklist
[13:52:18] 	* Fix profiling
[13:53:14] 	03(mod) Kazakh message updates - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=12145  +comment (10alefzet)
[13:57:53] 	03vasilievvv * r29742 10/trunk/phase3/includes/Title.php: Revert accidentally commited image moving stuff
[13:58:50] 	hello
[13:59:17] 	Hi.
[14:00:20] 	i have set wgAutoConfirmAge = 3600*24; and $wgAutoConfirmCount = 10; but the users dont get confirmed after 24 hours and 10 edits. has anybody an idea?
[14:00:49] 	i set this after the users signed up
[14:01:00] 	does it only work with new users?
[14:02:21] 	tnx all.
[14:04:51] 	03siebrand * r29743 10/trunk/extensions/cldr/LocalNamesEn.php: Add local names for gag, njo
[14:09:02] 	03vasilievvv * r29744 10/trunk/extensions/TitleBlacklist/TitleBlacklist.hooks.php: * Delay message loading
[14:10:12] 	Uhm, does anyone know why people have reported to have seen my Bot "walking" without been hidden? :)
[14:11:29] 	03raymond * r29745 10/trunk/extensions/Renameuser/ (3 files): Use {{PLURAL}} for log entry
[14:11:32] 	Filnik: there was a bug
[14:12:03] 	VasilievVV: but is the bots now hidden or not?
[14:12:08] 	(and if not, why?)
[14:12:18] 	No ideas. You can check
[14:12:53] 	VasilievVV: ok
[14:22:45] 	03raymond * r29746 10/trunk/phase3/languages/ (3 files in 2 dirs): 
[14:22:45] 	Localisation updates Kazakh translations
[14:22:45] 	(bug 12145) Patch by AlefZet
[14:23:11] 	03(FIXED) Kazakh message updates - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=12145  +comment (10raimond.spekking)
[14:24:10] 	Hm... May I ask someone to sysop me on mediawiki.org site?
[14:31:42] 	howdy everyone - is there an easy way to restrict the talk/discussion tab to authenticated users only? .. all other pages are world readable
[14:32:42] 	Talk pages reading:
[14:32:47] 	03raymond * r29747 10/trunk/extensions/DismissableSiteNotice/DismissableSiteNotice.i18n.php: 
[14:32:47] 	Localisation updates Kazakh extension translations
[14:32:47] 	(bug 7971) Patch by AlefZet
[14:33:19] 	reading and editing of talk pages needs to be restricted to authenticated users
[14:33:35] 	!secrets | Axman
[14:33:35] --mwbot--	Axman: MediaWiki was not designed with read-restrictions in mind, and may not provide air-tight protection against unauthorized access. We will not be held responsible should private information, such as a password or bank details, be leaked, leading to loss of funds or one's job.
[14:33:55] 	VasilievVV,  here you go :-)
[14:34:08] 	Simetrical: as you can see, Raymond_ is also helping you to filter "Localisation updates" now ;)
[14:34:19] 	Bdk[a]: thanks :)
[14:34:20] 	:)
[14:34:33] 	I can go without the smartass comments - i'm well aware of how mediawiki works
[14:35:46] 	03(FIXED) Kazakh i18n updates for extensions - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=7971  +comment (10raimond.spekking)
[14:37:10] 	03(mod) Warning got in includes/Title.php - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=11801  +comment (10roan.kattouw)
[14:37:22] 	Axman: :( I'm not 
[14:37:47] 	Axman: anyway, there're few hooks, where you can check for namespace
[14:38:02] 	Axman: I guess hooking into userCan and having %2 check on namespace can help you
[14:38:06] 	03(mod) Add Create permission to Protectedpages - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=12583  (10roan.kattouw)
[14:38:17] 	though of course, skin will have troubles
[14:38:19] 	:)
[14:39:55] 	03(mod) SelectCategory: selecting no categories causes a PHP warning - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=9847  (10N/A)
[14:40:37] 	I want to change the links in the navigation box? what should I be editing?
[14:41:45] 	03(mod) It's possible to create new empty pages using {{subst:}}, etc. - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=12621  +comment (10roan.kattouw)
[14:42:30] 	03(mod) A JavaScript constant to declare whether api.php is available - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=12622  +comment (10roan.kattouw)
[14:43:30] 	10domas|wikivacati: i'll take a look .. thanks01
[14:49:15] 	So I want to change the links in the left hand navigation box. How can I do that?
[14:50:10] 	10as sysop, goto yourwiki.com/index.php/MediaWiki:Sidebar
[14:50:59] 	Axman, that works
[14:51:40] 	another question. I lost the password for the sysop account, since I've been user another created user. How can I change the password for the sysop
[14:51:44] 	I have access to the db
[14:51:48] 	hello, all. I'd like to be emailed the last day's Recent Changes for some small mediawiki sites I run. Any recommendations ?
[14:51:50] 	but that field is a blob
[14:52:09] 	It's in the local network and the email is not set up.
[14:52:15] 	(recommendations how to implement it, that is, or if there's a program that does this already)
[14:52:44] 	10foobix: update user set user_password = md5(concat(user_id,’-',md5(’new_password’))) where user_name=someone
[14:53:44] 	Axman, sweet. Thanks
[14:54:08] 	10foobix: just fill in the blanks and that should work like a champ
[14:55:00] 	hrm. is there something equivalent to {{Special:Newimages/8}} where I can customize the output? I'd like to shrink it's display, remove captions, etc.
[14:58:01] 	hrm. i wonder if DPL could somehow get new images...
[15:02:51] 	03raymond * r29748 10/trunk/phase3/ (3 files in 3 dirs): 
[15:02:51] 	Localisation updates. Introduce formal address of German ("Sie")
[15:02:51] 	See next commit too.
[15:04:23] 	03(NEW) Lithuanian Wiktionary project need DynamicPageList extension - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=12625 15enhancement; normal; MediaWiki extensions: DynamicPageList2; (vpovilaitis)
[15:05:23] 	03(mod) Lithuanian Wiktionary project need DynamicPageList extension - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=12625  (10raimond.spekking)
[15:05:50] 	LIES
[15:07:44] 	centosian: There is an RSS feed
[15:07:54] 	Go to the recent changes page and click the RSS link in the sidebar
[15:07:56] 	RoanKattouw: hello
[15:08:07] 	RoanKattouw: yes, but I want it emailed to me in an email-program-readable way
[15:08:08] 	Then get an RSS reader to read it (see Wikipedia article on RSS reader)
[15:08:23] 	centosian: There are e-mail programs (like Thunderbird) that can also handle RSS
[15:08:26] 	VasilievVV: Hi
[15:08:34] 	can I convert the CSS to something HTML or ASCII? I'm using Eudora
[15:09:04] 	sorry, RSS
[15:09:18] 	!passwordsql | foobix
[15:09:18] --mwbot--	foobix: UPDATE user SET user_password=MD5(CONCAT("1-",MD5("password"))) WHERE user_id=1;  This will reset the password for the user with user_id 1 (the default admin account, usually WikiSysop) to "password".  If you do not know the name of the user account, execute: SELECT user_name FROM user WHERE user_id=1;
[15:09:42] 	centosian: No. Alternatively, you could watch every page and enable e-mail notification
[15:09:48] 	Maybe there's an extension for that
[15:10:03] 	foobix: or better use maintenance/changePassword.php
[15:10:04] 	RoanKattouw, thanks. 
[15:10:06] 	!extensions | centosian
[15:10:06] --mwbot--	centosian: MediaWiki has been built so it can easily be customized by adding extensions. This is usually a simple process. See  for instructions to install extensions, as well as for writing them. See  for an overview of known extensions.
[15:10:09] 	Second link
[15:10:16] 	RoanKattouw: would that miss new page creations? Also, wouldn't that send me one email per change?
[15:10:40] 	What's the format of the /MediaWiki:Sidebar page?
[15:10:44] 	centosian: No, it wouldn't miss page creations, and no, it would send you only one e-mail until you visit the page again
[15:10:47] 	I add links, but they dissapear
[15:10:54] 	That's weird
[15:11:01] 	It's some foo|bar type of value, that I don't now that it maps to.
[15:11:07] 	!sidebar | foobix
[15:11:07] --mwbot--	foobix: To edit the navigation menu on the left, edit [[MediaWiki:Sidebar]] using its special syntax. For more details, see .
[15:11:12] 	foobix: Foo maps to [[MediaWiki:Foo]]
[15:11:44] 	ah
[15:12:43] 	Axman: There is also something about namespace protection somewhere
[15:12:51] 	!access | Axman
[15:12:51] --mwbot--	Axman: For information on customizing user access, see . For common examples of restricting access using both rights and extensions, see .
[15:13:00] 	Second link, look through the TOC
[15:13:29] 	10RoanKattouw: ive been reading a few things, i'll take a look there as well
[15:21:07] 	03raymond * r29749 10/trunk/phase3/languages/messages/MessagesDe_formal.php: 
[15:21:07] 	Localisation updates.
[15:21:07] 	Set fallback to 'de'. Specific formal address messages follows soon.
[15:25:47] 	Raymond_: de-formal?
[15:26:08] 	VasilievVV: yeah, them Germans ;)
[15:26:15] 	VasilievVV: German has an informal (du) and a formal (Sie) word for "you" 
[15:26:22] 	yes
[15:26:24] 	As does Dutch BTW, but we don't mind that much
[15:26:30] 	The same with Russian
[15:26:43] 	'nl' has a formal localisation of mw.
[15:26:48] 	a lot of German companies are asking for a formal address in its wikis
[15:27:54] 	But writing interface messages in Russian with informal form of "you" looks pretty silly
[15:28:16] 	siebrand: Didn't know that. Informal is the default, I assume
[15:28:34] 	RoanKattouw: nope. Always "u".
[15:28:40] 	Oh
[15:28:49] 	RoanKattouw: looks so nice you didn't even notice it ;)
[15:29:02] 	RoanKattouw: or maybe you use an English UI..
[15:29:04] 	hmm, no one who has an idea?
[15:29:05] 	Oh you mean everything is formal; do we even have a formal/informal distinction for MW i18n ?
[15:29:11] 	siebrand: I never use a Dutch UI
[15:29:15] 	It'll just confuse me
[15:29:39] 	RoanKattouw: ah, that explains. There's no distinction in 'nl' MediaWiki UI. It's all formal.
[15:29:43] 	The logo in my wiki is small. The navigation box is flushed agains the logo. It looks like there's a large box area where the user can click
[15:29:44] 	Oh ok
[15:29:55] 	I'm trying to have the nav flush up agains the logo
[15:29:56] 	foobix: URL?
[15:30:01] *VasilievVV 	uses russian UI on russian Wikipedia
[15:30:08] 	RoanKattouw, local network
[15:30:10] 	Oh
[15:30:27] *siebrand 	users Dutch everywhere. Just to be able to keep track of the localisation ;)
[15:30:29] 	I'm editing skins/MonoBook.php
[15:30:38] *RoanKattouw 	has always used the English UI at wikis and wouldn't know what namespaces are translated to in Dutch
[15:30:53] 	But it's not straight forward  where that link is comming from
[15:31:03] 	Hmm maybe I should try the Frisian locale some time
[15:31:07] 	I don't know if it's Javascript, or Css that's placing that where it is.
[15:31:17] 	I don't know shit about that either, sorry
[15:32:12] 	RoanKattouw: Frysian is currently being worked on. Finally making progress on a centralised localisation for that. Was quite difficult to get someone to do that.
[15:32:30] 	I can imagine
[15:32:37] 	I speak Frisian myself, but not natively
[15:32:42] 	RoanKattouw: y 	 Frysk 	 100.00% 	 0.00% 	 44.68% 	 0.00%
[15:32:50] 	What are those percentages?
[15:32:54] 	100% of most used messages available
[15:33:05] 	0% wikimedia extensions
[15:33:12] 	44% of all core messages
[15:33:18] 	0% of all extensions.
[15:33:27] 	RoanKattouw: source: http://translatewiki.net/wiki/Translating:Group_statistics
[15:33:37] 	Hmm so core msgs aren't even complete for fy
[15:33:55] 	RoanKattouw: they are for Saterfries :)
[15:34:09] 	Huh
[15:34:33] 	That's that funny language they speak in parts of Schleswig-Holstein, DE right?
[15:34:34] 	RoanKattouw: English:  stq 	 Seeltersk
[15:34:48] 	RoanKattouw: yep. 2000 living speakers, I think.
[15:34:51] 	lol
[15:35:10] 	There's also *another* variant of Frisian being spoken on some Wadden islands in Denmark
[15:35:33] 	I'll just stick to Kleifrysk, I guess (~500,000 living speakers)
[15:35:49] 	RoanKattouw: Hindi is not complete either... That's Bad (tm). 	 18.35% and there are some 300M+ speakers.
[15:35:58] 	Holy hell
[15:36:04] 	Yeah that's something you wanna get fixed
[15:36:08] 	RoanKattouw: yep.
[15:36:29] 	RoanKattouw: watch the mailing lists. We are ;)
[15:36:43] 	RoanKattouw: ran into some grant money.
[15:36:44] 	BTW do we also have separate localizations for things like fr_CA and fr_BE or do they just use fr_FR?
[15:37:18] 	RoanKattouw: there is Walloon. Possibly also for Canadian French, but not certain.
[15:37:53] 	It's kind of silly that Hindi is so incomplete, while we are setting up a Limburgish Wikiquote ^^
[15:38:42] 	RoanKattouw: Limburgish must have 100% and 100% Wikimedia extensions before it can be created. So depending on how hard the community wants that new wiki, it will indeed be a lot better than Hindi :)
[15:39:04] 	lol
[15:39:11] 	RoanKattouw:  Limburgs 	 100.00% 	 47.63% 	 100.00% 	 14.90%
[15:39:18] 	Hmm not bad
[15:39:26] 	100% core, 48% WM exts
[15:40:01] 	nope. That scores top 30 best supported languages
[15:40:26] 	Well I'm sure we have at least 30 langs that are nearly at 100-100-100-100 ?
[15:40:32] 	German, French, the biggies
[15:40:42] 	Well 30 is a lot
[15:40:50] 	RoanKattouw: ready for a deception? We have 4.
[15:40:53] 	lol
[15:40:56] 	EN, DE, FR and ?
[15:41:13] 	Arabic, English, French and Dutch
[15:41:24] 	Well I don't mind langs at 100-100-99-100 because someone introduced a new interface message like yesterday
[15:41:26] 	lol
[15:41:33] 	German isn't even there
[15:41:43] 	RoanKattouw: current milestones are at 99%, 90%, 90%, 65%
[15:41:55] 	Right
[15:41:57] 	8 languages qualify.
[15:42:04] 	^^
[15:42:17] 	And I thought MW was doing so well at l10n
[15:42:32] 	RoanKattouw: it is no.
[15:42:35] 	now
[15:42:50] 	Well not as well as I imagined it did ;)
[15:43:07] 	RoanKattouw: I wrote a report beginning of this year.
[15:43:11] 	hello
[15:43:24] 	how many languages have more than one reviewer? ;)
[15:43:50] 	RoanKattouw: now that we are organised, the WMF has clear policies and some money is flowing in localisation, we have a nice environment to make serious progress.
[15:43:59] 	siebrand: Yeah, I saw it, lot of text, read through it rather quickly
[15:44:03] 	Good
[15:44:05] 	flyingparchment: 50%+
[15:44:56] 	RoanKattouw: #mediawiki-i18n is also getting busier, we get ~3 new accounts/day on Betawiki. That's up from 3/week beginning of last year.
[15:45:15] 	RoanKattouw: having a Betawiki URL in the MediaWiki namespace helps.
[15:45:30] 	RoanKattouw: but only for those running 1.12alpha, yet.
[15:45:39] 	siebrand: I also see a lot more BetaWiki updates than, say, a month ago
[15:45:40] 	03(NEW) Remove watchlist entries directly from My Watchlist - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=12626 15enhancement; normal; MediaWiki: General/Unknown; (LSander153)
[15:46:05] 	Between midnight and 15:00 I actually saw like 4 l10n commits
[15:46:11] 	RoanKattouw: committing was a bitch. Have command line export to files now, so it's mainly review that takes the time.
[15:46:27] 	RoanKattouw: and Grondin is working on it.
[15:46:44] 	siebrand: yeah and some people review after commit too, I hope :)
[15:46:53] 	RoanKattouw: in the process of getting command line export for all extensions, too. ~15 more to go out of 125 or so.
[15:47:14] 	Nikerabbit: hmm, you do :)
[15:47:16] 	Right
[15:47:16] 	03(mod) Remove watchlist entries directly from My Watchlist - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=12626  +comment (10roan.kattouw)
[15:47:38] 	Nikerabbit: and I update Betawiki right after a commit for possible syntax errors, etc.
[15:47:39] 	siebrand: yeah, have to keep care of my (fi) minions :)
[15:47:58] 	there should be php -l in pre-commit hooks :o
[15:54:39] 	hi there
[15:57:43] 	would somebody have a link explaining the difference of usage of the slash (/) in page names in the main namespace and other namespaces ?
[15:58:24] 	zif: In the main namespace, subpages are enabled
[15:58:39] 	So Foo/Bar is a subpage of Foo and will display a link to Foo on top
[15:58:53] 	is there anyway to view the job queue, to see what's in it?
[15:58:54] 	This is only true for the main and User namespaces IIRC
[15:59:02] 	Morbus: If you have database access, yes
[15:59:09] 	I do.
[15:59:28] 	Well there is a table called job I think
[15:59:32] *RoanKattouw 	checks
[15:59:35] 	ayup. thanks.
[15:59:46] 	Morbus: http://www.mediawiki.org/wiki/Job_table
[16:00:06] 	RoanKattouw: hum, http://en.wikipedia.org/wiki/Wikipedia:Naming_conventions_(technical_restrictions)#Forward_slash says that subpages are disabled 
[16:00:11] 	hrm. 
[16:00:21] 	:Statistics shows my queue is 5, but the table is empty.
[16:00:43] 	Then statistics probably isn't up to date
[16:01:02] 	zif: Hmm. Well they are enabled in the User: namespace I think
[16:02:06] 	RoanKattouw: yeah, I think they talked about User: and Talk:
[16:02:19] 	hi, how can i make my friend admin too?
[16:02:19] 	RoanKattouw: but is there a way to enable it for main:, then ?
[16:02:29] 	(haven't check the code, yet)
[16:02:36] 	There is no difference, really, only the links on top of the page linking back to the "parent" pages
[16:02:56] 	You can still create [[Foo/Bar/Baz]], it just won't display Foo > Bar > Baz on top
[16:03:28] 	hum, that's the only difference? hum, thanks for the info
[16:04:21] 	03raymond * r29750 10/trunk/phase3/languages/messages/MessagesDe_formal.php: Localisation updates. Specific German messages for formal address only.
[16:04:36] 	sorry, but i find nothing to make an user to admin :(
[16:04:54] 	T-Vairus: On your own wiki or on a Wikimedia wiki?
[16:06:39] 	my own wiki :)
[16:08:02] 	03(WONTFIX) Remove watchlist entries directly from My Watchlist - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=12626  (10roan.kattouw)
[16:08:07] 	T-Vairus: Special:Userrights
[16:08:56] 	thanks :) - what is higher? sysop or buerocrats?
[16:09:48] 	T-Vairus: Bureaucrats can do everything sysops can, but they can also promote and demote users
[16:10:05] 	that's not true
[16:10:13] 	oO
[16:10:17] 	bureaucrat's can only change user rights
[16:10:27] 	Oh yeah right
[16:10:33] 	oh i understands - thanks a lot dudes :)
[16:10:43] 	because it's mostly together with sysop rights, it seems that bureaucrats can do everything
[16:10:44] 	You probably wanna give someone *both* sysop and bcrat in that case
[16:11:00] 	Yeah, I remembered incorrectly. Pretty much every bcrat is also a sysop
[16:41:55] 	Hi, how can I set up my wiki that all external links are opened in new windows?
[16:42:28] 	MONK3000: I think you'd have to add something to your common.css, don't know exactly what and where though
[16:42:30] 	Anyone?
[16:43:41] 	that's just a style sheet
[16:44:02] 	I know, but can't you set the target attribute there?
[16:44:09] 	If not, it can't be done
[16:44:43] 	somewhere it must also be handled that the blue arrow is shown next to external links. i'd like to remove that aswell
[16:45:17] 	MONK3000: That blue arrow is an image somewhere in the skins/ directory. You can replace it by an empty image sized 0x0 pixels
[16:46:21] 	i think i deleted it a year ago or so, stil IE and Opera are displaying the mofo :)
[16:46:30] 	i want to handle it properly this time#
[16:46:45] 	hsd   K h d d sdjsd sak  ska d  sd kjlk a;d kd saduf aduye a; ueh as  a.hfyfa e        adhka x
[16:47:25] 	MONK3000: It's possible that the image was overwritten by an upgrade
[16:49:20] 	trying the 0x0px now
[16:50:54] 	but my point is it must be handled somewhere to call the image. i expected that could add the target there as well
[16:51:30] 	MONK3000: the arrow iummage *can* be disabled in common.css
[16:51:37] 	look at the plainlinks class, it does just that
[16:51:51] 	no need to delete or replace images
[16:52:22] 	that's the proper way then ok ;9
[16:52:35] 	as for opening external links in a new window: this was decided to be a Bad Thing in principle, and is thus not suppoerted. I think there are several extensions that let you do it, but they all use a special syntax, afaik
[16:52:57] 	bad decision tztz
[16:52:59] 	MONK3000: yes, that's the proper way. it's added using a css style, so it should be removed the same way
[16:53:33] 	MONK3000: i actually support that decision. where a page opens should be left to the browser / user. i annoys me a lot when websites pop up new windows.
[16:53:43] 	(actually, i have ff set to never ever do that)
[16:53:46] 	O TE EENKSHA PIDHIN
[16:54:01] 	HFKSD KF DF;JG  HLSDFUR SDI LRI LADS RUSD LJG ROL LAR;;RY GJRUSI SJ  R
[16:55:56] 	hmh .plainlinks only defines bg-color and the padding (in common.css and monobook/main.css). maybe i removed it already? but still IE and Opera call the image
[16:56:57] 	MONK3000: the thingy is a bg image, right-alligned, so it shows up just in the padding. setting bg-color and no padding will remove it, iirc.
[16:57:06] 	You're looking for a rule ".plainlinks img" or similar
[16:57:17] 	n/m me
[16:57:20] 	Duesentrieb: Not every Wiki is used as a normal website where you raed things. my site lives from opening links
[16:57:23] 	Dashiva: no  is used
[16:57:30] 	Dashiva: -plainlinks a, if anything
[16:57:46] 	i have bodyContent .plainlinks a {	background: none !important;	padding: 0 !important; }
[16:57:50] 	MONK3000: links, yes. windows?! ugh! 
[16:58:04] 	MONK3000: google lives from opening links. do they pop up windows?
[16:58:08] 	flyingparchment: Could you kick DiSiRe please?
[16:58:15] 	in real browsers windows are just tabs ;)
[16:58:23] 	Yes
[16:58:28] 	MONK3000: not per default.
[16:58:44] 	MONK3000: all browser i know will actually make a new window, per default
[16:58:50] 	which sucks 
[16:58:58] 	target=newwindow works in opera and firefox
[16:58:58] 	let the user decide
[16:59:17] 	*as new tabs per default
[16:59:19] 	err. that should be target=blank
[16:59:28] 	and that *will* open a new *window* in ff, per default
[16:59:51] 	anyway, as i said, let users decide. 
[16:59:56] 	opera is very aggressive about opening tabs instead of windows by default, because its tabs are also windows
[17:00:15] 	anyway, forcing a new tab or a new window is almost always wrong :)
[17:00:16] 	ah sorry i use onclick="window.open('...')
[17:00:33] 	about the css style: yes, that will do the trick, if you replace .plainlinks with .bodyContent
[17:00:44] 	MONK3000: using js is even worse.
[17:01:02] 	using js is great, i can disable js and open links where i want them
[17:01:05] 	it can't readily be overwritten by client config
[17:01:11] 	and it will not work for a lot of people
[17:01:24] 	works for my little php application i wrote though
[17:01:25] 	flyingparchment: hehe - unless you don't want to disable it
[17:01:34] 	MONK3000: works for *you*.
[17:01:42] 	do you ever find out for how many people using your site it doesn't?
[17:01:53] 	no.
[17:01:57] 	because they stop using your site,
[17:02:05] 	that is just for me. on my wiki i'd like to have a proper solution of cos
[17:02:16] 	all the blind people i know who use computers use lynx.  therefore, 100% of blind people use lynx, and none of them can use your website!
[17:03:08] 	MONK3000: the proper solution is: don't open new windows. and don't use JS for any central function. only for the iceing
[17:03:52] 	i just think that my site lives from the external links. every article includes an external link where the users browses to to get a download link. once on the other page i think it sucks to go back to my site
[17:04:21] 	again, consider google as an example. or any other decent search engine
[17:04:30] 	none of them pops up windows. for a good reason.
[17:04:39] 	hmh.
[17:05:06] 	that's absolutely correct. and they might be interested in it then me ;)
[17:05:11] 	*more
[17:05:52] 	anyway, gotta fix dinner
[17:05:53] 	bbl
[17:07:20] 	thanks
[17:07:50] 	AaronSchulz: PING
[17:08:07] 	POong, have to go in 2 min
[17:08:21] 	en-admins
[17:29:41] 	03(mod) Lithuanian Wiktionary project need DynamicPageList extension - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=12625  (10vpovilaitis)
[17:30:26] 	how can i make backups of my wikipedia - can i use mysqldumper?
[17:32:51] 	03mkroetzsch * r29751 10/trunk/extensions/SemanticMediaWiki/includes/SMW_QP_RSSlink.php: Be more careful when building RSS-Link (avoid illegal title errors if there are problems)
[17:33:21] 	03(mod) It's possible to create new empty pages using {{subst:}}, etc. - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=12621  (10Simetrical+wikibugs)
[17:33:32] 	T-Vairus, yes, that works.
[17:33:38] 	"de-formal" o_O
[17:34:16] 	Simetrical: New empty pages is really convenient when you're making an empty template for fallback
[17:34:17] 	Hello wiki fans...
[17:34:34] 	I have a question that I can not find a answer to on meta
[17:34:34] 	Dashiva, yes, I don't see why it shouldn't just be allowed altogether.
[17:34:47] 	Because it's a null edit? :)
[17:35:00] 	Is there a way to put an "Edit" button inside a table cell, when clicked only an edit page for that cell comes up?
[17:35:22] 	Please don't
[17:35:23] 	Dashiva, that's not why it's prevented, creating pages with no content is prevented explicitly.
[17:35:36] 	thanks simetrical :)
[17:36:00] 	Klemo, that works for sections.  Create a section inside the table if you want, with == header ==.
[17:36:25] 	yeah I was thinking that.. oh maybe a sub sub section so the is no lines...
[17:36:26] 	He also has to end the section inside the cell, otherwise the trailing table markup follows
[17:36:47] 	how do you end a section?
[17:36:48] 	Yes, well, nothing's perfect.  :)
[17:36:52] 	Klemo, start another section.
[17:36:53] 	I normaly just create another section
[17:37:42] 	maybe something to think about for the next version... Edit buttons inside table cells, added by CSS tag, kind of like sorting
[17:37:56] *Dashiva 	is confused
[17:38:38] 	didn't i just read a webpage that says you can add a tag to the header row that allows collum sorting?
[17:39:08] 	Klemo, yes, class=sortable.
[17:39:08] 	maybe there could be a tag for a cell that would allow editing of the contents of just that cell with out all of the "|" markup
[17:39:17] 	Maybe an extension.
[17:39:58] 	Might be better with a __SECTIONSEPERATOR__ or somesuch, I'd think
[17:40:03] 	But it's still evil
[17:41:03] 	why would it be evil?... My boss likes to keep spread sheets for our tasks... the Master Task Sheet is a mile long, and editing large tables in media wiki is pretty haynous
[17:41:26] 	so make it easy by allowing users to edit just a cell or row...
[17:41:52] 	any rate, its just a suggestion to make table editing less a pain in the arse...
[17:41:56] 	my 15 cents...
[17:42:55] 	wow
[17:43:04] 	adding === TEST === to a cell just renders text...
[17:43:22] 	03(NEW) pre-built search returns no matches - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=12627 major; normal; MediaWiki: Search; (filop)
[17:43:30] 	Klemo, it needs to start on its own line.
[17:44:13] 	hi
[17:44:34] 	well that kind of works...
[17:46:44] 	03(FIXED) Rollback on en.wiki - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=12534  +comment (10matthew.britton)
[17:49:02] 	03(NEW) Excess  tags are generated - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=12628 normal; high; MediaWiki: Page editing; (haridara)
[17:53:06] 	there is no other way to end a section other than putting another section?
[17:53:08] 	03mkroetzsch * r29752 10/trunk/extensions/SemanticMediaWiki/ (3 files in 3 dirs): Added Italian translation (by Davide Eynard & David Laniado)
[17:54:44] 	because, you if there was such a way to do that, there would be no need to have row/cell edit buttons..
[17:55:08] 	I could just wrap my cell content in a start and end section...
[17:55:16] 	Klemo, no, there's not.
[17:56:29] 	another thing which might make my life easier... is there CSS specific tags for each section/subsection edit button
[17:58:53] 	03rotem * r29753 10/trunk/extensions/Renameuser/SpecialRenameuser.i18n.php: Update for he.
[18:01:38] 	03mkroetzsch * r29754 10/trunk/extensions/SemanticMediaWiki/ (3 files in 2 dirs): Updated Chinese translations by Michael Chang
[18:01:52] 	04(REOPENED) Patch for SpecialUserrights. php to prevent logging of no changes (see URL) - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=12614  +comment (10matthew.britton)
[18:02:52] 	or are all the edit buttons linked to CSS class ( Which is what I would expect...)
[18:03:01] 	to one *
[18:03:43] 	03(mod) Patch for SpecialUserrights. php to prevent logging of no changes (see URL) - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=12614  (10Simetrical+wikibugs)
[18:03:46] 	03(ASSIGNED) Patch for SpecialUserrights. php to prevent logging of no changes (see URL) - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=12614  (10Simetrical+wikibugs)
[18:04:02] 	Klemo, the CSS class is editsection.
[18:04:46] 	03mkroetzsch * r29755 10/trunk/extensions/SemanticMediaWiki/languages/SMW_LanguageRu.php: Updated Russian language file by Dmitry Khoroshev.
[18:05:04] 	yeah I was thinking i could do something like === === Cell Contents ==== ==== .... then hide the sub section edit button...
[18:05:34] 	well ... hide the sub sub sub sub section edit button... lol
[18:06:23] 	I'm wondering how I can see what extensions wikipedia is using, because I copied some templates from wikipedia, and they're not working on my wiki
[18:08:58] 	Hi, I have a Mediawiki Site http://www.manipedia.eu/  , where I cant edit anymore existing pages. I have tried reinstalling Mediawiki (updating to latest version with no success). What is going wrong 
[18:10:09] 	When I try and do an edit, and then click submit, I am just brought back to the main index page, no changes done
[18:11:02] 	Kyrin: see [[Special:Version]] on Wikipedia, pay especial attention to ParserFunctions and other parser hooks
[18:11:12] 	Thanks much
[18:11:40] 	Anyone able to help, please? 
[18:11:53] 	It used to work till about a week ago?
[18:12:38] 	Petzlux did you change your LocalSettings file and change any of the permissions lines? or did you add any permission lines?
[18:12:56] 	03simetrical * r29756 10/trunk/phase3/ (RELEASE-NOTES includes/SpecialUserrights.php): (bug 12614) Prevent logging of null change on Userrights . . . for real this time.
[18:13:08] 	Not that I am aware off , I reinstalled and using the plain vanilla LocalSettings File that came with the update script
[18:13:17] 	kk
[18:13:24] 	still same problem
[18:13:36] 	Petzlux Have you tied emting your browser cache?
[18:13:49] 	emptying 
[18:14:17] 	Yes, no effect. Other Users trying to edit the wiki have reported same problem
[18:14:37] 	03(FIXED) Patch for SpecialUserrights. php to prevent logging of no changes (see URL) - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=12614  +comment (10Simetrical+wikibugs)
[18:15:09] 	petzlux: sorry, I don't know then... one of these wiki geniuses will figure it out I bet...
[18:15:39] 	you can try yourself on the site, the bizarre thing is it works for newly created pages, not for existing ones? wouldnt any problem be coming up regardless of if its a new or existing page? 
[18:15:59] 	Klemo:  thanks for your help
[18:16:11] 	petzlux, it works fine for me.  http://www.manipedia.eu/index.php/Category:SQL_Examples?action=history
[18:18:16] 	Simetrical:  can you try the following editing the following page, doesnt work for me: http://www.manipedia.eu/index.php/Keyboard_Shortcuts
[18:18:39] 	Works fine.
[18:19:01] 	Try logging out?
[18:20:13] 	Ups , got kicked out for some reason ?
[18:20:20] 	petzlux, try logging out?
[18:20:28] 	did I miss any messages? 
[18:20:35] 	[080114 13:19:39]  Works fine.
[18:20:35] 	[080114 13:20:01]  Try logging out?
[18:21:49] 	Simetrical:  Could this be a browser issue? Am using Firefox, can see the change you made to the page, but cant edit myself
[18:22:06] 	petzlux, I'm using Firefox too. Have you tried logging out?
[18:23:37] 	Simetrical:  Nope, no luck
[18:23:54] 	Interesting.
[18:24:00] 	You say you get redirected to the main page?
[18:24:04] 	yes
[18:24:09] 	Have you tried using a different browser?
[18:25:05] 	Simetrical: And then this page I just created, and did an edit, no problem
[18:25:07] 	http://www.manipedia.eu/index.php/Another_test
[18:25:59] 	Simetrical: Just tried IE , same problem 
[18:26:06] 	Quite odd.
[18:26:56] 	Tearing my hairs out here!!!
[18:29:52] 	Any idea why the edit page might come back with wpStarttime, wpEdittime, and the textarea empty?
[18:30:36] 	Any ideas????
[18:33:01] 	what's the recommended method for extending the login time on a mediawiki? is it a php.ini option or a mediawiki option ?
[18:35:31] 	login time?
[18:35:40] 	cookie expiration ?
[18:35:51] 	petzlux: Works for me
[18:36:01] 	Try clearing your MediaWiki cookies and/or logging out
[18:36:07] 	RoanKattouw_away:  what page ?
[18:36:13] 	Another test
[18:36:47] 	RoanKattouw:  that page works for me as well . please try to edit the Main Page 
[18:38:16] 	Weird, Main Page doesn't work for me
[18:38:40] 	thats what I  mean
[18:38:53] 	Cleared my Cookies, no effect
[18:40:19] 	That's weird, it suddenly broke for me
[18:40:53] 	RoanKattouw:  you might find that you can still create a new page and edit that one fine
[18:41:01] 	Weird
[18:41:08] 	I really don't know what could be going on here
[18:41:11] 	is there an easy way to do an internal link that goes right to the edit page of the article linked to?
[18:41:15] 	Have you checked the Apache error log?
[18:41:35] 	frieze: Well, easy is different, but here it is: [{{fullurl:Article name|action=edit}} link text]
[18:42:01] 	RoanKattouw:  no not yet
[18:42:53] 	03vasilievvv * r29757 10/trunk/phase3/ (RELEASE-NOTES includes/Skin.php): 
[18:42:53] 	* (bug 12536) User should be able to get MediaWiki version from any page
[18:42:53] 	* (bug 12622) A JavaScript constant to declare whether api.php is available
[18:43:31] 	RoanKattouw: thats the only two error messages : [Mon Jan 14 18:41:28 2008] [error] [client 83.244.143.154] File does not exist: /home/petzlux/public_html/404.shtml
[18:43:32] 	[Mon Jan 14 18:41:28 2008] [error] [client 83.244.143.154] File does not exist: /home/petzlux/public_html/favicon.ico
[18:43:47] 	Weird
[18:44:09] 	?
[18:44:12] 	oops
[18:44:19] 	stupid keyboard
[18:44:32] 	RoanKattouw: do you know if there is any plan to merge the apiedit-vodafone branch into the trunk anytime soon?
[18:44:32] 	is there a way to get a random number without using an extension?
[18:44:41] 	03vasilievvv * r29758 10/trunk/phase3/ (RELEASE-NOTES includes/AjaxFunctions.php): * Add caching to the AJAX search
[18:44:50] 	carl-m: That plan involves my free time, which is scarce right now
[18:45:07] 	I bet
[18:45:24] 	And unfortunately I'm looking to take more of it: http://pastebin.com/d3c7090ac
[18:46:14] 	lines 98 and 100 boggle my mind
[18:46:45] 	03vasilievvv * r29759 10/trunk/phase3/profileinfo.php: 
[18:46:45] 	Tweak profileinfo.php:
[18:46:45] 	* Added average calls per request column
[18:46:45] 	* Added time % column
[18:47:42] 	RoanKattouw:  could something to do with my hoster be to blame ?
[18:47:49] 	I don't know
[18:49:09] 	03vasilievvv * r29760 10/trunk/phase3/README: Update README
[18:52:56] 	RoanKattouw:  Im giving up for now, this is really depressing, thanks for the help anyway. 
[18:53:10] 	03vasilievvv * r29761 10/trunk/phase3/ (4 files in 2 dirs): 
[18:53:10] 	Autopromotion:
[18:53:10] 	* Add APCOND_INGROUPS
[18:53:10] 	* Fix comment
[18:53:10] 	* Don't use array_unique() since we have groups as *keys* of $wgAutopromote
[18:53:15] 	03(mod) Prefix index search and namespaces in Special:Withoutinterwiki - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=10049  (10Eugene.Zelenko)
[19:06:24] 	it is possible to make a pastebin within MW ?
[19:07:07] 	You could just have everyone who wants to pastebin make a new wiki page.
[19:08:01] 	no I just want a page within my wiki to act as a pastebin (because I could not find a pastebin with mIRC highlight and captcha to wrk) heh
[19:08:09] 	work*
[19:44:05] 	here is what I have discovered: if you try to get an edit token using POST, the wpStarttime and wpEdittime fields are not filled in, nor is the preview filled in. If you request an edit token with GET, they are all filled in coorectly. This seems like a bug to me.
[19:49:49] 	carl-m: why would get a token by a POST ?
[19:49:54] 	*you
[19:52:56] 	ialex: because my library was sending all queries by post, which should be safe. the real problem is that, instead of reporting an invalid request method, index.php returns half of the needed information when you post to it 
[19:53:16] 	it took me a long time to diagnose the problem
[19:57:07] 	carl-m: if you send a POST, that indicates you're already in a submission cycle
[19:57:30] 	Heh:
[19:57:32] 	!r 29760
[19:57:32] --mwbot--	http://svn.wikimedia.org/viewvc/mediawiki?view=rev&revision=29760
[19:57:52] 	:D
[19:58:12] 	brion: you seriously use the request method to tell what part of the cycle the request is supposed to be in? 
[19:58:21] 	Yeah, actually we do
[19:58:26] 	I agree that it sucks
[19:59:16] 	in a sane world, you only POST to things that should be POSTed to
[19:59:40] 	I realize that several mediawiki developers have had that opinion
[20:00:20] 	We should really look at action=edit versus action=submit
[20:01:33] 	this will do same problems
[20:02:02] 	No. Action=edit gets the edit form, action=submit submits it
[20:03:21] 	actually, action=edit is an alias for action=submit (only a session check more for action=submit)
[20:03:48] 	Well $wgRequest->getVal('action') is still there
[20:04:07] 	Although admittedly it sucks to hardcode 'edit' and 'submit' in multiple places
[20:05:29] 	brion, are you okay with r29705?
[20:05:31] 	!r 29705
[20:05:31] --mwbot--	http://svn.wikimedia.org/viewvc/mediawiki?view=rev&revision=29705
[20:05:51] 	that sounds pretty stupid to me
[20:05:55] 	what would be the point of it?
[20:06:06] 	brion, more efficient.
[20:06:18] 	brion, as I said, someone showed me his vandal bot's server load cut in half using rollback.
[20:06:34] 	surely less efficient -- taking away rollback UI will discourage people from using it
[20:06:39] 	they'll do manual reverts instead
[20:06:41] 	Er, maybe you're misreading?
[20:06:45] 	I'm not changing the UI at all.
[20:07:02] 	I'm only changing the backend permissions.
[20:07:03] 	"The 'rollback' permission now only affects whether rollback is made available through the interface."
[20:07:37] 	As opposed to affecting whether accessing a rollback URL (such as one constructed by a client-side script) will work.
[20:07:44] 	Which previously it also affected, and now it doesn't.
[20:08:28] 	just give the permission and be done with it
[20:08:32] 	hiding it from the ui makes no sense
[20:08:56] 	Maybe allow rollback for autoconfirmed?
[20:09:29] 	additionally that would appear to give rollback backend permission to those who normally wouldn't get it when the permission is igven to them, which seems odd
[20:09:57] 	brion, well, if we *don't* give it by default, is there any point in requiring user revert scripts to request and resubmit the page text, rather than using a more efficient mechanism that already exists?
[20:10:16] 	either give the permission or don't
[20:10:19] 	Simetrical: Not if they have rollback permission, which they should have
[20:10:25] 	there's no compelling reason to avoid giving it
[20:10:40] 	brion, I thought you said we weren't going to enable it by default.
[20:10:40] 	I say we just give rollback permission to such bots, they can be trusted with that I guess
[20:11:03] 	RoanKattouw, again, why should *any* script have to bother with the inefficient implementation?  What's the point?
[20:11:14] 	It shouldn't. It should have rollback permission
[20:11:28] 	RoanKattouw, any user can currently install a revert script.
[20:11:41] 	Oh right I was thinking bots
[20:11:46] 	Simetrical: if you give out the permission, just give it out then
[20:11:54] 	what's the fucking point of fucking around and going WINK WINK USE THIS SCRIPT
[20:11:59] 	How about $wgGroupPermissions['autoconfirmed']['rollback'] = true;
[20:12:03] 	that's just stupid
[20:12:11] 	brion, the only difference in the commit is improved efficiency.
[20:12:12] 	That's it.
[20:12:18] 	It allows no one to do anything they couldn't already do.
[20:12:21] 	I'm with brion here: hiding something from the UI yet opening the back door is stupid
[20:12:28] 	Exactl
[20:12:30] 	y
[20:12:37] 	That's why rollback shouldn't be as restricted as it is now
[20:12:44] 	That's a separate issue.
[20:12:48] 	The enwp community thinks different, though
[20:12:53] 	Assume that on some wikis, rollback will be restricted.
[20:13:06] 	In that case, does this change do anything but increase efficiency for, e.g., some bots?
[20:13:16] 	If not, is there any reason not to do it?
[20:13:24] 	Well the problem these bots have is not in core, it's in enwp's LocalSettinsg
[20:13:32] 	It's both.
[20:13:35] 	Or either.
[20:13:50] 	Even in that case, the point still stands for user revert scripts.
[20:13:57] 	This is *nothing* but a minor efficiency improvement.
[20:14:08] 	It has no other effects whatsoever, positive or negative.
[20:14:15] 	If such scripts want the benefit of rollback, whoever runs that wiki should allow them rollback
[20:14:21] 	. . .
[20:14:22] 	What you're intending to do is kind of a double moral
[20:14:47] 	So in other words we should add config options meaning "make random options gratuitously inefficient for this group of users"?
[20:14:47] 	You can't click a rollback link in the interface, but if you've got some brains you can construct a URL that will do exactly the same
[20:14:59] 	RoanKattouw, actually you need a script to construct the rollback token.
[20:15:07] 	It's a hash of the edit token and a couple other things.
[20:15:21] 	You can't do this without a script.
[20:15:24] 	edit token salted with page title and user whose contribs are reverted
[20:15:30] 	Yes.
[20:15:31] 	Also available from the API
[20:15:38] 	Can't be manually constructed.
[20:15:39] 	But the token thing is a good one
[20:15:47] 	Manual rollback is nigh impossible
[20:16:00] 	Right.  Correct.  So this only affects scripts.
[20:16:06] 	And only improves the efficiency thereof.
[20:16:11] 	What is the bad thing you see here?
[20:16:19] 	Anyway, I don't see why we don't just allow rollback for everyone, as we're still forcing 'manual' undoers to use inefficient stuff
[20:16:42] 	So I don't suggest your rev be reverted, I agree with it, but I think we should go even further
[20:16:59] 	Well, the argument there is precisely that untrusted users will be encouraged to rollback more than they should.
[20:17:06] 	03(mod) Provide accessibility of Wikipedia' s sister projects through *one* interwiki prefix - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=4285  (10gangleri)
[20:17:09] 	And/or won't necessarily understand that the rollback button has no confirmation, etc.
[20:17:15] 	It's a separate issue.
[20:17:19] 	03brion * r29762 10/trunk/phase3/ (RELEASE-NOTES includes/Article.php): 
[20:17:19] 	Revert r29705. Simply provide rollback permission if you want people to be able to use rollback.
[20:17:19] 	Otherwise you're going to end up with the restricted portions of it incorrectly allowed when they're not meant to be by accident, which seems pretty senseless.
[20:17:30] 	brion, what restricted portions of it?
[20:19:44] 	present example off the top of my head is the add'l 'bot' hiding method. in general, anything additional that may get in there would end up being forgotten that it's generally-accessible instead of restricted by the right
[20:19:59] 	[that particular one has another right attached to it, i hope, though]
[20:20:14] 	Bot-hiding method is 'hidebotedits'.
[20:20:27] 	Has been for some time.
[20:20:30] 	Only introduced relatively recently however
[20:20:58] 	I would have introduced it if it weren't before doing that, obviously.
[20:21:26] 	Anyway brion, in your opinion should we A) give rollback to anyone by default or B) give rollback to autoconfirmed by default or C) just keep the current situation ?
[20:21:33] *Simetrical 	gives up, is overruled
[20:21:44] *RoanKattouw 	curses the B ) smiley rendering as B)
[20:21:49] 	Simetrical: if you do split it out, i'd recommend splitting the permissions
[20:22:03] 	'rollback' and 'rollbacklink' ?
[20:22:03] 	brion, you mean like rollback and rollback-interface or something?
[20:22:04] 	03(mod) Excess  tags are generated - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=12628  +comment (10seventowers)
[20:22:04] 	but i don't think it's a particularly good idea to begin with
[20:22:33] 	It was just a (substantial) efficiency improvement for anti-vandal bots, which typically aren't given any kind of special status.
[20:22:35] 	with 29705 as it was, you'd have *no* ability to restrict rollback backend, which i rather dislike
[20:22:53] 	(it's trivial to add rollback permission to bot group, of course)
[20:22:54] 	I don't see why rollback is a restricted right anyway; any baboon can fetch the text of an earlier revision and replace the current text with tit
[20:23:31] 	It's no different from editing.  I don't see why it should be restricted separately.  Do we separately restrict the ability to undelete entire pages from the ability to undelete single revisions?
[20:23:38] 	RoanKattouw: it has specific abilities. first, the UI and posting behavior is different to encourage user efficiency for vandal cleanup
[20:23:55] 	second, it has the ability to hide all those edits (now controlled on a separate right)
[20:24:06] 	RoanKattouw, because we don't want to encourage random users to roll back people's edits in one click with a generic message and no explanation.  That's only good for specific narrow cases, i.e., vandalism.
[20:24:20] 	True
[20:24:35] 	But the thing is that rollback is much more efficient than undo
[20:24:41] 	it probably also bypasses plugins like spam blacklist and such
[20:24:49] 	If they go out of their way to devise some other method of doing it quickly, that's their business, but the first time you view a history page you shouldn't see a big "rollback" button and wonder "What does that do?".
[20:24:53] 	Yes, I think it does
[20:24:55] 	so there may be security issues on that front
[20:25:23] 	brion: does an list of all available parameters to interface messages exists?
[20:25:28] 	as such i'd recommend against being unable to restrict it
[20:25:32] 	rollback again, eh? there's a page on mediawiki now: http://www.mediawiki.org/wiki/Random_ideas_for_rollback_to_be_shelved_and_forgotten_about
[20:25:47] 	i was tired of the same IRC chat happening ad infinitum
[20:25:48] 	Well I'd say that anyone who actually uses anti-vandal bots should be eligible for rollback rights
[20:25:52] 	AzaTht: if it does i'm ignorant of it
[20:26:00] 	brion: haha
[20:26:04] 	anyway, i'm going to ignore any more rollback talk today as i'd like to actually get work done ;)
[20:26:05] 	It uses doEdit(), so it will trigger the ArticleSave hook.
[20:26:28] 	RoanKattouw, you're phrasing it as a policy decision, not a technical decision.
[20:26:43] 	brion: could it be possible to list number of parameters to Special:Allmessages?
[20:26:57] 	AzaTht: ?
[20:27:12] 	AzaTht: Look at SpecialAllmessages.php ?
[20:27:18] 	Oh wait you mean something lese
[20:27:20] 	*else
[20:27:38] 	RoanKattouw: the number of available parameters
[20:27:47] 	And what they do
[20:27:48] 	updateArticle uses the ArticleUpdateBeforeRedirect hook, but that's irrelevant to rollback anyway, it's meant to allow changing the redirect URL.
[20:27:54] *Simetrical 	shrugs
[20:27:55] 	the standard message might not use them all I think
[20:28:05] 	Well usually, the default messages use all available parameters and make it pretty clear what they are
[20:28:09] 	RoanKattouw: that also, but it's a bit more difficult
[20:28:15] 	AzaTht, that's not stored anywhere.  It's basically decided by the caller.
[20:28:25] 	Simetrical: aha
[20:28:30] 	So Allmessages has no access to it, anymore than it has access to how it will render.
[20:28:45] 	i see
[20:28:49] 	AzaTht: Think of it as C printf() with the $1 and $2 stuff being %s
[20:28:57] 	Pretty much, yeah.
[20:29:06] 	RoanKattouw: that's correct
[20:29:14] 	Except it doesn't crash the program if you add the wrong number of format strings to the message.  :P
[20:29:19] 	True
[20:29:21] 	hehe
[20:29:40] 	Simetrical: How does PHP react if some varargs function tries to pull more args than it's got?
[20:29:51] 	Would anyone care to field support type question?
[20:29:58] 	Sure
[20:30:06] 	RoanKattouw, the question doesn't really make sense.  Have you seen how varargs works in PHP?
[20:30:14] 	RoanKattouw: you get an array, which has as many things in it as it has
[20:30:15] 	You just fetch the arguments as an array with a library function.
[20:30:17] 	I'm looking it up right now
[20:30:20] 	Oh right
[20:30:32] 	what's really fun is when a *regular* function doesn't get enough args :)
[20:30:33] 	If you try to index past the end of the array, it probably gives you a notice and lets everything proceed.
[20:30:38] 	So you don't fetch them one by one in a loop the way C does
[20:30:52] 	Simetrical: Using undefined keys gives an E_NOTICE indeed
[20:30:54] 	you get warnings spit on your console, then it complains about using unset variables :D
[20:31:07] 	RoanKattouw, well, you use a loop, probably, just foreach, not for.
[20:31:08] 	So essentially, that parameter is NULL
[20:31:14] 	So there's no issue of indexing past the end.
[20:31:19] 	Yeah but you don't use a loop to get them in the first plac
[20:31:21] 	e
[20:31:29] 	Right.
[20:31:38] 	bakunin: What was your question?
[20:31:40] 	RoanKattouw: if you're asking what wfMsg() and co do, that depends on the exact method used to do the variable replacaement
[20:31:55] 	I'm using MediaWiki 1.11.0 with the default skin, but I can't the special type URLs, MediaWiki:Mainpage, MediaWiki:Sidebar, etc.  I suspect there is something obvious I'm overlooking, but can't find anything in the FAQ about it. 
[20:31:56] 	iirc right now any ones that aren't passed in just don't get replaced
[20:32:05] 	the system doesn't know how many parameters "should" be in the message
[20:32:07] 	I know pretty much what they do, just asking exactly how they do it cause I've never worked with varargs in PHP
[20:32:12] 	it just takes those that were given to it and replaces them in the string
[20:32:29] 	bakunin: You should go to MediaWiki:Sidebar as if it were an article
[20:32:42] 	i.e. go to www.example.com/wiki/MediaWiki:Sidebar
[20:32:43] 	(so eg not like C printf, where the format string controls how many vars it grabs of fthe stack)
[20:32:59] 	Yeah, I got that, it's actually better that way
[20:33:05] 	One of the few things PHP does well
[20:33:06] 	Varargs in PHP is better than in C, but it's still kind of awkward (not explicit enough).  Python's way is nicer.
[20:33:13] 	brion: but does all default messages use "all" parameters?
[20:33:29] 	every*
[20:33:33] 	AzaTht: probably, for the most part.
[20:33:35] 	RoanKattouw, it should have something in the type line, and give a fatal error notice if you pass too many args to a non-varargs function.
[20:33:37] 	there's no guarantee they do, howeve.r
[20:33:40] 	AzaTht: wfMsg() calls were generally written before the messages were
[20:33:46] 	RoanKattouw: When I try that I get a URL was not found on this server type error.
[20:33:48] 	some may be "extras"
[20:33:49] 	03(NEW) Watchlist edit broken on en-wikipedia - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=12629 major; normal; MediaWiki: Special pages; (ioeth.trocolar)
[20:34:01] 	hehe; someone filed a bug
[20:34:10] 	Like C's int fun( int args, ... ) or whatever C uses.
[20:34:29] 	bakunin: URL?
[20:35:09] 	Python has def function( arg1, arg2, *numberedargs, **namedargs ), which basically puts all numbered args beyond the second in numberedargs (a tuple), and all named args not already covered in namedargs (a dict).
[20:35:49] 	RoanKattouw: Unfortunately its behind a firewall, I tried this: http://www/prodwiki/wiki/MediaWiki:Sidebar my main page looks like this: http://prodwiki/index.php/Main_Page
[20:36:21] 	bakunin, you want http://prodwiki/index.php/MediaWiki:Sidebar.  Just type it in the search box and hit "Go".
[20:36:32] 	RoanKattow: oops, I flubbed that typing just now, was supposed to be http://prodwiki/wiki/MediaWiki:Sidebar
[20:36:53] 	bakunin: http://www.prodwiki.com/index.php/MediaWiki:Sidebar
[20:36:54] 	brion: about Special:Allmessages, could you do so the output is clipped, as for example Mediawiki:Commons.js does take a large ammount of space
[20:37:22] 	AzaTht: then you can't search,  that would be very bad
[20:37:30] 	Simetrical: Thanks, that worked!  
[20:37:34] 	MZMcBride: PHP error?
[20:37:35] 	true
[20:37:50] 	Thanks all, have a nice day.
[20:37:53] 	brion: perhaps a builtin search :)
[20:38:02] 	bakunin: In the future, just replace Main_Page with MediaWiki:whatever in the URL, or type MediaWiki:whatever in the search bo
[20:38:04] 	x
[20:38:06] 	RoanKattouw: it's giving a white screen; i thought PHP error, but only the parser's been synced today
[20:38:27] 	RoanKattouw: Thanks
[20:38:30] 	I confirm MZMcBride 
[20:38:38] 	MZMcBride: moment
[20:38:46] 	MZMcBride: bug confirmed
[20:39:05] 	i also get a white screen
[20:39:10] 	raw edit works though
[20:39:13] 	hehe; i know it's true; it just seemed that no one cared in -tech so i stopped trying to get someone's attention
[20:39:15] 	:)
[20:39:30] 	MZMcBride: always use the "wrong"(tm) channel :)
[20:39:40] 	heh.
[20:40:08] 	i get no failures currently
[20:40:26] 	heh. working now
[20:40:27] 	brion: how do you mean?
[20:40:29] 	brion: http://en.wikipedia.org/wiki/Special:Watchlist/edit
[20:40:37] 	that's where i am RoanKattouw 
[20:40:39] 	Hey
[20:40:41] 	It works for me :O
[20:40:42] 	that's why i avoided a bug report...
[20:40:56] 	WTF just happened there?!?
[20:41:05] 	got a white screen earlier, but none now that i'm watching the error log
[20:41:06] 	yea, works for me now also
[20:41:20] 	really fishy
[20:41:23] 	14(WFM) Watchlist edit broken on en-wikipedia - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=12629  +comment (10roan.kattouw)
[20:41:28] 	a watched pot never boils?
[20:41:35] 	Guess so
[20:41:47] 	MZMcBride: hehe
[20:41:55] 	A tree falling when no one
[20:42:02] 	's around doesn't make any noise, that kind of stuff
[20:42:51] 	all i'm seeing in the log is the occasional deep parsing loop failing on the one machine running xdebug
[20:46:56] 	ugh i hate shit like this :)
[20:46:59] 	PHP Warning:  urlencode() expects parameter 1 to be string, array given in /usr/local/apache/common-local/php-1.5/includes/GlobalFunctions.php on line 980
[20:47:12] 	something passing incorrect stuff to wfArrayToCGI().... somewhere.....
[20:47:27] 	it's like php never heard of stack traces :P
[20:47:31] 	That's hinky
[20:47:40] 	03(mod) Install TitleBlacklist extension - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=12366  (10N/A)
[20:47:44] 	brion: PHP can produce them, it just doesn't do it spontaneously
[20:47:45] 	(probably not relevant to the watchlist problem, just happened to see it :)
[20:47:58] 	i reopened the titleblacklist bug
[20:48:24] 	RoanKattouw, yep, whereas (for instance) Python throws an exception with stack trace every time you make a wrong step.
[20:48:42] 	Which can be convenient . . . although you'd better be sure to have exception-catching logic all over the place.
[20:49:01] 	brion, isn't it possible to hijack built-in error messages and (for instance) provide a stack trace?
[20:49:03] *Simetrical 	goes to look
[20:49:16] 	in theory yes
[20:49:26] 	unless they're fatal errors, which are usually the once we most want to catch ;)
[20:49:43] 	i believe xdebug will give you a backtrace on those, but it hurts performance pretty bad so we don't have it on generally
[20:52:14] 	http://us2.php.net/manual/en/function.set-error-handler.php
[20:52:36] 	Well, fatal errors are fairly noticeable, so I think they'd be tracked down regardless of whether we get a nice handy stack trace.
[20:53:31] 	the general problem with php error messages is that they only give you the immediate line. if the failure is in a function which was called with bad parameters, you don't get the feedback you need because you're in the wrong place.
[20:53:33] *brion 	stabs php
[20:53:43] 	Which is why printing a stack trace would be nice.
[20:53:47] 	*nod*
[20:55:26] 	As we already do for unhandled exceptions.
[20:55:40] 	ok, what graphic card should I buy?
[20:55:57] 	my radeon X850 went to the fathers
[20:59:30] 	man i can't even remember the last time i bought a graphics card
[20:59:47] 	whatever Our Father Jobs picks is good enough for me i guess ;)
[21:01:36] 	03(mod) Excess  tags are generated - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=12628  (10haridara)
[21:03:21] 	hey brion. I come to nag you about pdfhandler...
[21:03:37] 	is it still on your list ?
[21:04:13] 	yes, still on list.
[21:04:19] 	what are the external requirements btw?
[21:04:41] 	I suppose gs and convert
[21:04:54] 	I did not write it, I have to check
[21:06:21] 	brion: oh, and there's also pdfinfo
[21:06:37] 	03(NEW) Limit page moves onto titles with over 1, 000 deleted revisions to bureaucrats - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=12630 15enhancement; normal; MediaWiki: History/Diffs; (random832)
[21:09:46] 	ok i'm getting that watchlist failure.... but no erro rmsgs :P
[21:09:51] 	Weird
[21:10:09] 	brion: confirmed
[21:10:21] 	hm, if i stick something on the url it comes up fine
[21:10:27] 	Hard refresh has been disabled, unfortunately
[21:10:37] 	or action=edit
[21:10:40] 	Is Special:Watchlist cached
[21:10:45] 	?
[21:10:49] 	ok this might be a blank page stuck in a cache
[21:10:49] 	nice
[21:11:17] 	Special:Watchlist/randomcrap shows up fine, it's just /edit
[21:11:21] 	ok, did a purge of the url and it comes up again
[21:11:32] 	Yep, WFM
[21:13:20] 	03(mod) Watchlist edit broken on en-wikipedia - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=12629  (10brion)
[21:13:55] 	probably one person with a too-huge list hits the memory limit, then it gets cached for everybody
[21:14:22] 	latest versions of php should produce a 500 response code on that sort of thing, but i don't think we've upgraded
[21:15:45] 	cu all
[21:16:19] 	brion: Isn't there some way to config Squid not to cache empty page
[21:16:20] 	s
[21:16:24] 	Or headerless pages
[21:21:50] 	03(NEW) Linksearch page/limit controls missing - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=12631 normal; normal; MediaWiki extensions: LinkSearch; (artificial)
[21:23:31] 	03(mod) Excess  tags are generated - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=12628  (10seventowers)
[21:23:39] 	i want a special page where people can add/remove themselves to a list of predefined groups.  what's a better name from Special:Addmetoagroup?
[21:24:11] 	flyingparchment, how about use Special:Userrights?
[21:24:16] 	I was thinking of that recently, actually.
[21:24:19] 	ugh, not touching that code
[21:24:25] 	It's fairly clean these days, actually.
[21:24:29] 	RoanKattouw: in theory perhaps
[21:24:38] 	or we could have erorr pages that work right :D
[21:24:39] 	HTMLForm has been abolished, it subclasses SpecialPage, and works very nicely.
[21:24:55] 	well, these days it's not particularly restricted, is it?  so letting anyone use it wouldn't require any changes?
[21:25:06] 	night all
[21:25:07] 	No, you'd just need to tweak things a bit.
[21:25:14] 	maybe i'll fiddle then
[21:25:17] 	flyingparchment: You need to tweak some LocalSettings.php stuff
[21:25:20] 	But beware
[21:25:23] 	I was thinking that perhaps we could have two more arrays to allow adding/removing *yourself* from groups.  If that was your only access it would go straight to your account, with no switch form.
[21:25:37] 	There is no way (that I know of) you allow only changing your own groups, but not others'
[21:25:46] 	Yes, that would have to be added separately.
[21:25:54] *Simetrical 	wonders if it's appropriate for core . . .
[21:25:54] 	RoanKattouw: that's why we are talking about adding one.  :)
[21:26:02] 	Simetrical: i don't see why not if it's simple enough
[21:26:07] 	It should be very simple.
[21:26:11] 	i.e. you can either add *everyone* to the bot group or no one at all
[21:26:22] 	But flyingparchment why'd you want that?
[21:26:33] 	RoanKattouw, could be used for, e.g., allowing resignation of ranks.
[21:26:44] 	Probably other things too.
[21:26:47] 	RoanKattouw: we have a group on the toolserver wiki called "query-service".  people in this group receive notification of some bugs in our bugtracker
[21:26:59] 	RoanKattouw: people should be able to add themselves to this group themselves
[21:27:08] 	Right
[21:27:23] 	03(mod) Excess  tags are generated - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=12628  (10seventowers)
[21:27:42] 	Is it terrible if they can also add others? I mean people can unsubscribe too, right? So you could at least allow everyone to add to and remove from query-service
[21:27:54] 	And pray no one starts a userrights war
[21:28:23] 	you shouldn't be able to sign someone up to large amounts of email by clicking a couple of buttons on a website
[21:28:31] 	True
[21:28:52] 	there are also other users (like the one simetrical suggests) where that's not possible
[21:29:18] 	True
[21:29:52] 	neat, kdelibs actually compiled on windows
[21:29:54] 	wonder if it works
[21:30:02] 	flyingparchment: QT does anyway
[21:30:14] 	ok sounds like it was BetacommandBot's watchlist :)
[21:30:18] 	244226 entries :D
[21:30:20] 	Qt has supported windows for a long time... kde is just getting started
[21:30:30] 	brion: that was not it
[21:30:51] 	i have not loaded it until after this was brought up
[21:30:52] 	:)
[21:31:03] 	well some other schmuck with too big list :D
[21:31:05] 	HOLY HELL
[21:32:07] 	ugh, this watchlist clear is going to cause some replag :)
[21:37:31] 	14(WFM) Excess  tags are generated - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=12628  +comment (10Simetrical+wikibugs)
[21:37:59] 	Query OK, 244225 rows affected (7 min 34.54 sec)
[21:38:01] 	jees :D
[21:38:21] 	lol
[21:38:34] 	What did you do, Brion, trash his watchlist?
[21:39:03] 	Waiting for 10.0.0.236: 541 seconds lagged
[21:39:08] 	brion: if only there were a way to only delete some of the rows at once ;)
[21:39:36] 	RoanKattouw: I asked brion to clear the list
[21:39:44] 	Oh ok
[21:39:54] 	Heh, yeah, I'm seeing replag here.
[21:40:00] 	  
[21:40:02] 	      
[21:40:03] 	      
[21:40:05] 	      
[21:40:06] 	      
[21:40:08] 	      
[21:40:09] 	      
[21:40:11] 	      
[21:40:13] 	      
[21:40:14] 	      
[21:40:16] 	      
[21:40:17] 	      
[21:40:19] 	      
[21:40:20] 	      
[21:40:22] 	. . .
[21:40:25] 	you don't need to flood the channel, we get the idea
[21:40:33] 	Alright
[21:40:45] 	Huh?!? It's 638 now
[21:40:58] 	because the slave is still processing the query
[21:41:11] 	Oh right; so the lag will rise, then fall
[21:41:28] 	Which is pretty much what I'm seeing here http://en.wikipedia.org/w/api.php?action=query&meta=siteinfo&siprop=dbrepllag&sishowalldb
[21:43:07] 	You can just look at your watchlist, it has a very helpful message.  :)
[21:43:48] 	LOL
[21:44:08] 	hi, when i create a new line in the navigation, it doesnt show on every site at the navigation. example: i create a portal in the navigation - i see it from the mainsite, but when i go on the rulessite, the portal in the navigation is gone :(
[21:44:23] 	T-Vairus: Ctrl+F5
[21:44:45] 	action=purge might be more helpful.
[21:44:52] 	Yeah
[21:45:05] 	T-Vairus: index.php?title=Some_page&action=purge
[21:45:27] 	oh damn... yes u right - F5 is the magic key - thanks again dude :D
[21:51:00] 	Well the servers have recovered now
[21:51:23] 	Simetrical: Are there any other pages that display a similar lag message?
[21:51:54] 	RoanKattouw, not that I know of offhand.
[21:52:08] 	brion stalled the replication
[21:52:08] 	RC maybe?
[21:52:14] 	Oh lol
[21:55:55] 	bye guys :)
[22:04:57] 	03(NEW) {{USERSINGROUP}} magic word - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=12632 15enhancement; normal; MediaWiki: General/Unknown; (artificial)
[22:07:32] 	ok, svn up time
[22:08:31] 	03(mod) {{USERSINGROUP}} magic word - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=12632   +schema-change; +comment (10Simetrical+wikibugs)
[22:09:30] 	03(mod) {{NUMBERINGROUP}} magic word - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=12632  summary; +comment (10Simetrical+wikibugs)
[22:09:54] 	scap... begins!
[22:25:11] 	yay!
[22:31:22] 	03(mod) Excess  tags are generated - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=12628  (10haridara)
[22:39:56] 	just today I noticed irc.wikimedia.org, which appears to show a live stream of the Recent Changes in any given WMF wiki.  Is the code for the RC bot in SVN?
[22:40:15] 	also, is it truely "live" as in pushed by an extension? or is there polling involved
[22:42:36] 	It polls the RSS feed or something, I think.
[22:42:36] 	Dunno where the code is, but it's somewhere.
[22:42:48] 	thanks Simetrical 
[22:43:05] 	i was hoping that there was an ArticleSaveComplete hooking extension at work there :)
[22:43:35] 	Nope.
[22:46:29] 	03brion * r29763 10/trunk/extensions/Gadgets/SpecialGadgets.php: 
[22:46:29] 	Must call setHeaders() from execute() so non-monobook skins don't explode.
[22:46:29] 	note -- this sucks, we need to make that not be required. :)
[22:49:21] 	!rcbot | jimbojw
[22:49:21] --mwbot--	jimbojw: To create an IRC bot to display recent changes to your Wiki, follow the directions at .
[22:49:42] 	awesome, thanks Duesentrieb 
[22:49:43] 	Simetrical: rss would be much too slow. it actually listens to a stream of udp packets emitted by mediawiki
[22:49:47] *jimbojw 	pets mwbot
[22:49:53] 	Duesentrieb, ah, interesting.
[22:49:54] 	jimbojw: forget the per cruft, use netcat/socat
[22:49:59] 	*perl
[22:50:08] 	cool
[22:50:50] 	jimbojw: if you need it for multiple wikis, it gets more complicated (neeed to demux to multiple irc channels). wikimedia uses a special program for that - for which i don't know if it's documented. ask river, if need be
[22:51:20] 	or adapt the perl cruft to do that :)
[22:52:33] 	ah, i see - /includes/RecentChange.php is actually responsible for posting out that UDP signal
[22:53:04] 	interesting
[22:53:14] 	hiya, i would like to create a password to enter a wiki i installed, could anyone help me? I've looked through the site but unable to find anything :(
[22:53:22] 	i thought that would have been extension material
[22:54:00] 	wietse: did you forget your password?
[22:54:06] 	no no
[22:54:26] 	i'd like to actually set a password for people to access the wiki (even view it)
[22:54:44] 	wietse: are you talking about an HTTP authentication layer?
[22:54:50] 	i've installed it on a server
[22:54:56] 	not sure
[22:55:11] 	haven't done it before so not sure how it works
[22:55:22] 	jimbojw: that hack is older than the extension hook architecture. it would sure be nice to have a decent hook for that, though perhaps a realy plugin interface would be needed, with two dimensions: format and medium/protocol.
[22:55:29] 	not sure if it is something within the wiki or outside of that
[22:55:31] 	this is on my "to do when i am completely bored" list
[22:55:51] 	!access wietse
[22:55:51] --mwbot--	For information on customizing user access, see . For common examples of restricting access using both rights and extensions, see .
[22:56:06] 	thanks mwbot
[22:56:16] 	03(mod) Warning got in includes/Title.php - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=11801  (10brion)
[22:56:17] 	it's a bot, dude :P
[22:56:20] 	it's here to serve
[22:56:33] 	:-)
[22:56:39] 	Much like you!
[22:56:46] 	heh.
[22:56:53] 	no, i'm here to procrastinate
[22:57:01] *Duesentrieb 	ogles long running mysql query
[22:57:11] *Duesentrieb 	pokes server with pointy stick
[22:57:25] 	03(mod) Allow Special:Log/rights to be sorted by rights changed - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=12571  (10mike.lifeguard)
[22:58:40] 	Duesentrieb: my "completely bored" list consists of creating an abstract Factory class and having all object instantiations go through it - so extension developers can override built-in classes ( like Article or Image ) without hacking the core
[22:59:03] 	good old dependency injection via factory pattern.
[22:59:31] 	works for Java
[22:59:33] *jimbojw 	ducks
[23:00:30] 	unfortunately, the only way to modify methods of objects at runtime in PHP is to use a PEAR extension
[23:00:47] 	so an abstract hookable factory seems a reasonable surrogate
[23:00:53] 	IMO
[23:01:00] 	you don't need to do that. you can build a dynamic proxy class that forwards all but a few calls and member accesses
[23:01:08] 	mediawiki already uses this for deferred initialization
[23:01:14] 	yeah, but how do you get it in there
[23:01:16] 	i have used it in extensions to to scary shit
[23:01:28] 	03(mod) File links on image description page should be in alphanumeric (alphabetic) order - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=12532   +easy; +comment (10brion)
[23:01:28] 	yes, how indeed
[23:01:39] 	for the most important bit, the Article calss, there's a hook
[23:01:41] 	right - because there happens to be an ArticleFromTitle hook in Wiki.php that you can exploit
[23:01:46] 	exactly
[23:01:53] 	i've seen your code ;)
[23:01:57] 	hehe
[23:02:23] 	my problem is that that doesn't help all the new Article($title) calls
[23:02:37] 	and all the other classes for which there aren't hooks
[23:02:58] 	could be solved by one simple method:
[23:03:07] 	wfNew(classname, args...)
[23:03:11] 	which would have a hook.
[23:03:18] 	simple
[23:03:35] 	yeah
[23:03:41] 	probably two hooks
[23:03:46] 	one for hijacking creation
[23:03:55] 	and one for doing things to objects right after creation
[23:04:00] 	maybe
[23:04:51] 	hi
[23:05:03] 	is it possible to add a hyperlink without http:// ?
[23:05:10] 	i would like to launch a custom handler
[23:05:14] 	like opensim://
[23:06:19] 	Duesentrieb: you think wfNew() is better than Factory::newFromArgs() :)
[23:06:31] 	Hello, I was wondering if there is a way to insert a line into an article? for instance, if I have a template that expects {{{bla}}} and I could make a small form that would give |bla=... ?
[23:07:35] 	jimbojw: i don't cxare if it's a top level function or a static function in a class. but i'd suggest not to use one factory instance per class, but one single global instantiator method, with hooks.
[23:07:42] 	Wiredtape: Could you give an example?
[23:07:51] 	phrearch: in your LocalSettings.php, add this towards the bottom:
[23:07:52] 	$wgUrlProtocols[] = 'opensim://';
[23:08:05] 	i have no control of the wiki
[23:08:10] 	just wanting to add a custom link
[23:08:20] 	but i can ask the administrators
[23:08:31] 	phrearch: then you are restricted to the protocols supported for external links on that wiki.
[23:08:31] 	Duesentrieb: certainly - there's no need to be instantiating Factories all over the place
[23:08:38] 	exactly
[23:08:48] 	thanks
[23:08:49] 	well, in php, there isn't.
[23:08:52] 	in java, there is :)
[23:08:53] 	i think we're all on the same page here
[23:09:01] 	(arguably)
[23:09:02] 	wah wah wah
[23:09:18] 	well, if we were using Ruby or Python, this would all be a moot point anyway
[23:09:25] 	actually, i'd be much happier if mediawiki was java.
[23:09:39] *jimbojw 	smells flamebait
[23:09:52] 	Dashiva: for instance, I have an FAQ namespace on my wiki, and I have a Template:FAQq (faq question) - basically it expects {{{q}}} and {{{a}}} I wanted to make a small javascript that would handle insertion into {{{a}}} (of course this would involve some sort of php extension)...
[23:10:44] 	CharInsert maybe
[23:11:01] 	Dashiva, is this an extension
[23:11:01] 	?
[23:11:06] 	yes
[23:11:12] 	let me take a look.
[23:12:09] 	03(mod) {{NUMBERINGROUP}} magic word - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=12632  (10artificial)
[23:12:35] 	Duesentrieb: if you're serious about moving to Java, JAMWiki probably needs volunteers (http://jamwiki.org/wiki/en/)
[23:12:40] 	03(mod)  formatnum formats template parameters instead of the returning result - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=12560  (10brion)
[23:13:31] 	jimbojw: i'm serious about wishing it was java. i don't think i'm serious about *porting* and *migrating*. i'm afraid that would be a nightmare
[23:14:13] 	my morbid curiosity is getting the better of me - why would you rather it be Java?
[23:14:26] 	Dashiva, CharInsert could work if it weren't searching for an input box... I basically need something that would insert a line/"text" and save it, reload page and you would have an updated page *(future could do this with ajax as well)
[23:15:09] 	03(mod) Change $wguname/$wgNodeName/wfHostname behaviour - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=12586  +comment (10brion)
[23:16:16] 	14(WFM) dumpBackup.php does not backup Main Page and other pages - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=12593  +comment (10brion)
[23:16:42] 	jimbojw: i'd be curious about an Erlang-based wiki...
[23:17:06] 	Erlang?
[23:18:24] 	Wiredtape: a programmign language with built-in support for clustering
[23:18:40] 	hm... yaws, the erlang webserver, seems to feature a wiki per default
[23:19:02] 	03(mod) {{NUMBERINGROUP}} magic word - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=12632  (10Simetrical+wikibugs)
[23:19:09] 	14(INVALID) Create new prefix COM: for Commons - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=12600  +comment (10brion)
[23:24:19] 	03(mod) Activate Extension:LanguageSelector on wmf wiki - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=12604  (10brion)
[23:24:38] 	Duesentrieb, so, do you know of any way to make an insertion into an article?
[23:25:01] 	Çäðàâñòâóéòå! Åñòü ðóññêîÿçû÷íûå?
[23:25:32] 	Wiredtape: using an input form on a wiki page? no, not off-hand. there are some extensions that can create new pages, or append new sections, based on form input.
[23:25:47] 	for example this one:
[23:25:48] 	Äîáðîé íî÷è! Åñòü ðóññêîÿçû÷íûå ïîëüçîâàòåëè?
[23:25:49] 	!inputbox
[23:25:49] --mwbot--	http://www.mediawiki.org/wiki/Extension:Inputbox
[23:26:04] 	Alex45092: try in english please - or at least use utf8
[23:26:34] 	Duesentrieb, inputbox creates a new article doesn't it? Is there a way to create a new section?
[23:26:47] 	correction: append a new section?
[23:26:52] 	?action=edit§ion=new&preload=Foo
[23:27:01] 	iirrc, inputbox can be configured to do a lot of different things with the input. and there are more extension
[23:27:04] 	s like that#
[23:27:14] 	14(DUP) Special:Statistics shouldn't show page views if they aren' t being tracked - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=12609  +comment (10brion)
[23:27:15] 	03(mod) Break messages used in Special:Statistics down further - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=5619  +comment (10brion)
[23:27:37] 	yea, preload also exists.
[23:27:59] 	but preload only works on blank new pages or blank new sections
[23:28:24] 	but, it is native, doesn't require Inputbox
[23:29:06] 	Hi! I have a problem, I install mediawiki. But when trying to configure refers to the 'DirectAdmin Login Page. " What to do? Maybe the wrong file attributes?
[23:29:08] 	14(WFM) Edit summaries - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=12612  +comment (10brion)
[23:29:17] 	Splarka, I think we talked about this before.. I need something that could work on an existing page.. or atleast manipulate it into a newly formed page.. (meaning it would take the wikitext, add something/delete something, save it and reload it)
[23:29:31] 	Wired: javascript?
[23:30:18] 	Splarka, I'm thinking that would take to long for the user?
[23:30:23] 	"too"
[23:30:40] 	would be quite a mess to do it with javascript
[23:30:46] 	depends on hoiw complex you wanna be
[23:31:05] 	^how
[23:31:05] 	Wiredtape: the tricky part when impmenting this in an extension is to reliably find thte correct line to insert at.
[23:31:08] 	There's a great example of a website that does this .. wikinvest.com does this for their user profiles.. it allows them to append education/work data...
[23:31:25] 	you have to consider comments, nowiki stuff, parts of the page comming from templates, etc
[23:31:41] 	Any good "Email Article to friend extensions.  I've tried a couple, but not exactly what the client is wanting.
[23:31:49] 	I'm sorry, please help me...  I have a problem, I install mediawiki. But when trying to configure refers to the 'DirectAdmin Login Page. " What to do? Maybe the wrong file attributes?
[23:31:52] 	if you're gonna make an extension, have you considered storing it outside the article database?
[23:31:55] 	Duesentrieb, of course.. but I was thinknig to solve this with the use of template variables.. so it would search for {{{q}}} for instance...
[23:32:33] 	aBSDaemon, I made one, you can take a look and ill send you the extension.. http://www.wiredtape.com (any ns==0)
[23:32:41] 	should be possible to add that functionality to inputbox
[23:33:02] 	how do i combine two PHP arrays into one?  does + work?
[23:33:23] 	Splarka, I have.. I'm just puzzeled how wikinvest did it without that.. you can see the data on &action=edit
[23:34:23] 	Splarka, they use javascript of course.. but they must have a way to make the insertions through php
[23:35:24] 	well, there are plenty of hooks
[23:35:33] 	http://www.mediawiki.org/wiki/Manual:Hooks
[23:35:44] 	flyingparchment, I think there's a function for that? (depends how you want to do it?)
[23:35:45] 	have you heard the phrase "By hook or by crook" ?
[23:36:16] 	Splarka, :) 
[23:36:49] 	03brion * r29764 10/trunk/phase3/ (RELEASE-NOTES languages/messages/MessagesRo.php): * (bug 12617) Decimal and thousands separators for Romanian
[23:36:50] 	it comes from (allegedly) the ancient medieval custom of serfs being allowed only to take wood they found on the ground, could pull off a tree by hand, or could pull off a tree with a shepards hook. Any other means of wood gathering being considered criminal.. "by hook.. or by crook"
[23:36:52] 	03(FIXED) Add $separatorTransformTable for ro:wiki - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=12617  +comment (10brion)
[23:38:18] 	14(INVALID) Page count not updated. - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=12618  +comment (10brion)
[23:38:42] 	Good to know.. is there a wp article for that? :)
[23:39:00] 	or wikiquote... 
[23:39:37] 	for using proper hooks vs hacking the code like a crook?
[23:40:01] 	no, lol... for the alleged story .. 
[23:40:06] *Splarka 	asks the InterNette
[23:40:12] 	http://en.allexperts.com/q/Etymology-Meaning-Words-1474/hook-crook.htm
[23:40:21] 	nice...
[23:40:26] 	flyingparchment: array_merge() ?
[23:40:50] 	so, anyway, how do I request an articles wikitext?
[23:45:35] 	Wiredtape, is there a way to set the script to email from the user instead of from a group?
[23:45:53] 	03(WONTFIX) cleaner URLs for editing, previewing skins, etc. - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=12619  +comment (10brion)
[23:46:57] 	aBSDaemon, silly me.. i haven't disabled the old email extension.. use the one with the icon.. it asks for an email address, name and so on.. of course you can modify the css and make it look better (ill get to that part someday) :)
[23:47:23] 	and making it into ajax is also fairly simple.. I can do that tomorrow...
[23:50:26] 	03(mod) It's possible to create new empty pages using {{subst:}}, etc. - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=12621  (10brion)
[23:51:36] 	This hook looks interesting for my needs: http://www.mediawiki.org/wiki/Manual:Hooks/EditPage::showEditForm:fields -- but how do I read $editpage (or how do I pull the wikitext from it so i can scan it?)
[23:52:09] 	Wiredtape, that's exactly what I'm needing
[23:52:12] 	both
[23:53:01] 	brion bugzilla blogging?
[23:53:36] 	aBSDaemon, so you want it now or would you rather wait for the ajax?
[23:53:50] 	14(DUP) serve SVG where supported by UA, not PNG - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=12623  +comment (10brion)
[23:53:55] 	03(mod) SVG client side rendering - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=3593  +comment (10brion)
[23:54:13] 	hi
[23:55:03] 	got a problem, hope someone can help...  Warning: pg_query() [function.pg-query]: Query failed: ERROR: function concat(text, "unknown") does not exist LINE 1: ..._timestamp)...
[23:55:11] 	DatabasePostgres.php on line 533
[23:55:27] 	this happens when trying to delete an uploaded file
[23:56:28] 	03(mod) Remove watchlist entries directly from My Watchlist - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=12626  +comment (10brion)
[23:56:30] 	03(mod) Should be able to remove articles from regular watchlist view - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=424  +comment (10brion)
[23:56:32] 	thats what happens when you use PG! :)
[23:57:56] 	03river * r29765 10/trunk/phase3/ (4 files in 3 dirs): 
[23:57:56] 	* $wgGroupsAddToSelf, $wgGroupsRemoveFromSelf: allow users to add or remove
[23:57:56] 	 themselves from specified groups via Special:Userrights.
[23:58:09] 	I could use it as is.
[23:58:52] 	aBSDaemon, do you want me to send it to an email?