[00:00:05] Alright, that might work.. [00:00:05] see http://www.mediawiki.org/wiki/Manual:Parameters_to_index.php#Parameters_that_are_needed_to_save [00:00:41] but it might only be possible to create new pages, if you can't get wpEdittime [00:00:50] anyone know for sure? [00:02:56] yes, wpStarttime and wpEdittime are required for all edits [00:04:08] MW: right, but for a new page, both can be assumed [00:04:17] can the edittime be faked for an existing page? [00:04:46] just about to ask that [00:05:01] it appears edittime is the timestamp of the first revision [00:05:12] top you mean? [00:05:23] like, initial creation [00:05:51] er wait, nvm [00:06:02] MrDoomBringer: it might be easier to make an extension that can read a log of a web bug in the game form [00:06:13] yeah, last revision [00:06:33] *MediaWiki picked a page to test it on where the first and last revisions were on the same day :P [00:06:41] Eaheh.. That would be a problem.. [00:06:45] eg, the game form might load an image from http://wiki.name/w/webbug.php?stat=value&stat=value&type=gif [00:06:54] you don't have dev access to the wiki in question? [00:07:05] I have sysop access [00:07:13] I'm making things difficult :P [00:07:32] hrm well, what browsers would be transferring the info? just yours, or anyone's ? [00:07:49] Uh, lemme check real quick [00:07:51] it appears that you can fake edittime [00:08:00] It's going to be a server, actually [00:08:00] to what extent I don't know [00:08:13] so, yes, anyone [00:08:47] right, so you can't relax cross-site ajax restrictions, but you could sign the scripts or mutually associate the two sites maybe, but it'd be an ugly hack [00:09:21] then the form could fetch the edit page remotely, including the user's login, edittime, edittoken, and such, and then re-submit it with data [00:09:33] I'm afraid you've lost me again, I have rudimentary self-taught HTML basics.. [00:09:34] but that'd take a javascript guru [00:09:42] I do get the second part there.. [00:09:58] The problem, as you stated, would be a lot of Java, which I have little knoledge of [00:10:11] javascript != java ^_^ [00:10:16] yeah, wpEdittime can be faked completely [00:10:18] much more complicated than I first anticipated [00:10:27] http://en.wikipedia.org/wiki/AJAX [00:10:28] Ah, that simplifies things a bit [00:10:51] basically, it lets a web page fetch data from another web page, live, without reloading the page [00:11:00] *MediaWiki changed it to 0123456789 and it still went though [00:11:08] and without having to anticipate the data when loading the page the first time [00:11:14] of course, I was logged in, so I don't know what happens with anons [00:11:33] That makes sense [00:11:49] but it doesn't work across domains, because that is insecure [00:11:55] right [00:12:05] in fact, livejournal had to give everyone a subdomain to disable ajax "friend"-viruses [00:12:55] anyway, try copying the form, from an anonymous user, with the edittoken of \+ [00:13:09] Uh.. [00:13:11] and be sure to generate the starttime and edittime fields too [00:13:18] WMF is quite slow now [00:13:19] and see if it works, heh [00:15:51] 03brion * r28978 10/trunk/phase3/includes/api/ApiOpenSearch.php: [00:15:51] Add an optional limit parameter (range 1 to 100) to the OpenSearch suggestion interface. [00:15:51] Will be using this in adaptor to the Apple Dictionary widget's search, since we're having performance problems with the special-purpose search backend. [00:17:27] What would "wpRecreate" use as data? [00:17:53] you should be able to ignore that, it would only be needed if the page was deleted recently [00:17:54] Just a boolean true? [00:18:00] Ah, alright [00:18:51] timeline: [edit page loaded] -> [user edits page] -> [page deleted] -> [user submits save] -> [page is not instantly recreated as the page was just deleted, user is asked to verify that they want to recreate the pag] [00:19:10] Alright, let me see if I can get this right the first time.. [00:19:11] reason: pages getting deleted, and re-created seconds later with something like {{delete}} added [00:19:12] http://wiki.roblox.com/index.php?title=User:Mr_Doom_Bringer/playerdata&action=submit;wpTextbox1="HelloWorld";wpStarttime=123456789;wpEdittime=15;wpEditToken=+/ [00:19:20] Sound right? [00:19:24] nope [00:19:31] yeah, thought as much [00:19:32] you can only submit those parameters via POST [00:19:47] I see [00:19:48] action="http://wiki.roblox.com/index.php?title=User:Mr_Doom_Bringer/playerdata&action=submit" method="POST" [00:20:03] ah, alright, how would I include the data? [00:20:13] via things like [00:20:23] inside the
[00:20:27] same as MW does it [00:20:44] ok, let me see if it works from the in-game DHTML window [00:22:52] so you need: [00:22:55] name="wpEditToken" value="+/" [00:23:14] name="wpStartTime" value="[current epoch timestamp]" [00:23:27] name="wpEdittime" value="[current epoch timestamp]" [00:23:40] name="wpTextbox1" value="[data]" [00:23:59] oop name="wpStarttime" [00:24:08] darn irregular camelcasing [00:24:15] heh [00:24:30] all within the header? [00:24:41] (and inconsistant too) [00:25:22] *Splarka nods [00:25:26] er [00:25:26] no [00:25:31] all within the [00:25:59] Ah, ok [00:27:15] http://p.defau.lt/?QKZhmBMkLyIZXl7eNqek0w [00:30:45] Anyone knows how to make mediawiki just accessable to registered users? [00:31:00] Problems with the Lua compiler, almost got it.. [00:31:03] And how to register users manually? :P [00:31:50] timofonic> have a look at the FAQ [00:32:13] hashar, http://www.mediawiki.org/wiki/Manual:FAQ ? [00:32:32] your answer should be there :) [00:32:37] Thanks [00:33:33] 3.13 :) [00:34:04] timofonic: http://www.mediawiki.org/wiki/Help:User_rights is a fun read too [00:35:03] I started using the extension CategoryTree which works well in Monobook, but not in the Cavendish style (from Mozilla) which I uses. Any clue? [00:35:27] Splarka, thanks :) [00:45:47] Is it due to a $wgBreakFrames error? [00:47:46] 03(NEW) Incorrectly truncated multibyte UTF-8 char - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=12444 minor; normal; MediaWiki: Internationalization; (alexander.sigachov) [00:58:16] Hi all. I've got a problem with $wgUseTidy that is driving me nuts. I've got a line in my tidy.conf file that specifies "show-body-only: yes", but tidy seems to be ignoring it and wrapping the HTML of my articles in header and footer HTML tags. Does anyone have any idea why this might be happening? [00:58:53] so, you have double-nested HTML tags in your page source? [00:59:00] Yes. [00:59:44] Right after it begins the body Content, right under "
", I get the following: "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd"> [00:59:44] [00:59:45] [00:59:45] test [00:59:46] test [00:59:48] [00:59:50] [01:00:25] hmmm... then something in your config is messed up, or tidy is messed up (or both). I really don't know [01:01:20] I've gone over the config a zillion times, and it's identical to what is recommended on http://www.mediawiki.org/wiki/Manual_talk:%24wgUseTidy#Example_Tidy_Configuration_File [01:05:27] hey :) [01:05:28] does anybody know from where to download the new mediawiki? (1.12alpha (r28966)) [01:05:41] svn.wikimedia.org [01:05:55] mark: Where there? :-) [01:06:04] use subversion [01:06:09] there are no snapshots [01:06:47] http://svn.wikimedia.org/svnroot/mediawiki/trunk/phase3 [01:07:25] twincest: Oh, twincest, is that all? :-) [01:07:43] yes, that's the svn path for head [01:07:59] if you really want a particular revision for some reason, you can use svn ro -r 28966 http://... [01:08:12] fhmphp [01:08:16] there is a bug in php [01:08:21] no wai [01:09:44] okay, thanks, good night [01:09:45] :) [01:16:34] http://translatewiki.net/sandwiki/extensions/Translate/xxtest.ph garg [01:27:34] Hey guys, I'm confused on the use of {{ifeq:}} and {{#expr:}}. I'm trying to say 938=1, else "number"=938+1, incrementally (939, 940, etc.) How can I express this? [01:28:15] 938 != 1, right? [01:28:22] yeah [01:30:39] {{#ifeq:938|1|938=1|{{#expr:938+1}}=938+1}} [01:30:45] is that what you're getting at? [01:31:02] or did I misinterpret the request? [01:35:52] that gives 938, 939, etc., but it needs to output 1, 2, etc., respectively [01:39:28] {{foo|938}} returns 1, {{foo|939}} returns 2, {{foo|940}} returns 3, etc? [01:40:22] Splarka: Exactly [01:41:50] so something like {{#expr:{{{1|938}}}-938}}... [01:42:02] *Splarka pokles skizzers to do the details, his brain is melted by JS [01:42:16] lol [01:42:22] I keep thinking {{{number|{{#if:938|1|{{#expr:938+1}}=foo}}}}}, but get lost after the = [01:42:29] OH [01:42:48] not sure how to keep adding for each use of foo [01:44:15] {{#expr:{{{1}}} - 937}} [01:44:39] that'll get the number you have to add to 938, plus 1 [01:45:24] *Skizzerz is completely unsure of the intent of such a template, so he can't give very advanced code snippets [01:50:15] {{{number|{{#if:938|1|{{#expr:{{{1}}} - 937}}}}}}}? or {{{number|{{#expr:{{{1}}} - 937}}}}}? those print 938 [01:50:59] *Sasoriza was never very good at math [01:51:32] the first one will return 1 [01:51:48] (hint: use ifeq, not if) [01:52:20] the second will take whatever you put into the parameter and subtract 937 from it (so if you put in 938, it would return 1) [01:56:52] Yes! [01:57:00] Haha! It's working! [01:57:08] uh, good :) [01:57:16] what's working...? [01:57:41] I need to get anonomous edit access back onto the wiki of mine, but I was able to edit a page using a completely separate HTML script! [01:58:32] Thank you all, have a nice night [02:00:53] Ah, there we go... Skizzerz to the rescue once again. Thanks everybody! [02:00:54] rar [02:03:15] rar indeed [02:03:23] *Skizzerz isn't sure what he just did [02:03:55] ohai Sasoriza, I JUST figured out who you are xD [02:04:02] ?? [02:04:10] who am I? [02:04:23] *Sasoriza is still trying to figure that out [02:04:28] *Skizzerz helped you with that image message thingy a while back, didn't I [02:04:35] oh, yep [02:04:56] I knew I'd make an impression eventually [02:05:00] yay for half-attentiveness :P [02:14:13] 03(mod) enable gadgets on en.ws - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=12426 (10gangleri) [02:16:14] 03(mod) enable gadgets on en.ws - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=12426 (10gangleri) [02:31:32] http://localhost/MW/index.php/Special:Version [02:31:41] Notice: Undefined offset: 1 in C:\Server\xampp\htdocs\MW\includes\SpecialVersion.php on line 265 [02:31:45] ugh, flood [02:35:41] hmm? [02:36:20] localhost? no Windows machine running on my localhost. localhost isn't a very useful URL... [02:37:46] heh [02:39:14] *Skizzerz doesn't think he meant for you to follow the link... [02:39:41] *Werdna eats AaronSchulz [02:40:09] *Skizzerz eats a cookie [02:40:48] *Skizzerz goes back to figuring out how to code a GroupPermissions Manager [02:41:31] ooh fancy [02:41:35] might want to chat to brion first. [02:41:45] uh, ok [02:41:50] I asked him about it a few weeks ago, and he said it didn't work too well last time. [02:41:54] ah :P [02:42:12] *darkcode created a primitive namespace manager awhile ago [02:42:17] is there a function that lets you get all of the groups and their associated rights? if not, then that might be a problem [02:42:43] (or a combination of functions that acheives this goal) [02:43:48] Skizzerz: there's $wgGroupPermissions... [02:43:58] *Skizzerz is amazingly retarded [02:44:17] of course if you can assign stuff to it, you can retrieve it too >_> [02:44:31] *Werdna slaps Skizzerz [02:44:37] look at User::isAllowed [02:45:09] then just make a the particular directory that the extension is stored in writable by the webserver so you can generate the config file via a special page [02:45:40] (or perhaps a subdir of that directory... might be safer) [02:46:59] 03nikerabbit * r28979 10/trunk/extensions/Translate/ (7 files): * Cleaning interfaces [02:47:01] why not a table within the database? [02:47:09] because that requires schema changes [02:47:33] having it just require the .php file makes it compatible with the current schema [02:47:34] ok [02:47:58] use a MediaWiki: page? [02:48:17] was thinking about it, but it's not restricted enough [02:48:50] and it's a misuse of the mediawiki namespace [02:49:00] which is for the translation of /interface messages/ [02:49:41] well Spamblacklist, Userblacklist, and the Gadgets extension must all be broken then I guess [02:49:41] my current method requires not only for you to have the permission to access the special page, but for you to know a login name and password as well (which only people with access to the filesystem should know) [02:50:09] darkcode: not broken, but yes, they would ideally store their configuration elsewhere. [02:50:52] they wouldn't be very useful if they did [02:53:00] the ability to destroy a URL is insigificant compared to the power of Userrights [02:53:11] maybe there should be an Extension namespace for configuring extensions within the wiki, if MediaWiki: should be for interface messages only [02:53:20] Splarka: huh? [02:54:05] a sysop can be trusted to edit the spam blacklist, but there is a reason userrights are restricted, a sysop shouldn't be able to grant themselves checkuser by editing an interface message ^_^ [02:54:22] exactly, which is why I made it a special page... [02:54:41] 03brion * r28980 10/trunk/tools/apple-dictionary/ (. index.php): OpenSearch <-> Dictionary.app lookup [02:54:47] skiz: this was in reply to: use a MediaWiki: page? [02:54:49] well you could use the useCan hook to restrict it further couldn't you? [02:54:54] oh [02:55:06] if you trust it with a sysop, then MediaWiki: is good [02:55:08] *Skizzerz is using a login form to restrict it further [02:55:17] if not, then Special: ^_^ [02:55:28] so that only people who have access to the filesystem anyway can use it [02:55:34] Spamblacklist, Userblacklist, and the Gadgets can be trusted with sysops, apparently ^_^ [02:55:35] (or anyone that they tell) [02:58:00] maybe I should remake my Namespace extension into a special page... [02:58:08] perhaps [02:58:50] right now it makes use of MediaWiki:Namespaces for adding new namespaces [03:00:30] brion: Interface messages are for customising or translating the interface, not for configuring the wiki, correct? [03:00:37] yes [03:00:44] though there a few horrifying misexamples ;) [03:01:17] indeed. [03:01:23] *Werdna is responsible for one of these. [03:01:25] brion: by the way, since you tried to code a user permissions manager earlier, can you give any pointers on how to make it run smoothly (or anything NOT to do)? [03:01:41] did i? [03:02:03] *Skizzerz smacks Werdna with a trout [03:02:17] I asked him about it a few weeks ago, and he said it didn't work too well last time. [03:02:29] brion: how is special:version for you? [03:02:53] Skizzerz: I said that he said there had previously been problems encountered, not that he encountered them himself. [03:03:05] AaronSchulz: fine? [03:03:07] yeah, but it made me assume that he tried it >_> [03:03:09] why do you ask [03:03:20] 03(mod) Rename the "Image" namespace to "File" - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=44 (10voyagerfan5761) [03:03:24] *Skizzerz smacks himself with a larger trout for assuming things [03:03:38] Skizzerz: long ago there was a half-done attempt at having group definitions in the database [03:03:43] oh [03:03:51] I'm not using the database [03:03:57] it didn't work very well and had performance and bootstrapping problems [03:04:07] eventually i just wiped it all out and used nice simple arrays [03:04:27] *brion gonna review TitleBlacklist until either he finishes or dinner appears on a plate [03:04:42] I'm thinking of generating a config php file with the grouppermissions arrays (much like how the installer generates localsettings.php) [03:05:09] Skizzerz: make sure it's a separate file [03:05:13] it will be [03:05:33] brion: Notice: Undefined offset: 1 in C:\Server\xampp\htdocs\MW\includes\SpecialVersion.php on line 265 [03:05:40] *AaronSchulz is confused [03:05:44] probably in a server-writable subdirectory of the extension file [03:06:11] AaronSchulz: trunk or branch? which extension triggers it? [03:06:20] trunk [03:06:41] *brion needs a couple 30" god-damn monitors [03:07:13] I turned of all extensions [03:07:24] opps [03:07:28] wrong url [03:07:49] brion: I'm on dual 22" [03:08:51] *darkcode has only a measly 15.5" widescreen monitor [03:09:06] hmmm, mac pro w/ 4gb ram and two 30" screens is only $7094 [03:09:21] but i beat you whiners would rather have a couple servers instead [03:09:27] *bet [03:09:37] naw, get the screens :) [03:09:59] brion: I don't care, but I think ant might ;) [03:10:09] brion: I think it is FlaggedRevs [03:10:54] *brion peeks [03:11:40] brion: if you change FlaggedRevs.php, can you make the special page v1.0 while you are at it? :) [03:12:24] flaggedrevs looks ok on my local test [03:12:52] so, flaggedrevs is going to be live on wikimedia soon, eh? [03:13:18] erik's got some planned schedule of beta test rollouts [03:13:26] impressive [03:13:27] no big rush to production [03:13:30] ui needs lots of work :) [03:13:30] well done, AaronSchulz [03:13:32] *Skizzerz think it would make an extra mountain of work for the sysops, etc. [03:13:49] but as long as it's pretty much working we can actually get it to someplace people will play with it [03:13:56] then everyone can fight over the ui :D [03:14:00] hehe, one perennially postponed plan down, only one to go [03:14:05] *Werdna looks meaningfully at brion [03:14:07] hehehe [03:14:22] single...[cough cough] [03:14:38] brion: what's holding up centralauth anyway? [03:14:51] too much other stuff to do :) [03:14:58] as in, what is there left in it [03:15:08] brion: I don't mind doing some polishing on it, if you're too busy [03:15:10] i'm rescheduling it to start in a small-scale beta next month or so [03:15:33] polish is always welcome :) [03:15:45] brion: if there's a TODO list... [03:16:05] that's one of several things i've got to get together :) [03:16:11] hehe [03:17:04] well.. feel free to throw things to do at me [03:26:04] brion: what needs doing? [03:26:42] i'm in process of putting a list together... [03:26:53] okay, cool [03:27:00] one thing that might be fun though... [03:27:18] ...i was thinking of doing the beta period for centralauth as an invite system [03:27:34] oh? [03:27:38] so we can start it small scale and let people join in slowly while we fix up the 'how'd that get in there' probs [03:27:49] probably not hard to cobble together a system for that [03:27:50] uhuh [03:27:51] interested? [03:27:55] sure [03:27:58] cool :D [03:27:59] on one condition [03:28:15] you have to let me give out the invites :p [03:28:21] hehe [03:28:29] ok, back to working on confirmaccount [03:28:31] as long as i get one eh :) [03:28:35] :D [03:28:39] 03aaron * r28981 10/trunk/extensions/FlaggedRevs/ (FlaggedRevs.php SpecialMakeReviewer.php): [03:28:39] * Stop errors from spewing at special:version [03:28:39] * Bump to v1.0 [03:28:41] oh, that reminds me [03:28:47] *Skizzerz wants one too :P [03:28:52] i've seen some reports of confirmaccount and confirmedit breaking each other [03:29:03] who said that? [03:29:07] lemme see if i can find it [03:29:18] lol, they are supposed to work *together* [03:29:48] Werdna: I'm doing some massive updates per Sanger [03:29:57] http://bugzilla.wikimedia.org/show_bug.cgi?id=12194 [03:30:03] I saw that on mediawiki wiki [03:30:09] might be a local config problem or something [03:30:16] i didn't quite get to trying it yet [03:31:30] brion: so, allow a user who has CentralAuth enabled to invite any other user to use CentralAuth, and make an invite a requirement for Special:Mergeaccount? [03:31:49] pretty much :) [03:31:51] at least as an option [03:32:03] it might make sense to just make a general invite system and have it plug in [03:32:14] then make wikipedia invite-only [03:32:20] and charge for invites [03:32:31] yay, no more vandalism :D [03:33:05] brion: limited number of invites? [03:33:14] probably wise [03:33:35] definately invite the active nicks here ^_^ [03:33:56] you can't invite an IRC nick... [03:34:07] well... the people that USE the nicks [03:34:10] brion: also, should I build it into the default central-auth.sql, or add a patch? [03:34:31] a patch seems more sensible - most people won't want the inviting cruft. [03:34:39] *nod* [03:34:51] having an unused table doesn't cause any problems [03:34:59] requiring a patch makes it more annoying for people who do want it [03:35:27] hmm.. good point [03:36:19] (i could actually see people using an invite-only wiki, for closed/close communities it's a nice way to avoid abusive users) [03:36:35] it's not an invite-only wiki [03:36:44] it's invite-only account-merging. [03:36:57] well that's another reason to consider making it a separate invite system and plug it into that :D [03:37:00] why not just rate limit it? [03:37:03] then we can reuse it for other things [03:37:33] oh I see.. so you want me to build it into core, and let CentralAuth plug into it. [03:37:39] twincest: it can be useful to start with people who more or less know what they're doing and/or know people who do [03:37:53] then you've got a social network for them to beg for help to when it breaks [03:38:03] rather than whatever random percentage of people happen to show up [03:38:03] brion: you know that after about 3 hops, it'll go right back to the people with no clue though ;) [03:38:12] :) [03:38:24] but they ask their invitees. [03:38:29] inveiters* [03:38:32] inviters, even [03:40:30] brion: OK, my evil plan atm is to create an Invitations extension, with a special page Special:Invite, of which CentralAuth can be one user. [03:42:34] 03(mod) ConfirmAccount extension spammed by bots from XXXX sites - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=12194 +comment (10JSchulz_4587) [03:42:43] weird... I did print_r($wgGroupPermissions), and some of the permission keys had empty values [03:45:00] it appears the empty value means false, but I'm slightly confused as to why it's showing in the first place... [03:49:56] Werdna: whee [03:50:27] Skizzerz: print_r sucks [03:50:30] use var_dump or var_export [03:50:38] meh, it works enough for me [03:50:49] apparently it doesn't eh [03:50:57] or it would be telling you that your false values are false [03:51:23] it's saying stuff like "[userrights] => " instead of "[userrights] => 0" [03:51:52] *brion points at the suckage :) [03:52:46] yeah, fine, I'll do the var_dump or var_export when I actually need to see if the values really are getting assigned or not. I was only worrying about the structure of the array when I did the initial print_r [03:53:17] mmm [03:53:32] spinnage keish [03:54:40] *brion spins [03:58:02] brion: I should include such features as checking whether someone is invited to a feature in a separate object, right? [03:58:15] *qsheets wonders which way brion spins [03:58:23] i.e. as static memers of a class Invitations, contained in Invitations_obj.php [03:58:41] qsheets: to make some spinnage for AaronSchulz's keish [03:58:56] lol [03:59:31] brion: $wgMemc->set() is in millisecs right? [03:59:34] Werdna: hmm [03:59:50] AaronSchulz: the expiry time is in seconds [04:01:16] auh oh [04:01:50] well that would explain some things at least [04:02:30] brion: what dost thou thinketh? [04:06:55] i hope you mean "quiche" [04:07:04] unless you're eating this guy: http://en.wikipedia.org/wiki/Keish [04:42:29] how do i request commit access? [04:43:36] you ask brion or Tim nicely. [04:44:32] both are away [04:44:46] brion@pobox.com [04:44:53] Tim's on holidays, probably not worth annoying him [04:45:11] give him a few lines showing that you're fairly trustworthy, and what you'll use it for. [04:45:38] "fairly trustworthy" means that four or five patches of yours have been committed. [04:45:38] NO COMMIT ACCESS FOR YOU [04:46:29] yay, about 50% with that manager... [04:46:34] *Aviator sobs [04:46:46] oh wait, forgot about i18n... so 30% [04:46:56] heh. [04:47:04] *Aviator runs to a corner [04:47:05] I'm writing the very-back-end for Invitations. [04:47:10] cool [04:47:15] Skizzerz: you moved $wgGroupPermissions into the database in only a few hours? [04:47:28] twincest: I'm not moving it into the database [04:47:29] twincest: it's not in the database. [04:47:37] how does it manage them then? [04:47:38] I'm generating a php file [04:47:43] He's auto-generating userperms.php or something, and including that. [04:47:44] haha [04:47:48] that'll never get committed [04:47:55] it'll be an extension [04:48:07] why not do it properly and put it in core [04:48:09] *Skizzerz doesn't see the point in it being part of core, in all honesty [04:48:12] needs to be done anyway really [04:48:21] I'd love to make a Special:Configuration [04:48:26] but really, far too much effort :p [04:48:38] that and namespaces are the most annoying shell bugs at the moment, because there's no reason shell access should be needed [04:49:30] but if I get the php version working, I'll start working on moving it into core, but that's going to be a monstrous headache for easy implementation [04:49:34] maybe interwiki import configuration too [04:49:39] *moving it into the database [04:50:22] I'd probably borrow from your interface. I hate doing interfaces. [04:50:35] although I did have a pretty good idea for how to do an interface for Userrights. [04:50:58] we already have Userrights... unless you're talking about some other kind of it [04:51:09] I mean a new interface for it [04:51:20] Werdna: drag & drop ftw [04:51:24] list out all the groups as checkboxes, grey out the ones that can't be changes. [04:51:29] changed* [04:51:57] ah [04:52:13] *Skizzerz is hiding the ones that can't be changed [04:52:24] nah [04:52:27] er, is tentatively hiding [04:52:27] it's misleading that way [04:52:34] put the ones that can't be changed in [04:52:37] haven't got around to making that part of it yet :P [04:52:41] twincest: :P [05:13:56] crap [05:14:37] shit [05:14:49] Werdna, is that address the real one? [05:15:21] brion@pobox.com is the real one [05:15:30] indeed it is. [05:16:45] simply leave the unsettable ones off the list, but make them addable anyways by editing the form [05:16:51] oh, wait, rotem already did that, hah [05:17:16] oops, my bad [05:18:31] apparently, thunderbird didn't handle mailto links very well [05:21:30] Splarka: touche! [05:39:41] 03werdna * r28983 10/trunk/extensions/Invitations/ (. Invitations.php Invitations_obj.php README invitations.sql): [05:39:41] Add partially complete Invitations extension. [05:39:41] In brief, the invitations extension works like the gmail invite system. It is intended to be used for CentralAuth, first. [05:39:41] Although the back-end code is complete, it is untested, and the extension still [05:39:42] lacks an interface. I am hoping to rectify that at some point in the future, [05:39:46] although I do not object to somebody else jumping in and doing it (they'd [05:39:48] probably do a better job!) [05:39:50] yay, straight-written code had only 3 syntax errors. [05:48:46] 03(CLOSED) Set sitename for Lithuanian Wiktionary - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=12427 +comment (10vpovilaitis) [05:50:09] 03(CLOSED) Set logo for Lithuanian Wiktionary - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=12429 +comment (10vpovilaitis) [06:15:44] 04(REOPENED) Create namespaces for Lithuanian Wiktionary - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=12428 +comment (10vpovilaitis) [06:32:05] 03(FIXED) Create namespaces for Lithuanian Wiktionary - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=12428 +comment (10jeluf) [06:47:18] 03(NEW) Edit link magic word - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=12445 15enhancement; normal; MediaWiki: General/Unknown; (MZMcBride) [06:54:44] 03(CLOSED) Create namespaces for Lithuanian Wiktionary - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=12428 +comment (10vpovilaitis) [06:55:39] Werdna: would you be willing to write a quick magic word patch? [06:56:34] are there any devs about? [06:57:42] Betacommand: hello [06:57:52] MZMcBride: hay [06:58:05] caused a little ruckus today, eh? [06:59:02] mz that is SOP [07:00:39] true [07:01:07] no one's usually around at this hour... it's busier during 8-4 EST [07:01:51] and 8PM-12AM EST [07:03:06] MZM: it might be possible also, to allow full URL syntax, eg [link text], to allow local URLs as well as full, eg... [/w/index.php?title=Foo?action=edit edit] (per 12445) [07:03:51] yeah, that's an option too, though kinda uglier :) [07:05:04] or, simply trim all external links that match the wgServer + wgScript, of that [07:06:00] that too [07:06:10] anyway, you do realize that handling this in any way could cause problems on static dumps and live mirrors? [07:06:30] well, lets say, "inconsistencies" ^_^ [07:06:42] are we worried about live mirrors having problems? [07:06:56] the [/w/ trick would allow oldid URLs to work as well without a separate magic word... [07:06:58] well, to take another example [07:07:41] for a page that is meant to be copied: This page is from Wikipedia, the original location is [{{fullurl:{{FULLPAGENAME}}}} {{FULLPAGENAME}}]. [07:07:59] i know what you mean :) [07:08:17] *Splarka nods [07:08:25] don't section [edit] links cause issues too then? [07:08:35] sure, but why make it worse? [07:08:55] but yah, the best solution would be to allow a syntax to have local-ish external URLs.. [07:08:58] well, a magic word could have an automatic wrapper... [07:09:13] you mean [/w/ or something? [07:09:18] [{{localurl:{{FULLPAGENAME|action=edit}} foo] [07:09:20] yah [07:09:58] yeah, i just keep thinking that there could be security issues, though i guess not... [07:10:12] and then caveat the hell out of it, that "this will cause a link to be local even if the page is mirrored or transwikified, use with friggen caution" [07:10:44] yeah, but the parser could be nice and wrap the links in a span or id [07:11:14] why not just class it? like [07:11:28] [07:11:35] mmm, conflicty [07:12:17] could use an obscure name, wpExtLocText or something [07:13:26] Betacommand: are you taking notes? [07:25:27] hmmm [07:27:18] MZMcBride: I wonder if one could just add $wgScript to $wgUrlProtocols [07:27:42] hmm... [07:28:14] it wouldn't work with the stylepath, api, or article paths, but would work with any /w/index.php url, if it worked [07:31:01] Splarka: can you test it? [07:31:21] I can't, asking someone else to in PM [07:31:33] but ya got a room full of ~160 devs ^_^ [07:31:42] what are you trting to do? [07:31:54] i can do it... i think [07:32:08] twin: http://bugzilla.wikimedia.org/show_bug.cgi?id=12445 [07:32:36] pondering if adding $wgScript to $wgUrlProtocols will simply allow URLs like [{{localurl:foo|action=edit}} foo] [07:32:45] i don't understand the problem. why is http://... a problem? [07:32:50] no idea [07:32:56] it's screwing up Special:linksearch [07:33:07] *Splarka is not here to understand problems, just suggest insane solutions and find silly bugs ^_^ [07:33:24] which makes it difficult to find self-references and bad content citations [07:33:25] MZMcBride: in what way? [07:34:01] the results are full of pages that contain a template that contains [{{fullurl:... [07:34:43] are there any cases where {{fullurl}} URLs should be shown in that list? [07:35:14] i don't think so [07:35:24] i'm not really sure [07:35:45] not that editlink is necessarily a bad idea, but i don't think the problem has been thought through properly [07:36:07] the magic word was one solution; splarka pointed out others [07:36:30] though all of them have the possibility of screwing up dumps and mirrors [07:36:47] Splarka: i can test now [07:36:58] self-referencial 'edit this page' links are already a problem for mirrors [07:37:07] because usually, you can't edit the mirror [07:37:27] that's what i said; but on the other hand, we probably shouldn't make the problem worse [07:37:41] and then caveat the hell out of it, that "this will cause a link to be local even if the page is mirrored or transwikified, use with friggen caution" [07:37:59] what line do you want me to add to localsettings? [07:38:17] Splarka: yes, but my point is that it's already a problem; this doesn't add a new one. it just changes it slightly [07:38:46] anyway, you do realize that handling this in any way could cause problems on static dumps and live mirrors? [07:38:46] don't section [edit] links cause issues too then? [07:38:46] sure, but why make it worse? [07:39:06] *Splarka invites twincest to discover the magical world of recent scrollback ^_^ [07:39:08] it doesn't make it worse [07:39:12] please read what i'm saying [07:39:13] it adds to it [07:39:17] no it doesn't [07:39:23] the problem is already there, this just changes it slightly [07:39:33] that is, his "solution" will cause *more* local links to exist [07:39:39] which will make the problem bigger [07:39:57] why will it cause more links to exist? [07:40:10] bigger number of problem instances = worse problem [07:40:18] no, it will cause more local links to exist [07:40:19] why will there be more problem instances? [07:40:21] if people can make them [07:40:25] why will it cause more local links to exist? [07:40:38] because MZM will go about trying to clean up http://en.wikipedia.org/w/index.php?title=Special%3ALinksearch&target=en.wikipedia.org%2Fwiki%2F [07:41:03] those links already exist. whether they're local to the mirror or remote to wikipedia makes no difference [07:41:16] by changing unneeded {{fullurl:}}s to {{localurl:}}s [07:41:26] (and in general, on a mirror, links constructed that way will still point to the local mirror...) [07:42:04] Splarka: http://pruebita.com [07:42:54] IS there a MediaWiki to HTML parser that is one step setup??? [07:43:00] ...or close to [07:43:04] twin, to summarize... [07:43:06] self-referencial 'edit this page' links are already a problem for mirrors [07:43:06] qsheets: mediawiki [07:43:12] hahaha [07:43:39] seriously. it's probably easier than any other solution [07:43:46] no, really...i can't have this part going through to the layout [07:43:47] you could write a maintenance script or something to do it [07:43:56] 1. there are few local links in the body of a mediawiki page, and no wikicode to make them arbitrarily... [07:44:00] 2. adding a way, will create more of them [07:44:15] 3. if you have more, the "problem for mirrors" extends scope [07:44:27] Splarka: what's this? http://en.wikipedia.org/wiki/Template:Edit [07:45:06] well, lets see what it does... [07:45:07] anywhere that's used will show up in Special:Linksearch [07:45:17] {{edit|foo}} -> [07:45:24] [edit] [07:45:33] vs an edit section... [07:45:41] mTitle->getNamespace() == NS_IMAGE ) ); [09:39:53] VasilievVV: i can only assume you didn't mean to commit that ;) [09:40:05] twincest: sure [09:42:50] 03vasilievvv * r28987 10/trunk/phase3/includes/Article.php: Oops, forgot to remove this debug stuff [10:06:55] Why page moving is in Title, not in Article? And why we use Special:Movepage, not action=move? [10:09:25] hi [10:18:56] 03grondin * r28988 10/trunk/extensions/Farmer/Farmer.i18n.php: Fix french message key [10:23:57] Hello [10:24:24] hello [10:26:21] Hi Wiredtape :) [10:45:52] hello, does anyone run a wiki based on the nightly compiles? (other than wikipedia)... [10:48:30] ^_^ Love the new commit... finally image redirects... [10:48:49] Are we going to be seeing image moving at some point in time, hopefully? [10:50:27] Dantman|FS, they fixed image redirects? so how does it work now? something like this-> [[blabla|[[Image:bla.jpg]]]] ? [10:50:45] Eh? no [10:51:11] It looks like the latest HEAD of MediaWiki now works with redirected images [10:51:27] As in, you have a image named Foo.png, and a redirect at Bar.png [10:51:47] And using [[Image:Bar.png]] will display Foo.png because Bar.png is a redirect. [10:52:07] ok.. didn't know that could be usefull? :) [10:52:16] ^_^ of course [10:52:43] If you've ever run a wiki you know the annoyance of people uploading images with absolutely crappy filenames [10:53:03] You have to download the image, and reupload it at another filename [10:53:26] Which is really annoying if you have a slow connection at the time, or have a lot of images to change the names of. [10:53:57] ^_^ Now that images can be redirected, it'll be easier to create the ability to rename an image without it breaking things. [10:54:24] hmm... never thought of that... :-) [10:54:31] And public wiki communities will LOVE!!! that, as it'll save hours uppon hours of godafwull troubles [10:55:24] wouldn't it have been easier though to create a maintanence script to do that? [10:56:04] public communities have no access to maintenance scripts [10:56:21] Besides, without redirects you still run into issues with things breaking [10:56:36] Which is why the redirects are nice [10:56:40] point taken ... [10:57:03] by the way, do you happen to know if there is a reason monobook was set to xhtml transitional? [10:58:24] Hi! [10:58:35] hi genjix [10:58:41] Probably so that using things like target="_blank" wouldn't give validation errors [10:58:59] Not like that matters... MediaWiki is TagSoup anyways, it's not even XHTML [11:00:14] Dantman|FS, true... I'm just going crazy from IE6 and the different browser quirks.. people over at #css have told me if I set it to strict it would fix it.. but it hasn't.. and I don't want to mess with mw... :) [11:01:00] ^_^ well... if MW did use actual XHTML, IE wouldn't even be rendering it... [11:01:44] why not? [11:01:58] IE isn't XHTML compatible, even IE7 isn't [11:02:17] Send out a XHTML page with the proper mimetype and IE returns a document tree [11:02:23] really? I was under the impression it had a workaround.. [11:02:30] Ya... ^_^ TagSoup [11:02:34] hehe [11:02:39] lol [11:02:55] No soup for you! [11:03:43] I was actually working on a site of my own... Developer profile mostly (^_^ easer to get hired with a good site of examples) [11:04:11] I nicely coded it to conditionally output valid XHTML or valid HTML depending on what the browser supports [11:05:04] I was actually talking about that just yesterday.. (cond doctypes) .. It seemed like it would make a mess of things... did it work out good for you? [11:05:18] Actually, it wasn't just the doctype [11:06:03] XHTML and HTML have different markup to consider [11:06:43] Gimme a sec [11:06:48] Dantman|FS: [11:06:56] argh.. [11:07:16] *VasilievVV can't find backspace key on the keyboard [11:07:17] so seperate html files? did you do this through apache, or php? [11:07:30] siebrand have time? [11:07:36] It's a php site... no, not seperate files [11:07:36] VasilievVV, lol [11:09:42] 03(mod) Auto-noindex user/user talk pages for blocked users. - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=11443 (10huji.huji) [11:16:10] Wiredtape: This is the source seen in IE [http://wiki-tools.pastebin.com/m68f381ce] And this from Firefox [http://wiki-tools.pastebin.com/d5f1d888d] [11:19:41] Dantman|FS, thanks [11:19:57] I actually use functions to generate the tags [11:21:25] That's what I was thinking... is this a per-tag thing? [11:21:52] Depends [11:22:24] I do script and link tags, as well as some others with specal functions [11:22:41] But tags can genericly be opened and closed [11:22:54] Dantman|FS, what's the php function to return browser? [11:22:54] It's similar to the Xml class in MediaWiki [11:23:01] Doesn't need one [11:23:20] (The tricky thing, is making a list of what HTML tags to close) [11:23:56] Dantman|FS, sounds good :) [11:24:03] Every browser sends a Accepts header to the server when it makes a request [11:24:10] In there is a list of mimetypes it'll accept [11:24:37] I search that list for 'application/xhtml+xml' (The legal XHTML mimetype)... [11:24:50] If it sens it, I return XHTML with that mimetype [11:24:59] Otherwise, I return HTML with text/html as the mimetype [11:25:00] else strict... [11:25:09] Nah... both are strict [11:25:18] HTML strict, and XHTML strict [11:25:39] oh yeah.. i didn't notice that on first look :) [11:26:51] hmm.. but if you wanted to do the same, but checking for doctype compatibility? [11:27:06] Eh? [11:28:13] I am assuming there is a way to check what browser someone is using through php.. and then you could set strict/transitional... [11:28:36] Ick... there is... but it's not that trustworthy [11:29:19] User agent strings... there's a whole project dedicating to constantly compiling a list which is used to detect the browser [11:29:27] But it's quite easy to trick that [11:30:06] oh? [11:31:04] Opera lets you pick from one of 3 user agent strings... One of those mascarades Opera as IE... Konqueror lets you completely customize the User-agent string you send [11:31:37] So while you /can/ detect a browser... Someone could actually be using one browser while their detected User-agent says another [11:32:00] Not to mention that you have to constantly keep that list up to date because of new browsers coming out [11:33:09] yes, but writing css\html for non-ie browsers isn't that big of a deal.. it's IE that drives me crazy.. and IE/windows doesn't let you pick a user agent string.. (afaik) [11:33:48] Well... technically you're not supposed to [11:34:10] The User-agent is the info sent out telling what browser you're using [11:34:42] If you're sending out fake results, then what's the point of detecting things to try and make things compatible. [11:36:13] Dantman|FS, but I would say less than 5% of the internet users know how to do these things.. and so, compatability wise you would be going for the other 60%+ IE users.. [11:37:12] I would for instance, run a small script to check what user agent, and if IE direct to one php, and if other than to a different php [11:37:28] O_o [11:37:45] one script can handle everything on it's own if you detect things [11:38:13] 03(mod) Show message on user page of blocked users - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=2044 +comment (10huji.huji) [11:38:34] But really... I see little need for browser detection [11:38:47] yes, well I would probably prefer to keep things "tidy" so I would #include a script on an if basis... [11:39:01] (can you #include based on if? :) ) [11:39:31] Dantman|FS, if I don't than what solution do I have for MW and IE6? [11:39:37] well ya... but you're missing how apps like MW workd [11:39:51] We don't need to pick out an include ""; [11:39:59] We don't even use echo [11:40:48] As for IE6... take a look at monobook, conditional comments [11:41:03] Best way to make IE hacks work to fix things [11:41:22] And they don't cause validation errors [11:41:41] Dantman|FS, point taken. However, I think in my case the problem is with the doctype.. [11:42:07] The doctype is as loose as it's going to get [11:42:07] if you have IE6 you can take a look.. (im using a custom skin) [11:42:17] Nope, IE7 [11:42:29] IE7 works fine :P [11:42:40] transitional is the loose version of the doctype [11:43:10] I am aware of that.. :) [11:43:39] I don't see how using strict would make things render better, it's more likely for it to cause issues [11:44:55] Dantman|FS, that's what I thought as well. But the guys over at #css seem to think differently.. apparently hasLayout has something to do with it... [12:04:14] *qsheets throws Aviator accross the room... FLY [12:05:57] hey ive istalled a pop up on my mediawiki with common.js. Now the popup opens on every page but i want that the popup only opens on the mainpage? [12:07:09] bango: what version of mediawiki? [12:07:34] one moment please [12:08:10] 1.11.0 [12:08:30] do a check for wgPageName [12:08:54] §wgMainpage ? [12:09:05] no, in JS [12:09:11] if (wgPageName=="Main_Page") [12:09:12] bango: MediaWiki:Mainpage [12:10:04] okay i will add it to common.js [12:10:07] well [12:10:16] add it as a conditional to whatever activates the popup [12:10:40] you might find it easier to add to the top of whatever function is activating it: if(wgPageName!="Main_Page") return; [12:11:06] note that "Main_Page" might not be your mainpagename, view the page source of your main page and look for var wgPageName = [12:13:38] VasilievVV, MediaWiki:Mainpage? [12:20:20] mhm [12:20:29] how can i show you the text= [12:20:49] so that i must not poste it in mirc [12:21:14] !pastebin | bango [12:21:14] --mwbot-- bango : Please do not paste more than 2-3 lines of text into the channel as it disrupts the flow of conversation. Instead please use a pastebin such as and post a link to your paste in the channel. [12:23:01] http://pastebin.com/d497b7c46 [12:23:10] i did it, but it doesnt work. [12:23:11] What is the use of Mediawiki:Mainpage? [12:25:43] mm [12:27:18] Mediawiki:Mainpage defines what page is the main page [12:27:29] Splarka, thanks. [12:27:30] eg, what the logo links to, what the 'mainpage' sidebar link points to [12:27:35] bango: it goes in with the function [12:27:55] Splarka, would it define what index.php first heads to? [12:28:01] yes [12:28:14] bango:, don't put