[00:00:32] <^demon> Splarka: Mind if I quote you in my svn commit msg? [00:01:56] <^demon> nevermind. reverted. I'll go break something else tonight. [00:01:57] 03demon * r36536 10/trunk/phase3/ (6 files in 4 dirs): Self revert, this isn't really implemented right. [00:02:57] oop, belated, but you can quote me next time [00:03:08] (or in the bug) [00:03:19] basically, I think you'd want... [00:05:26] in the template skins: [00:05:33] [00:05:34] [00:05:59] *Splarka checks the specs on [00:08:42] > Multiple xml-stylesheet processing instructions are also allowed with exactly the same semantics as with LINK REL="stylesheet". [00:09:04] okay, so that's good to go [00:09:14] and then, in cologneblue/nostalgia/standard: [00:09:17] [00:09:17] [00:15:10] 03(mod) Print stylesheet should be explicitly editable (e.g. MediaWiki: Print.css) - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=2889 (10herd) [00:46:32] uuquit [01:17:01] What parameter determines the path used for css and js files? [04:11:39] evening all. [04:12:12] Within an extension, is there a variable that stores the lasted-modified time for the current page? [04:13:06] I'm trying to implement a simple caching system based on the contents of a file on-disk. [04:13:52] Specifically, if the last edit to the current page is more recent than the mtime on a file, I want to do taskA(), otherwise, taskB() [04:15:24] Hawson: I'm guessing $wgArticle->getTimestamp(). [04:16:18] Hmm...that might be perfect. [04:16:39] is there a nice reference for utility functions like this? [04:16:40] depends on what definition you need, of "modified" [04:17:13] Splarka: a simple last edit to the page. Major or minor, don't care about who did it. [04:17:23] grep ^_^ [04:17:45] Hawson: Use the source, Luke! What documentation MW should be ignored because it is out of date and thus inaccurate. [04:17:56] s/MW/MW has/ [04:18:02] davidmccabe: heh, noted. ;-) [04:18:16] *Splarka did mean grep the source... [04:18:28] I'll happily dive into the source if needed, but if (good) docs exist, that's faster. :) [04:19:00] you could also use getTouched() [04:19:11] but that would show things like null edits and log actions too, I believe [04:19:24] I only need content changes [04:19:32] Hello, is anyone else experience problems downloading the latest mediawiki http://download.wikimedia.org/mediawiki/1.12/mediawiki-1.12.0.tar.gz? The download doesn't start for me [04:20:08] Thanks in advance [04:20:39] works fo rme [04:20:45] wfm [04:20:47] tried a mirror? [04:21:14] I googled for a mirrow but couldn't find one [04:21:37] Could you please send the link to a mirror? [04:21:54] *Splarka wonders if there are any [04:22:31] http://coblitz.codeen.org/download.wikimedia.org/mediawiki/1.12/mediawiki-1.12.0.tar.gz [04:22:33] try that [04:23:03] Thanks OverlordQ, Splarka, Hawson [04:23:08] also http://www.washington.edu/computing/web/publishing/mediawiki-1.12.0.tar.gz [04:23:32] Got it [04:39:58] 03river * r36537 10/trunk/tools/tsutils/ (7 files in 3 dirs): days2date, date2days: convert between yyyy-mm-dd and days since 1970 [04:51:08] sadly, no luck so far. there are a few functions for classes that I'm not using, and not quite sure how to import them either. [05:00:24] Is there a hook which is called each time a link is encountered in parsing, or else an official regexp for finding links, or something? [05:02:50] you want the link (as it appears I would guess) in the wikitext and not the link as is registered (for a given page) in the db? [05:04:12] Splarka: I think so... [05:05:38] well, for example, if you have [[Foo|foo]] and [[Foo|bar]] on a page... [05:05:47] oh, right [05:05:49] in the links on the page you'll just get foo listed once [05:05:52] I want them as they appear in the wikitext. [05:05:55] right [05:06:20] well, you could pass them through the preparser (outputting the XML parse tree) and strip out most of the ambiguity [05:06:35] er, pass the wikitext that is [05:07:00] it kind of blows my mind that we have such a thing. [05:08:10] So, $wgArticle is a global that I can import, right? [05:08:15] yes. [05:08:18] hrm. [05:08:23] welcome to MW. [05:08:38] *Splarka looks for the page [05:08:46] I'm trying to call $wgArticle->getTimestamp(), but it's throwing the error: [05:09:02] "Trying to get property of non-object" [05:09:06] davidmccabe: http://www.mediawiki.org/wiki/Preprocessor_ABNF [05:09:10] (at such-and-such line) [05:09:17] Hawson: where are you calling this from? [05:09:30] Hawson: $wgArticle may not exist yet if you didn't use a proper entry-point. [05:09:34] an extension. and I've got global $wgArticle [05:09:45] paste? [05:10:02] it's a hacked up copy of GraphViz. One sec [05:10:37] What does it mean for a link to be nested? [05:11:09] david: what context? [05:11:13] will paste the relevant parts... [05:11:20] I mean, some "link" syntaxes are nested [05:11:20] Splarka: that grammar seems to allow nested links. [05:11:33] Hawson: in a pastie, not here, obviously. [05:11:36] [[Image:Foo|10px|thumb|A picture of a [[Foo]] seen here.]] [05:11:49] Splarka: oh, gotcha. [05:12:18] http://rafb.net/p/cn28RE32.html [05:12:24] davidmccabe: of course. :) [05:12:40] it's bombing out on line 49 [05:12:45] you might pokle tim about how to find links in wikcode in the same way the processor does [05:12:48] (49 of the pastebin) [05:13:20] I've clipped out lots of the "computation" code--stuff not related to the problem at hand [05:13:43] hm, that's weird. [05:13:46] aye [05:14:12] all I want to do is compare the last modified time of the current page, to the timestamp on a file on disk [05:14:28] this check should be done each time the page is rendered [05:14:48] It doesn't seem like a hard problem [05:14:56] nope. [05:15:08] the full message in error_log is: [05:15:12] Sun Jun 22 01:05:23 2008] [error] PHP Notice: Trying to get property of non-object in /www/htdocs/wiki/vel_anegia/extensions/GraphViz.php on line [49] [05:15:14] Hawson: wouldn't it be even more efficient to store the text that generates the graph in an associated file, and diff that with the next you get from the article? [05:15:39] checking the mtime of a file is a lot faster than running a diff [05:15:40] (this has nothing to do with your current prob. of course) [05:16:23] and given that I'm actually running this wiki on a Celeron 533, I actually care a bit about the speed of the checks here. :) [05:16:38] No point in rendering the graph each time [05:16:54] wow, gee. [05:16:58] and running a full checksum on the graph "source" is just as bad [05:17:19] in this case, a simple tiemstamp check is both fast, and sufficient for my current purposes [05:17:19] anyways, maybe you shhsould paste your full code. [05:17:26] coming up [05:18:30] http://rafb.net/p/SSdROd23.html [05:18:35] full code. [05:18:52] I've already made a few changes from the "stock" graphviz ext. [05:19:04] I've never been able to get it to work correctly out of the box. :-/ [05:19:22] you'd think someone would pick up maintainership. [05:19:45] try var_dumping $wgArticle and seeing what it is? [05:19:55] anyhow, it bombs on line 90 [05:21:18] just "NULL" [05:21:29] which agrees with the error message, at least [05:22:46] weird. dunno. [05:23:17] I just tried moving 'global $wgArticle' outside the file--no luck there either [05:24:07] oh, you know what? I think it's a hook thing [05:24:34] hmm...no, maybe not [05:25:12] since it has to parse each time. (I was thinking--incorrectly--that it was running when a page is updated) [06:41:42] Hmm .. works fine when I run it [06:41:48] Why is teh cronjob teh fucked? [06:44:27] Alright, should be good now [06:49:22] 03river * r36538 10/trunk/tools/ts-specs/TSnano.spec: new spec [06:49:53] *Werdna boredly stalks random people [06:52:19] *Splarka gives mwbot a bandaid [06:52:37] Was running him on the wrong server :) [06:54:35] !meh [06:54:35] --mwbot-- I don't know anything about "meh". [07:15:33] <_wooz> lo [07:24:46] mwbot, make me a sandwich [07:27:56] Krimpet: i use mwbot as a blog, see http://toolserver.org/~amidaniel/chanlogs/zero2hero/ [07:28:48] wouldn't be so bad except it separates by day: http://toolserver.org/~amidaniel/chanlogs/Splarka/ [07:29:04] which should start annoying ami at some point [07:29:18] zero2hero: ooh, I'll have to remember to recite my grocery list in-channel and print out mwbot's log. :D [07:30:28] Krimpet: no, that's from /msgs [07:46:01] twitter2wiki http://socialsynergyweb.net/cgi-bin/wiki/WorldDemocracy/2008-06-22_Talk greetings :) [07:46:51] question, can you control the apache server from your webhost remotely. Like can I use its commandline from home. [09:38:45] in a fresh install where are the user groups information stored again? [09:51:18] also i've recently isntalled mediawiki on my local wamp setup and i get this following error message when i try to use it: [09:51:19] The page isn't redirecting properly [09:51:19] [09:51:19] [09:51:19] [09:51:20] [09:51:22] [09:51:24] [09:51:26] [09:51:28] [09:51:30] Firefox has detected that the server is redirecting the request for this address in a way that will never complete. [09:51:33] [09:51:35] [09:51:37] [09:51:39] * This problem can sometimes be caused by disabling or refusing to accept [09:51:43] cookies. [09:52:00] I'm soo sorry, that wasn't ment to happen, that will teach me for not chucking it into notepad forst [10:43:18] hi, can anyone tell me if it is possible to construct a template such that when it is added to a page, it adds a different page to specified category? I've been through the help pages on categories and templates (and advanced templates), but I can't see how it can be done/ [10:44:29] "different" in what way? [10:44:53] do the two pages have any predictable relationship? like being the same title in two namespaces? one being a subpage of another? etc? [10:45:01] different as in a page that isnt the one with thetemplate on it [10:45:07] oh [10:46:30] well, what i'm trying to do is to have a template that will add a page to a category (creating it if it doesn't exist), and to a subcategory of that category, again creating it if it doesnt exist. the problem is that i don't know how to make the second category a subctegory of the first automatically [10:47:01] well "creating it if it doesn't exist" is kind of abstract [10:47:15] sorry [10:47:20] um [10:47:20] on a practical level, categories always exist, they just return nothing if they don't have any members [10:47:28] oh, okay [10:47:31] i was unaware [10:47:38] now, a category description page is independant of the category existing [10:48:01] (is a bit counterintuitive) [10:48:11] no sweat [10:48:12] so... [10:48:52] you want to add, say [[Category:variables]] and [[Category:programming]] to page [[Foo]], and have Category:variables a subcategory of Category:programming at the same time? [10:49:10] yeah, exactly [10:49:15] with something like {{cat|cat1=Category:programming|subcat1=Category:variables}} ? [10:49:57] yeah [10:50:31] I don't think that is possible [10:51:00] you can easily do it by adding [[Category:programming]] to the variables category description page, making it a subcategory [10:51:24] but not from a completely unrelated page [10:51:39] well, unless you wrote an extension specifically to do that maybe... hmm [10:51:43] sounds fragile though [10:51:45] ah [10:51:53] how would it know to update? it'd be hard [10:52:15] it's funky [10:52:32] let's say you had that working [10:52:44] and then you deleted [[Foo]] [10:52:57] would the subcat no longer be a subcat? [10:53:02] so you cant make edits to a page using wikitext on another page? [10:53:20] well, as i envisioned it, yes [10:53:23] well, you can with transclusions.. [10:53:35] if [[Target]] transcludes [[Template:Foo]] [10:53:37] but you'd need, basically... the category description page to transclude part of the article [10:53:47] then editing {{foo}} edits Target [10:53:55] heh [10:53:58] you can transclude articles [10:54:01] {{:Foo}} [10:54:06] it's an evil idea [10:54:14] the whole thing is, yes [10:54:35] I am guessing you want to semiautomate the generation of a category tree [10:54:40] heh, ill just reread the tranclusions passage [10:54:42] and put it all in the hands of the article's [10:54:42] yeah [10:54:48] you would then, of course, throw the whole markup of [[Foo]] inside [10:54:54] except for your category tags [10:54:56] well [10:55:00] you can use [10:55:16] (most people say "is there such a magicword??" at this point) [10:55:18] same-same [10:55:29] right [10:55:34] hi, how do I get Special:Popularpages to show pages from arbitrary namespaces? [10:55:36] archd: how good are you at php? [10:55:39] basically, the idea is just weird [10:55:42] it seems to be not showing those from a custom numespace I added [10:55:54] the category tree should be independent of the items categorized within it [10:56:42] Splarka: no experience at all [10:56:52] checkers: you have to define it in $wgContentNamespaces [10:57:29] hmm.. is there a way to add a page to a category by including certain wikitext in the category page? [10:57:37] thanks Splarka [10:57:48] http://www.mediawiki.org/wiki/Manual:$wgContentNamespaces [10:57:48] absolutely; just as if you added any other page to a cat [10:58:04] [[Category:This]] [10:58:17] that's the standard method [10:58:50] of course, ppl will argue my next point [10:59:02] and i'll concede that in a general reference work, it may not apply [10:59:16] but in a narrow field, i say that categories should form a tree, in the strict sense [10:59:29] and each content page belong to one and only one category [10:59:41] and each category belong to one and only one supercategory [10:59:52] i mean, say, adding [[category:variables]] to [[category:programming]] by putting something in the body of [[category:programming]] [11:00:03] my wiki suddenly got error 500 (while running perfectly) and I have no access for the Apache error log. what do I do? [11:00:04] the exception being [[:Category:ROOT]] [11:00:18] archd: yes, just like that [11:00:19] http://mirc-il.com/wiki/ [11:00:26] oh, good [11:00:27] it used to run smoothly [11:00:40] adding that exact markup [11:00:44] there [11:02:06] Xiong: it would be nice, in such a case, to have parallel category methods [11:02:20] like, a navigation category namespace, a maintenance category namespace, etc [11:02:42] but lacking that, it should be perfectly viable to have parallel trees in the strict sense [11:02:53] Dor: you can ask someone who has access to the apache log. and you can complain that you do not have access. access to error logs is important for keeping a website safe. [11:03:02] oops, I was wondering why it wasn't working, I had this: `$wgContentNamespaces[102] = 1;` instead of `$wgContentNamespaces = array( NS_MAIN, 100, 102);` [11:03:03] that is, have no two categories in the same article meet anywhere in the tree except the root [11:03:13] Duesentrieb, yes well, in shared host I don't really have the option [11:03:14] i only blorted one philosophy of the matter [11:03:22] Dor: why not? i do. [11:03:32] this sounds like it will be way too much of a headache to implement [11:03:37] philosophy is the exercycle that keeps the flab of dogmatic religion away [11:03:40] Dor: logs can be set up per vhost. which is The Right Thing [11:03:52] taking things as fact without challenging is the human faltering [11:03:56] parallel trees are okay, i guess, so long as they are orthogonal [11:04:15] i will tolerate almost anything so long as it is nicely orthogonal [11:04:35] Xiong: hehe, go and enforce that on enwiki :) [11:05:06] Xiong: anyway,"facette classification" is a well known concept [11:05:28] [[Category:Famous persons who died in 2004 by having their spleen sold on eBay]] [11:05:43] i tried that duesen [11:05:48] now i try nothing there [11:05:52] :P [11:06:01] it wasnt the extraction of the spleen, but the EBAY SALE that killed them [11:06:09] en has been taken over by munchkins and gangsters [11:06:17] i tried people to get tho do this when categories were introduced on dewiki. not a chance [11:06:18] i predicted the slide years ago [11:06:32] SMW! [11:06:35] enwiki is worst wrt categories though. close to useless [11:07:05] it's a failure of intellectual honesty and rigor [11:07:21] [[Category:Famous people]] [[Died::2004]] [[Cause of death::Spleen sold on EBay]] [11:07:25] the basic flaw at the root of the entire effort was the assumption that people are generally rational [11:07:48] untrue [11:07:56] Wikipedia only works in practice [11:08:00] or that the general population has training in logic [11:08:04] Xiong: well, "rigorous" being the opposite of "flexible", it's not always the right thing... [11:08:04] there was no assumption except "can't be any worse than what we tried" [11:08:47] Hi! [11:08:47] I'm going explaining you my problem... :) [11:08:51] disagree; certain ad hoc assumptions became enshrined, if not as axioms, then as moral principles to be defended to the bitter end [11:09:24] My wiki is http://wiki.rockciclopedia.com I'd like to don't show the categories in the main page, how can I do? I've tried to add __HIDDENCAT__ to the categories pages but it doens't work... look at the main page... How can I fix this problem? [11:09:50] only the main page? [11:09:54] __HIDDENCAT__ is new, 1.13 only i think. [11:09:55] *Xiong looks [11:10:04] and it would apply to all pages where the category is used [11:10:16] LolliRock: the easiest way is to hide the category bar using css [11:10:31] LolliRock: the body tag has a per-page css class, which can be used to make per-page styles [11:10:38] wikipedia uses it to hide the title of the main page [11:10:39] Duesentrieb ok and is it possible to hide the category bar only in the main page? [11:10:45] yes, see above [11:10:47] !css [11:10:47] --mwbot-- To change styles for your wiki, go to one of the MediaWiki:xxx.css wiki page and put your custom styles there (sysop/admin rights required). MediaWiki:common.css is for all skins and should be used for content styles. MediaWiki:monobook.css is for the MonoBook skin (default), etc. For more information, see !skins and [11:11:14] I see a discussion about logic and ethics [11:11:18] and mention of axioms. [11:11:19] ok... and what code Do i Have to put? [11:11:20] yay :) [11:11:22] somethign like body.page-Main_Page #categories { display:none; } [11:11:26] check the class names/ids though [11:11:26] http://en.wikipedia.org/wiki/Common_Pipistrelle#catlinks [11:11:26] lollirock, you've messed up account creation, it goes to your forum [11:11:34] shh brion, go back to stalking us [11:11:39] er, werdna [11:11:51] Xiong: what? [11:12:01] http://en.wikipedia.org/wiki/Winston_Churchill#catlinks [11:12:32] body.page-Main_Page #catlinks {display:none;} if your main page is [[Main Page]] [11:12:33] when i attempt to create an account on your wiki i'm diverted to a forum instead [11:12:36] (work form there) [11:12:56] http://wiki.rockciclopedia.com/index.php?title=Speciale:Entra&type=signup&returnto=Pagina_principale [11:13:03] Xiong: Yes, 'cause when a user register to my wiki is registered automatically to my forum also [11:13:09] http://en.wikipedia.org/w/index.php?title=Black_Rat&diff=220856615&oldid=220848003 [11:13:09] Xiong: I use an extension to do that [11:13:10] don't want that [11:13:28] why don't you want that? [11:13:30] have you also buggered up the user pages? [11:14:03] it's not quite correct to say a user signing up onwiki also signs up on forum [11:14:06] Xiong: do you take offense to people customizing their mediawiki beyond recognition and then asking for support on it? ^_^ [11:14:19] rather, he's forced to signup on forum and perhaps that signs him in onwiki [11:14:24] Xiong: you're free to not register on my wiki [11:14:29] um, to some extent, yes [11:14:29] simple [11:14:50] but when i go to help, it makes it tough if i can't go about it using standard techniques [11:15:25] hey, it's your universe, do as you like [11:15:50] but then it can be difficult for guests and visiting firemen to find their way around [11:16:03] hmm.. I'm lazy. I'll do the abuse filter log first. [11:16:52] @Xiong: Mmh... I think I'll open a poll on my site to let users deciding if that is right or wrong... [11:17:38] i happen to think that monobook is a very difficult skin, hard for humans and hard for my browser [11:17:51] so first thing i do when checking a wiki is login and change to cologneblue [11:18:01] Mmh.... I've tried to add body.page-Pagina_principale #categories { display:none; } but it doesn't work [11:18:12] Pagina_principale 'cause my wiki si italian [11:18:19] WWW -> Wiki Wide Web? heh [11:18:20] but then, on my wiki, i've left monobook as an option, because i know other visitors may expect it to be available [11:18:23] 03(mod) #globalWrapper class is not applied correctly when specific page content is entered - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=14582 (10danny.b) [11:18:25] no matter my personal feelings [11:19:04] What about a universal cookie storing an array with your prefered skins... first one available is selected... heh [11:19:24] would be very nice -- also interface language, of course [11:20:11] Xiong: um, what? [11:20:14] Xiong: why not use a greasemonkey to use and cascade &useskin=cologneblue? [11:20:17] The trick would be the ss-caches... We'd have to make sure squid/varnish could actually understand the values in these cookies and cache based on their tokens [11:20:17] Accept-Language: en-au? [11:20:33] Splarka: no? [11:20:34] :) [11:20:42] neither body.page-Pagina_principale #catlinks {display:none;} or add body.page-Pagina_principale #categories { display:none; } work!! :( [11:20:42] LolliRock: #catlinks [11:20:52] i don't like messing with the browser bar address; i often tend to edit it by hand [11:20:55] I don't think #categories is standard (unless in older versions?) [11:21:18] LolliRock: .catlinks [11:21:18] i've tried with #catlinks also... [11:21:18] http://wiki.rockciclopedia.com/index.php?title=MediaWiki:Monobook.css [11:21:19] Werdna: just put away the boomerang and it's a deal [11:21:25] Though, the tricky part is a global cookie has potential for it to be messed with by any bad site... anyone could just delete your cookie prefs, or change them to one you don't want, just by visiting their site [11:21:39] with the . (dot) instead of the # ? [11:21:42] yes. [11:21:52] NO [11:21:52] either [11:21:58]