[12:31:38] Duesentrieb: what is the solution then ? the problem with the definition list, is there is no easy way to style it without wrapping a div around it. [12:32:04] 03(mod) Special page for recent changes in unwatched articles - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=13063 (10jlerner) [12:32:15] Duesentrieb: you dont think indented text should be marked as so with some actual markup ? [12:32:17] exobuzz: how would that be better with a blockquote? [12:32:28] aww [12:32:38] >> Indent? [12:32:39] e7th04sh: hold on a sec. it's a css hack [12:32:52] exobuzz: html is suposed to be semantic. it does nto have markup for "indent". [12:33:20] exobuzz: an indentation is nither a quite, not a definition. so

and

...etc [13:17:33] and some magic transclusion things, [13:17:46] and some formatting things, , [13:17:49] onlyinclude? [13:18:14] Werdna: because I thought it worked :) Anyway, I'm writing a better merging function now [13:18:22] xml tags throw a whole 'nuther monkey wrench in,
$etc
[13:18:43] you make a very good point... you think multi line input to a parserfunction looks better... so it shall be
[13:19:04] also bad are the ambiguities twixt {{template}} {{magicword}} {{parserfunction}} {{{templateparameter}}}
[13:19:22] VasilievVV: what's wrong with array_merge?
[13:19:24];) The nice thing about parserfunctions is rather than multiline input, you use parameters
[13:19:24] *Splarka shrugs, personal opinion
[13:20:26] in what cases do i get the corrupted file upload error?
[13:20:30] Oh god... nowonder EditPage is so horrid
[13:20:35] so they're both bad?
[13:20:38] i tested it now with two mediawikis...and its the same problem
[13:20:42] s are good for things that generally need to preserve whitespace, linebreaks, and integrity without accidental parsing, like pre, poem, math, hiero, imagemap
[13:20:54] i even increased the limitrequestbody in php.conf
[13:21:32] #parserfunctions are good for... things that depend on wikicode, {{#ifexist:{{TALKSPACE}}:{{PAGENAME}}|[[{{TALKSPACE}}:{{PAGENAME}}|talk page]]|no talk page}}
[13:21:44] Werdna: it doesn't handle subarrays. E.g. if default settings have $wgGroupPermissions has key 'sysop' and overriden $wgGroupPermissions has key 'sysop', the eariler will be overriden by the latter, not merged
[13:21:49] since you seem to want to pass parameters to it, you *can* use #tag, but seems easier to just make it a pareserfunction
[13:22:00] what some people do is make both, #foo and [13:22:08] jlerner: thx
[13:22:31] *Shaiaqua starts looking up how to make a parserfunction
[13:23:40] http://www.mediawiki.org/wiki/Manual:Parser_Functions
[13:23:50] er
[13:24:06] Help:Parser Functions
[13:24:08] I think.
[13:24:09] an example anyway
[13:24:16] Werdna: may you show me a configuration, under whcih you did actually test r39582?
[13:24:37] what's r39582?
[13:24:43] this has been my test case, most of the time
[13:25:01] Shaiaqua, start here: http://meta.wikimedia.org/wiki/ParserFunctions
[13:25:02] Werdna: interwiki userrights changes
[13:25:24] hello
[13:25:30] (no commit this time, Splarka)
[13:25:37] hah
[13:25:40] Lady_Aleena: um...
[13:26:04] Splarka, I know I know it's a redirect
[13:26:14] VasilievVV: I didn't test it directly. I tested each component of it. I can see the problem, though.
[13:26:16] wrong wiki, wrong topic...
[13:26:16] Erm...
[13:26:28] that is about the extension called ParserFunctions
[13:26:29] Lady_Aleena: We're talking about writing
[13:26:29] I used this to test the siteconfiguration stuff
[13:26:31] $c = new SiteConfiguration; $c->settings = array( '+x' => array( 'foo' => array('fred' => 'bar'), '+home_wiki' => array( 'george' => 'boo' ) ) ); $x = array( 'baz' );
[13:26:40] Oh, I will shut up now.
[13:26:44] and then used debugging output statements to verify that everything else ws going okay.
[13:26:52] Werdna: that's why you got broken-by-design feature which didn't handle anything correctly
[13:27:08] it's not broken by design.
[13:27:28] It misses some impotant parts
[13:27:29] It's broken because it needs to use a better array merging function.
[13:28:26] Werdna: not only. Read http://lists.wikimedia.org/pipermail/wikitech-l/2008-August/039056.html
[13:28:54] $magicWords['example'] = array( 0, 'example' ); what is the zero?
[13:29:59] The first item in the array is case sensitive
[13:30:23] anyone ever tried the flowplayer extension?
[13:30:24] The 0 is used as a placeholder to bump things over so that 'example' won't be case sensitive
[13:30:27]$
to split a section in two columns as described here : http://www.mwusers.com/forums/showthread.php?p=27736 [13:46:48] Werdna: $wgConf will, if we use it correctly [13:46:54] but next section starts in the right column first, and only later flows to whole width [13:47:07] what would be the best way to force next section after columns ? [13:47:16] I wouldn't want to risk making some established snafu by just replacing the code with my mod and assuming ownership without checking in here first :-P [13:53:55] hello, what's the best solution for php5 + apache2-mpm-worker on debian? [13:54:21] recompiling php5? using php5-cgi? [13:56:15] T_T EditPage::getPreviewText horrifys me [13:57:26] Man I wish EditPage::getPreviewText was protected... [13:57:41] VasilievVV: can you look at this? http://pastebin.ca/1179644 -- it doesn't seem to work properly. [13:58:10] it doesn't work properly. [13:58:20] Werdna: array keys are usually not arrays [13:58:32] yannf: could recompile. be careful with what modules you use not all are thread safe. easier to use fcgid probably and php-cgi [13:59:09] VasilievVV: Where do I make that assumption? [13:59:20] $out[$key] is not an array key. [13:59:30] Hmmm [14:01:35] It works for me [14:02:00] *SignpostMarv wasn't aware PHP supported arrays as keys [14:02:34] thought it was scalar values only ? [14:02:38] VasilievVV: http://pastebin.ca/1179646 [14:03:28] SignpostMarv: it is [14:03:48] exobuzz, yes, i'd recompile if possible: so does any of the problematic php modules are needed for mediawiki? [14:03:51] Werdna: mixing numeric and string indexes is a very bad idea [14:04:13] VasilievVV: ooh, that's it. [14:04:16] stupid, stupid, stupid. [14:04:27] well spotted [14:05:21] 0 != '0' [14:05:34] if both are present, anyway [14:06:25] SignpostMarv: not really. For me, "0" == 0 [14:06:40] SignpostMarv, in PHP, 0 == '0'. [14:06:50] VasilievVV: heh, I know, I fixed a bug in wiki sets with that error a few hours ago :P [14:07:03] ah wait, it's false it's not equal to, since a non-empty string equates to true, yes ? [14:07:08] $ php -r 'var_dump( 0 == "0" );' [14:07:08] bool(true) [14:07:10] you couldn't remove a wiki set from a global group [14:07:24] $ php -r 'var_dump( (bool)"0" );' [14:07:24] bool(false) [14:07:32] SignpostMarv, welcome to PHP. [14:07:49] I hope you enjoy your stay. But I doubt it. [14:07:59] I'm quite familiar with PHP, I'm just using === and !== because of the loose typing quirks [14:08:14] 0 != '0' [14:08:23] That's not right. 0 !== '0' is right, sure. [14:08:46] aah this one again [14:09:03] 03werdna * r39701 10/trunk/phase3/includes/ (GlobalFunctions.php SiteConfiguration.php): [14:09:03] Use a nice fancy deep-merge function for merging arrays in SiteConfiguration [14:09:03] (newly written), and move the merging of config options and globals deeper into [14:09:03] get(), so we don't need to duplicate the same functionality in the entry points. [14:09:07] i have the answer you know.. .. php is sh*t.... [14:09:10] :) [14:09:26] But that doesn't help you for array indexes. $arr = array(); $arr[0] = 'foo'; $arr['0'] = 'bar'; var_dump( $arr ); [14:09:27] array(1) { [14:09:27] [0]=> [14:09:27] string(3) "bar" [14:09:27] } [14:09:34] Yay PHP! [14:09:57] SignpostMarv: comprasions aren't the only point. All standard functions messes up arrays with numeric strings. Try var_dump( array_reverse( array( '4' => 1, '5' => 2 ) ) ); [14:10:39] I believe that's what uasort() was invented for :-P [14:10:46] I think the big mistake with php happened when someone went "I want to build something to help me with my website.. i know. ill call it personal home page" and it was all downhill from there [14:10:53] I believe that's what real languages like Python were invented for. [14:12:01] i love that bug from earlier.php -r "strtotime('+1000000000000 hours');" only php can be so great. [14:12:02] ^_^ Try comparisons in Lisp, they're even more confusing' [14:13:05] 03werdna * r39702 10/trunk/phase3/includes/ (GlobalFunctions.php SiteConfiguration.php): *ahem* self-revert. Something weird happened with getAll() [14:13:13] 03(mod) Watchlist categories - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=5875 (10jlerner) [14:13:16] far too late to be coding. [14:13:17] exobuzz, isn't strtotime() just a wrapper around a GNU library function? [14:13:17] night, all. [14:13:58] Simetrical: dunno [14:14:36] i like php more than cobol though.. that has to be said [14:14:59] More than BASIC. [14:15:05] More than, oh, punch cards. [14:15:07] Not saying much. [14:15:11] im not one of those people who think that computer languages should be like written/spoken languages [14:15:26] I think Python does that well. [14:15:28] I added 'mm' to $wgFileExtensions -- directly in includes/DefaultSettings.php -- I tried uploading such a file, and got an U;load warning: "The file is corrupt or has an incorrect extension. Please check the file and upload again." [14:15:34] It's readable without being verbose or hard to predict. [14:15:45] Like using "and" instead of "&&" and "not" instead of "!", that kind of thing is nice. [14:16:02] Simetrical: im all for readable. i mean like cobol where its "ADD SOMETHING TO SOMETHING AND SUBTRACT SOMETHINGELSE" [14:16:12] SQL too. [14:16:14] Simetrical: & is a common operator [14:16:19] The problem with that is that it's verbose and hard to predict. [14:16:35] and/or do work in php and most other languages you know [14:16:38] VasilVV, did I say it wasn't? [14:16:50] ^_^ (and foo bar) [14:17:00] Simetrical: that means it's readable [14:17:22] i like to do "unless something {}" instead of "if (! something) {]" in perl [14:17:23] :-) [14:17:28] VasilVV, no it doesn't. Perl is also common, that doesn't mean it's readable. [14:17:52] ^_^ (when ...) (unless ...) [14:17:58] exobuzz, that does seem like an interesting construct, but ultimately not a very necessary one. [14:18:02] 03siebrand * r39703 10/trunk/extensions/Translate/Translate.i18n.php: Tweak 'translate-tag-noedit' [14:18:08] i wrote a nasty piece of perl earlier. i will functionalise it and clean it up but.. here we go.. http://malus.exotica.org.uk/~buzz/ft.txt [14:18:20] parsing a binary file in perl..ehm [14:18:25] !wgwgFileExtensions? [14:18:25] --mwbot-- I don't know anything about "wgwgfileextensions?". [14:18:30] !wgFileExtensions? [14:18:30] --mwbot-- I don't know anything about "wgfileextensions?". [14:18:42] :/ [14:18:48] !wg FileExtensions [14:18:48] --mwbot-- http://www.mediawiki.org/wiki/Manual:%24wgFileExtensions [14:19:10] Simetrical: perl has lots of stuff like that. but so does php. php is so function bloated.. [14:19:36] Python doesn't. :) [14:19:51] why preg_match and preg_match_all. wtf is wrong with one function with a parameter ? [14:20:04] (loop for i from 1 to 10 summing finally (format T i)) ^_^ [14:20:05] Simetrical: python is on my "to do" list :) [14:20:08] How about ereg_match? [14:20:17] And strpos. [14:20:24] yeh strpos.. wtf.. [14:20:25] And . . . how many other variants are there? Five, ten? [14:20:34] Which all do basically the same thing in slightly different ways? [14:20:46] do we need split and explode ?. why are parameters always different orders [14:20:47] Okay, separating simple string functions from regex ones is sensible enough. [14:20:55] Dantman: "doesn't work" for whatever reason. [14:21:01] split and explode do two different things [14:21:22] And preg_split? [14:21:54] :/ What's wrong with preg_split other than taking up memory just like anything else that uses arrays? [14:22:14] i know there are differences. but thats the thing.. other languages make do with one function which does what you need, php needs 3 or 4 with slight differences [14:22:28] exobuzz, Python does have separate regex functions. [14:22:28] what about join and implode then ? [14:22:33] Because they aren't a core part of the language. [14:22:34] they are the same [14:22:45] Or die() and exit(). [14:22:53] Or &&, and, AND. [14:23:01] Simetrical: it is different in that cas.e i mean the problem is php has everything in the core. it should be split out [14:23:10] yeh [14:23:13] exactly [14:23:37] 03(mod) Indicate where the thread was moved to on placeholders - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=15133 (10minuteelectron) [14:23:44] Except that since every extension is a C module, you need to be root to install it, and therefore third-party authors can't rely on the presence of any optional modules. [14:23:46] 03(mod) On thread history page move items, indicate target and source - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=15134 (10minuteelectron) [14:23:54] 03(mod) Visually subdue deleted replies - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=15135 (10minuteelectron) [14:24:07] Whereas in, say, Python, most modules are loaded as-needed by import statements, and the modules are usually written in Python anyway. [14:24:11] 03(mod) Let users choose to hide deleted threads\replies - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=15136 (10minuteelectron) [14:24:19] mysqli_escape_string mysqli_real_escape_string etc.. do we need two calls depending whether it looks at the characterset ? [14:24:20] bah [14:24:21] 03(mod) On archive pages the radio button for "Any date" is not shown - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=15137 (10minuteelectron) [14:24:24] backwards compatibility sucks [14:24:28] 03(mod) Format dates in a more attractive manner on archive pages. - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=15138 (10minuteelectron) [14:24:30] exobuzz, if you agree that &&, and, AND shouldn't all be allowed, and you currently use Perl, you'll like Python. :) [14:24:36] 03(mod) Show a warning message, or optionally prevent, users from editing each others replies - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=15139 (10minuteelectron) [14:24:45] That's what Python 3 is breaking reverse compatibility. \o/ [14:24:52] In a reverse-compatible way, though. :) [14:25:00] Simetrical: i will have a look [14:25:05] let's recode mediawiki in python [14:25:08] would be faster! [14:25:09] :D [14:25:22] ill start.. gimme a few days [14:26:52] If only Brion would take that idea seriously. :( [14:26:59] do_you_think_there_is_a_competition_with_php_coders_on_who_can_make_the_longest_function_call() ? apart from one guy.. whos name is FrdBlgs who prefers short functions [14:27:40] whose [14:30:11] exobuzz: yes -- they don't want to be confused with perlmonks [14:34:30] 03vrandezo * r39704 10/trunk/extensions/SemanticMediaWiki/includes/SMW_GlobalFunctions.php: aligning code [14:41:23] Hi guys, can i get here some info on how to modify some parts of mediawiki? [14:42:12] When you ask mor specific, yes [14:44:12] ok. First of all, I would like to know, how to enable characters like _ in user name... [14:45:42] I have already removed the rule, when first character of any user name is capital, but can't find where to modify, to let users use _ in thir user names [14:46:04] thegr0m: I fear _ is not possible. all _ will be substituted to ' ' (space) [14:46:54] Ugh... T_T that kind of stuff is ugly business without my title rewrite [14:46:56] 03(mod) divs with display: inline still have paragraphs added around them ( are treated as block-level by the parser) - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=15246 summary; +comment (10Simetrical+wikibugs) [14:47:20] ^_^ You'll have to poke/pay me enough to ever get that done [14:47:35] 03(mod) Definition list abused for indentation purposes - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=4521 +comment (10Simetrical+wikibugs) [14:47:43] thegr0m: and how do you have removed the rule, that first character of any user name is capital? [14:48:35] http://moeffju.net/blog/2006/06/14/case-sensitive-mediawiki/ here's the code. But for earlier versions. I have modified it a bit and now it works [14:48:58] This is why I think, that everything is possible. But I can't understand where to start... [14:50:31] The funny thing is that, that if you try to login with wrong name, in the message you can see all this _. But it is not compared with database. It is getting all _ away, before compare... [14:51:15] quick poll: if you see Google adsense on a wiki, are you turned off (1) not at all, (2) not enough to leave if curious about the wiki (3) significantly (4) severely (leaving forever)? [14:52:59] How to allow _ only in user names? :( [14:53:41] Smaug, (1)-(2) here. [14:55:04] thegr0m: patch the code. it is not possible out of the box. and you will get a lot of sideeffects. I propose not to do that [14:55:08] thegr0m: wiki page name normalization is applied to user names. that means all _ are converted to speaces [14:57:29] Yes, I understand, that page normalization is applied to names, but how to disable it only for user names? I don't have open registration. I'm trying to make cross-server registration with one online browser game, where from I can get only few things over the GET request... [14:58:11] Simetrical: ty for the response [14:59:03] I already did a parser of those things and registration to database. But there are nicknames, containing _ and ()... Everything is registering fine. But I can't login, be cause it is cutting off _ [14:59:05] hello [15:01:14] Can somebody please point me, atleast, on this function, which is cutting off all illegal characters? [15:06:41] hey all, was wondering if anyone has had the chance to work with semantica? [15:07:37] semantica? [15:09:15] yup [15:09:21] http://www.semanticresearch.com/products/semantica-wiki.php [15:12:52] "Use open standards such as Wikimedia and KML-based GIS viewers to push and pull data from multiple sources quickly and easily" [15:13:00] that sounds like complete and utter bullshit [15:13:12] Wikimedia is an organization, not a standard. [15:13:22] "KML-based"? [15:13:25] very basic question: once I change a setting in LocalSettings.php, how do I "apply" the changes? [15:13:41] e.g. get them to take effect [15:13:41] Ah, I see, that's a real standard. [15:13:48] mpuccio, they take effect immediately. Reload the page. [15:13:54] thanks [15:14:16] the use of "Wikimedia" there is even trademark abuse, if they don't have a contract with Wikimedia Inc. [15:14:53] 03(mod) Mediawiki categories User-Javascripts - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=14573 +comment (10dan_the_man) [15:15:22] 03gri6507 * r39705 10/trunk/extensions/WhiteList/SpecialWhitelistEdit.php: test check in [15:16:19] Duesentrieb, Simetrical, so as i see it, it's pretty much bs? [15:18:50] Hi all [15:19:10] I have qaustion [15:20:03] !ask | Sterkebak [15:20:03] --mwbot-- Sterkebak: 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. [15:20:27] I wan't to update . I will Just overwrite all file's but then i must open siteadmin.php? [15:20:55] there is no file by that name in a standard mw install [15:21:03] Or must i do more? [15:21:17] How did you install MW? [15:22:41] I did not [15:23:01] I have got it and now i have to update [15:23:41] I am the new nwoer [15:23:48] Owner [15:24:49] So if i overwrite all files the files will be uptodate. [15:25:36] But then i must update the database [15:25:51] How must i do that [15:28:03] 03mkroetzsch * r39706 10/trunk/extensions/SemanticResultFormats/: Add base directory for new SMW extension to collect extra query result formats not shipped with SMW [15:29:42] php maintenance/update.php [15:29:57] You need access to the command line (Look for "SSH" access on your host) [15:30:11] Ack, tim is gone [15:33:10] 03(mod) Preview with forms on the bottom - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=10874 (10nathan) [15:35:18] but next section starts in the right column first, and only later flows to whole width [15:35:22] what would be the best way to force next section after columns ? [15:49:36] 03vrandezo * r39707 10/trunk/extensions/SemanticResultFormats/README: Initial README [15:55:31] Hi brion [15:56:00] *Jack_Phoenix waves to brion and ialex :-) [15:56:22] hello Jack_Phoenix :) [15:56:25] rar [15:58:00] wazzap [15:58:07] hi brion :) [15:58:37] hello [15:58:47] Werdna: for r39701: the problem might be that GlobalFunctions.php is loaded after executing LocalSettings.php [15:58:59] hello Nikerabbit [16:08:50] Can someone point me on that file and function, what is replacing _ with spaces? THanks [16:09:04] in names, I mean... [16:10:02] str_replace ? ... [16:10:16] oh [16:10:53] it's probably the same as for the titles, so be careful [16:12:53] DarkoNekom thanks, but I mean, where is this function in MW :) [16:13:17] I just want to know how is it working and what exactly replacing... [16:14:01] thegr0m: Title::new*(), Title::secureAndSplit(), all in includes/Title.php [16:14:02] 03brion * r39708 10/trunk/phase3/ (3 files in 3 dirs): [16:14:02] Revert r39416, 39421, 39427, 39489, 39515, 39516 "Rewrite allpages to use a "tree" of index range subpages (similar to the starting page) rather than one increasingly large range list (bug 13902)" [16:14:02] Breaks several aspects of paging behavior: [16:14:02] * Subpage parameter seems to have incorrectly changed from starting point to prefix [16:14:05] * We've lost forward and backward paging links [16:14:23] ialex, thanks. I'll check it... [16:14:44] thegr0m: and maybe a lot of other places too ;) [16:14:53] :) [16:16:01] I Just want to know what exactly is it removing, when registering a new user... For example if I'll enter _thegr0m_, it will return just thegr0m. Even without spaces. And I need to know what else will it replace, and how is it removing smaces from begining and from the end... [16:22:46] ialex, thanks. That helped. Found everything, I was looking for :) [16:24:59] Hi, i asked this before, but due to unfortunate event i lost the information i received. Could someone give me again information on how to hide page title via this css hack? [16:25:28] 03rotem * r39709 10/trunk/ (2 files in 2 dirs): Localization update for he. [16:28:00] 03(mod) Link with action=edit&redlink= 1 leading to existing page should redirect to plain view - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=15242 +comment (10mormegil) [16:30:04] 03vrandezo * r39710 10/trunk/extensions/SemanticMediaWiki/includes/ (SMW_GlobalFunctions.php SMW_QueryProcessor.php): Moved SMW_QueryProcessor::$formats to $smwgResultFormats [16:40:58] brion: does the hack in Skin.php to "Hide the styles from Netscape 4 without hiding them from IE/Mac" is still needed? [16:44:07] ialex: is that a hack that we killed last month? or a different hack? [16:44:26] brion: the CSS hack [16:45:08] Skin::getSiteStyles() and Skin::getUserStyles() [16:45:49] the point is that SkinTemplate based skins don't have it [16:48:29] I wonder, is there some why to bind a function to a hook, without parsing *whole* function code unless hook happens? [16:49:33] way* [16:49:43] my brain rots [16:52:52] autoloading [16:52:57] put the function in a class [16:53:03] and then put the class into the autoloader [16:53:08] thanks [16:53:13] then make the hook the callback [16:54:46] ialex: my recommendation is to ditch all the old-style skins and replace them with a consistent SkinTemplate-based system [16:55:03] i updated the CSS loading on SkinTemplate to use consistently but haven't touched the old skin code as it's nasty [16:55:35] brion: i did that for non-SkinTemplate skins [16:55:52] (using for CSS styles) [17:18:21] Hi [17:18:32] Can i change userrights from sysop to siteadmin in the database? [17:20:31] siteadmin is a right, not a group [17:20:56] unless you created a new group using $wgGroupPermissions [17:22:11] ialex how do i give myself the right from the database? [17:23:02] Sterkebak: you don't, you need to assing it to a group using $wgGroupPermissions and then you can assing the group in the database [17:24:05] 04(REOPENED) Postgres - Searches that include a number as a term always return no results - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=15229 summary; +comment (10brion) [17:24:16] 03(mod) PostgreSQL/pgsql support (tracking) - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=384 (10brion) [17:24:52] 03(mod) can't search for numbers (no numerical search) - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=42 (10brion) [17:26:19] 03nikerabbit * r39711 10/trunk/extensions/Translate/tag/Hooks.php: * Bugfix [17:28:45] ialex is there a way to give me other rights from the database then? [17:29:08] Sterkebak: as I said, only groups [17:29:40] But as i look at the groups i don't see the group Toezichthouder :( [17:29:48] But that is my admin group [17:30:54] brion: oy [17:31:10] *AaronSchulz redoes allpages again [17:33:14] Sterkebak: Toezichthouder is the displayed name, internal name for the sysop group is "sysop" (the one that will appear in the database) [17:33:32] okey i have three of them :s [17:33:43] I will drink a cup of coffee and try again. [17:33:46] yo yo [17:33:51] *brion is caffeinated \o/ [17:33:53] Thanks for your time :) [17:34:14] np :) [17:34:43] Caffeine is the work of the devil. [17:36:19] devil, hehe [17:37:39] more dangerous by far than the devil is the superdevil! [17:37:44] hurm. 15229 wfm, but then again I"m using a current version :( [17:45:01] 03(mod) LinkSearch results should use as much of the path as is provided, not simply search by domain - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=15218 (10brion) [17:45:20] Werdna: https://bugzilla.wikimedia.org/show_bug.cgi?id=15219 [17:47:05] 03(mod) Disable renaming of wiki sets to already existing names - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=15219 (10brion) [17:47:35] what's the purpose of patrolmarks? [17:47:43] http://en.wikipedia.org/wiki/Wikipedia:Village_pump_(technical)#.22go.22_button_in_watchlist <-- Hmm, agree, looks kind of ugly. [17:47:44] being able to see which edits are patrolled? [17:50:26] 03aaron * r39712 10/trunk/phase3/ (3 files in 3 dirs): redo allpages again less obtrusively [17:50:58] 03(mod) Allow .ttf files - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=15195 (10brion) [17:51:11] brion: better? [17:51:14] Is there an example I can view somewhere that will show me how to use the #if parser function (in a template) to show a field if a value has been entered, and not show the field if there is no value in it? [17:51:22] 03greg * r39713 10/trunk/phase3/maintenance/updaters.inc: Update to match postgres/tables.sql [17:52:14] I have been reading http://meta.wikimedia.org/wiki/Help:ParserFunctions, but I still can't get it to work [17:52:16] AaronSchulz, what's the purpose of patrolmarks? [17:52:17] being able to see which edits are patrolled? [17:52:35] yes [17:52:47] and patrol to view and set as patrolled, ok [17:52:57] 03nikerabbit * r39714 10/trunk/extensions/Translate/tag/ (Hooks.php Tag.php Utils.php): * Bugfixes [17:53:37] it's not documented anywhere [17:53:42] and [17:53:42] 'wgOnlySysopsCanPatrol [17:53:50] appears on meta and config! [17:54:16] *Platonides mumbles about removed options confusing people [17:54:33] for example: I have "{{#if: {{{example-field|}}}||}}" ... the field name shows up whether the field is filled out or not, and I also get "{{#if: ||}}" showing up on my page [17:55:20] do you have parserfunctions? [17:55:45] if it's something I had to add, then I guess not. I thought it was built into mediawiki :\ [17:55:51] 03ialex * r39715 10/trunk/extensions/OnlineStatus/OnlineStatus.php: [17:55:51] Tweaks for OnlineStatus: [17:55:51] * Fix PHP5.3.0aplha1 errors [17:55:51] * Hide "Status" tab in personal urls in Special:Preferences, there's already a switch on that page [17:56:45] 03(ASSIGNED) Special:BlockIP said to be broken in PostgreSQL - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=15148 +comment (10greg) [17:57:57] !parserfunctions | Seiken [17:57:57] --mwbot-- Seiken: "Parser functions" are a way to extend the wiki syntax. ParserFunctions is an extension that provides the basic set of parser functions (you have to install it separately!). For help using parser functions, please see . For details about the extension, see . [17:58:00] brion: I really can't stand allpages with the uber long lists that lag firefox [17:58:28] 03(mod) can't search for numbers (no numerical search) - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=42 (10reed) [17:58:43] Platonides: thanks, I am reading up on how to install now. [17:58:58] 03(mod) Postgres - Searches that include a number as a term always return no results - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=15229 (10reed) [18:00:53] 03(mod) Non-breaking spaces in wikilinks should act like regular spaces - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=15248 summary (10matthew.britton) [18:01:18] 03(NEW) Non-breaking spaces in wikilinks should act like regular wikilinks - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=15248 15enhancement; normal; MediaWiki: Page rendering; (matthew.britton) [18:04:52] 03aaron * r39716 10/trunk/extensions/FlaggedRevs/FlaggedRevs.hooks.php: Make isFileCacheable() faster [18:05:02] 03(mod) Special:BlockIP said to be broken in PostgreSQL - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=15148 (10greg) [18:05:16] "said to be" [18:09:39] hi all, I would like to design my main page with boxes and borders like at the main page onmediawiki.org or wikipedia.org. Do I have to do this by CSS or are there finished templates? [18:10:12] 03(mod) LinkSearch results should use as much of the path as is provided, not simply search by domain - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=15218 (10mike.lifeguard) [18:10:33] sn4ke, people usually use tables. [18:11:03] hi Simetrical [18:12:12] 03brion * r39717 10/trunk/phase3/ (8 files in 5 dirs): [18:12:12] * $wgAjaxSearch has been removed; use $wgEnableMWSuggest instead. [18:12:12] I never liked the take-over-the-screen UI of the old ajax search; the new suggestion search does the same thing in a more natural, integrated way. [18:12:29] 03rotem * r39718 10/trunk/phase3/languages/messages/MessagesHe.php: Localization update. [18:13:26] 14(INVALID) Ajax search feature does not work in modern skin - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=14444 +comment (10brion) [18:13:34] Simetrical: but wikipedia and mediawiki is made with CSS?! [18:13:43] Platonides: it works! :P [18:13:50] 03(mod) Special:BlockIP said to be broken in PostgreSQL - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=15148 (10overlordq) [18:13:51] sn4ke, MediaWiki uses CSS for most things. Wikipedia content often uses table-based layout. [18:18:47] Simetrical: but the main page of wikipedia is made in CSS as well?! or is a layout like this possible with tables? [18:19:01] sn4ke, the main page of Wikipedia, last I checked, is table-based. [18:20:01] Simetrical: ah ok, thx [18:20:32] 03aaron * r39719 10/trunk/extensions/FlaggedRevs/FlaggedRevs.class.php: tweak cache expiry [18:21:16] 03(mod) Special:BlockIP said to be broken in PostgreSQL - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=15148 (10greg) [18:25:57] 03brion * r39720 10/trunk/phase3/includes/HTMLDiff.php: fix for html diff on page with images -- ImageNode $attributes member was incorrectly declared as private [18:26:01] guyvdb: doing some quick testing w/ $wgEnableHtmlDiff before switching it on on test/en.labs for some public testing [18:26:27] i seem to be missing the diff headers, so we're missing the revision information and paging links [18:26:35] ah [18:26:39] yeah that could be [18:26:54] can you fix that up real quick and i'll go ahead and turn it on there? thx :) [18:26:59] okay [18:28:00] iirc, you can't see category and interwiki link changes [18:28:16] brion: I could also show the old diff, but that will increase the execution time even more [18:28:58] mm, well it might be good if we can conditionally switch it in for testing purposes [18:29:04] (and we need the header in either case) [18:29:31] *ialex would like an user option for $wgEnableHtmlDiff (I know that we don't like new options) [18:30:22] does anyone know, if it is possible to make internal link to another namespace? [[namespace:article]] doesn't work [18:30:47] that should work [18:31:01] 03aaron * r39721 10/trunk/extensions/FlaggedRevs/FlaggedRevs.hooks.php: Make overrideRedirect() faster [18:31:30] sn4ke: if it's category: or image: those are special-cased -- to make a regular link, put a : in front like [[:Image:Foo]] [18:32:30] brion: ah okay so [[:category:article]] is possible as well?! [18:32:52] yeah [18:33:06] ah ok thx [18:44:08] 03ialex * r39722 10/trunk/extensions/OnlineStatus/ (OnlineStatus.js OnlineStatus.php): Put ajax function in the OnlineStatus class [18:44:13] another question: subcategories and categories are apart. It is possible to take them together? [18:44:45] 03aaron * r39723 10/trunk/extensions/FlaggedRevs/FlaggedRevs.hooks.php: remove more grannyware [18:45:56] 03guyvdb * r39724 10/trunk/phase3/includes/DifferenceEngine.php: added diff headers for HTML diff [18:45:58] brion: done, r39724 [18:47:23] 03ialex * r39725 10/trunk/extensions/OnlineStatus/OnlineStatus.php: Whitespaces fix *sigh* [18:49:58] guyvdb: you should add a new message instead of hardcoding "HTML diff" [18:50:15] ;) [18:50:47] how would you translate HTML diff in another language? :p [18:50:58] yeah I need to do i18n for the differ itself too [18:51:01] it's on my todo [18:51:09] guyvdb, you could translate "diff", anyway. [18:51:22] Maybe transliterate "HTML" (although I've noticed that usually people don't transliterate jargon). [18:52:12] "diff" doesn't say anything to most of the people (neither does HTML, but that's besides the point) [18:53:14] and we should avoid jargon, so it could be like "Differences in rendered versions" or something [18:53:28] the word diff is used everywhere in the code [18:53:39] all the links say "previous diff" [18:53:41] etc [18:54:26] two bads don't make the third good :) [18:54:44] Nikerabbit, "rendered" is jargon. [18:55:12] jargon is jargon [18:55:13] "graphical" might be more comprehensible, but I'm missing context here. [18:55:24] "Jargon" is not jargon. :) [18:55:35] Simetrical: you right, but I'd imagine it would be more understandable than diff [18:55:43] Well, yes, for sure. [18:56:09] it also helps translators the more we try to avoid them [18:57:16] I'm all for changing the term HTML diff, the problem is that I don't know a suitable replacement, I'm not a native english speaker [18:57:24] Visual Compare has been sued before [18:57:31] just Compare [18:57:47] Inline Compare [18:57:57] Where's the message shown? [18:58:13] "Visual" or "graphical" is good. [18:58:21] "Visual comparison" sounds good. [18:58:24] "Inline" is bad, more jargon. [19:00:52] thx guyvdb, i'll take a look in a sec [19:01:47] 03aaron * r39726 10/trunk/extensions/ConfirmAccount/SpecialConfirmAccount.php: minor refactoring [19:02:18] is there any particular order in which the MessagesEN.php entries are? [19:02:31] no [19:02:41] just somehow related messages together [19:03:57] (and don't forget to update maintenance/language/messages.inc ;) [19:07:10] guyvdb: ok we've still got a security issue -- text is getting de-html-escaped [19:07:37] :s [19:07:49] ialex: what's ???messages.inc for? [19:07:53] example: write in a page this text "" [19:07:56] then html-diff it [19:08:23] it should show as plain text, but instead it's output as raw html and the script is executed [19:08:54] guyvdb: it lists all messages in MessagesEn.php, just add the message name at the correct place [19:09:40] 03guyvdb * r39727 10/trunk/phase3/ (3 files in 3 dirs): 'Visual Comparison' added to messages [19:11:52] guyvdb, messages.inc is used by maintenance scripts to ensure that messages are consistently grouped and ordered in all message files. [19:12:01] Actually, though, why can't you just parse MessagesEn.php instead? [19:12:09] Hmm . . . [19:13:00] 03(mod) Excessively large offset specified in {{#time:}} causes timeout - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=14898 minor->04CRIT; normal->high; +comment (10krimpet) [19:13:26] 03aaron * r39728 10/trunk/phase3/includes/Article.php: function is not static [19:13:35] brion: I don't really understand where these characters get decoded [19:13:42] the parser encodes them as special chars right? [19:13:57] then the XML parser decodes them? why? [19:14:18] apparently PHP's strtotime() can cause an endless loop if passed certain values -_- [19:15:04] what is grannyware? [19:15:10] Krimpet: php bug open? [19:15:26] Nikerabbit, yes. [19:15:32] http://bugs.php.net/bug.php?id=45822 [19:15:52] also, grannyware, is that like granny panties? :p [19:16:17] brion: htmlspecialchars on parsed characters will do the job? [19:16:25] 03siebrand * r39729 10/trunk/phase3/ (2 files in 2 dirs): Fixes for r39727 [19:19:42] brion: done [19:19:49] 03guyvdb * r39730 10/trunk/phase3/includes/HTMLDiff.php: Escape characters that were unescaped by the parser [19:20:17] Nikerabbit: grannyware == sloware [19:23:17] 03rotem * r39731 10/trunk/phase3/languages/messages/ (MessagesEn.php MessagesHe.php): Localization update. [19:27:05] 03aaron * r39732 10/trunk/phase3/includes/Skin.php: use quickUserCan() [19:36:40] 03dale * r39733 10/trunk/extensions/MetavidWiki/ (9 files in 6 dirs): updates for advanced search bill support [19:37:48] guyvdb: better thx :D [19:38:28] *AaronSchulz does a page view [19:38:30] wfMsgReal 7.43% 10.45% 15514 5171.33 0.24 0.76 1261.59 3930.76 [19:38:39] 15514 calls :) [19:41:56] guyvdb: and a way to switch between regular and html diffs would be great [19:42:31] 03ialex * r39734 10/trunk/extensions/CheckUser/CheckUser.php: Per report of Thogo on IRC, declare checkuser-log right in $wgAvailableRights [19:44:40] 03(mod) Postgres - Searches that include a number as a term always return no results - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=15229 (10reed) [19:49:21] 03(NEW) Make flagging from difflinks for unsighted pages less error prone - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=15249 15enhancement; normal; MediaWiki extensions: FlaggedRevs; (pbirken) [19:51:23] *qsheets asks brion to by him a PUMA [19:51:25] 03(NEW) Link tracking should use latest version, not the last reviewed - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=15250 15enhancement; normal; MediaWiki extensions: FlaggedRevs; (vasilvv) [19:54:06] brion: you want a flag to enable the html diff, and then still have the option to use the old diff? [19:56:22] maybe a way to switch it in/out for any request, so people can test side by side [19:56:32] 03btongminh * r39735 10/trunk/phase3/includes/api/ApiLogin.php: Provide some info on which case value was not handled in ApiLogin [20:10:07] 03(mod) Link tracking should use latest version, not the last reviewed - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=15250 (10JSchulz_4587) [20:12:36] Hi gang, what folder is the default logo in? [20:15:00] 03(mod) Make flagging from difflinks for unsighted pages less error prone - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=15249 (10JSchulz_4587) [20:17:21] 04(REOPENED) Change configuration for autopromotion on de.wikipedia.org - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=14975 +comment (10pbirken) [20:18:09] hi... I've created this template, which includes the same different template with a lot of different parameters.. this second template might be empty though depending on the parameters [20:18:17] 03(mod) Change configuration for autopromotion on de.wikipedia.org - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=14975 (10JSchulz_4587) [20:18:57] my problem is, that when the templates are empty, mediawiki interpret it, as though I want a
tag [20:19:19] is there any way to avoid that? [20:20:08] Avoid extra newlines when calling the template. [20:20:36] 03(mod) Change configuration for autopromotion on de.wikipedia.org - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=14975 (10pbirken) [20:21:09] alnokta, can I borrow your arabic-ness? [20:21:53] yeah, but it is much prettier/easier to edit when it on seperate lines [20:22:16] is there no way to turn off this behavior? [20:22:18] *alnokta hands over arabic to kibble [20:22:24] :-) [20:22:27] alnokta, see pm [20:25:56] 03(mod) Change configuration for autopromotion on de.wikipedia.org - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=14975 +comment (10raimond.spekking) [20:28:33] num num num korean bbq [20:28:54] RickZilla: skins/common/images [20:29:06] brion-lunch: Thanks [20:40:17] 03(mod) Skins "Simple" and "Modern" are broken - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=14954 (10wiki.bugzilla) [20:50:47] 03(NEW) A to Z display of inline queries - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=15251 15enhancement; normal; MediaWiki extensions: Semantic MediaWiki; (laurent) [20:53:08] i have a style issue about log files, in general -- not limited to MW [20:53:18] anybody go ahead and jump in [20:53:47] do you like to create a log file with a static filename, such as 'mythisorthatlog' [20:54:10] or a timestamped filename, such as 'log20080820' [20:54:11] ? [20:54:50] 03(mod) Switch Wikipedia search button to go straight to Google site search - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=15238 (10brion) [20:55:14] obviously the former will overwrite while the latter will not [20:55:19] i see advantages to each [20:55:42] does anybody, hopefully with more experience than i have, care to indicate a preference? [20:55:55] 03(mod) Last modified display date not the last one - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=6263 (10wiki.bugzilla) [21:01:52] Xiong: well, the unix pholosphy says: do separate things with separate tools. this is why unix programs generally write everything to a statically named log file, and there's a separate tool, logrotate, for renaming and comrpessing logs every so often [21:02:10] i think that's the best way [21:02:15] the philosophy is what i'm after [21:02:45] i've seen -- i don't know how many -- applications create log files; some always use the same filename, others datestamp it [21:03:12] i havn't sen many that use timestamps on their own. [21:03:54] well, i don't like the idea of timestamped logfiles during development, because i'm likely to do dozens of test runs, one after the other, and then have to throw the things out [21:04:07] but i'm trying hard to create re-entrant code [21:05:02] i think i have to rely, at a higher level of abstraction, that the caller will have the sense not to supply the same log filename already in use, if he's "not the same guy" [21:05:21] uh? [21:05:47] usually, you write everything to a single log file, really [21:05:54] ah, here i am in the low-level logging package, creating a new log object with a supplied filename [21:06:05] some systems even use a single log for all prgorams (that's why the syslog service exists) [21:06:33] two different higher-level packages could be calling me; shouldn't write to the same log -- ah, don't think it makes sense in my project [21:07:04] this is not any kind of system log but rather a progress report on a given run [21:08:02] the utility logger thing doesn't know, really, its larger context [21:08:34] ha ha, enough to drive a guy crazy, trying to keep track of all the things he's not supposed to know [21:08:41] thanks for the tipe [21:08:47] is it possible to use # on entry's names? [21:08:51] for example [[#mediawiki|#mediawiki irc chan]] [21:09:01] ooh, bad [21:09:24] 03(mod) DynamicPageList - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=11112 (10mostafa) [21:11:00] http://en.wikipedia.org/wiki/Main_Page#irc [21:11:40] Shown: # is used to reference sections, just as usual in urls [21:11:59] Shown: you can do whatever after the | though [21:12:07] i know [21:12:22] there's no way to name an entry so [21:12:28] http://en.wikipedia.org/wiki/User:Xiong/Sandbox [21:12:47] actually the link is completely valid [21:12:54] but it's relative [21:13:53] In the process of moving a mediawiki instance from one server to another, it seems like the categories have lost track of what pages should be in them (as in, the [[Category:Foo]] link is present on the page, but the page is not listed under Category:Foo) [21:14:05] Is there any way I can globally refresh all of the categories? [21:27:25] 03ialex * r39736 10/trunk/phase3/ (includes/EditPage.php languages/messages/MessagesEn.php): (log message trimmed) [21:27:25] Tweak for r29770 (yes, some months after it): [21:27:25] If the wiki is read only mode, don't show a "database locked" error if [21:27:25] previewing or showing diff, but rather the "normal" action with a warning that [21:27:26] the database is currently locked. If the user is trying to save the page, he'll [21:27:28] see the preview page instead. Currently, if you are saving or previewing a page [21:27:30] while the database is locked, you'll see the "database locked" error with the [21:37:32] hi guys. on my wiki, i have my main page protected so only i can edit it. what would be the best way so users could submit a page to me so i can review it and add it to the front page? [21:38:24] make them add a template to the page which categorizes it; review the contents of the category periodically? [21:39:01] hmm im not sure what u mean exactly or how to do that [21:39:04] do u have an example? [21:46:16] !categories | b00st3d [21:46:16] --mwbot-- b00st3d: http://www.mediawiki.org/wiki/Help:Categories [21:46:21] !templates | b00st3d [21:46:21] --mwbot-- b00st3d: For more information about templates, see . The most common issues with templates copied from Wikipedia can be fixed by installing ParserFunctions and enabling HTML Tidy . [21:46:41] b00st3d: but the simplest way is to tell people to put any suggestions on the talk page of the main page [21:46:45] that's what talk pages are for. [21:50:09] thanks dues [21:50:48] 03vrandezo * r39737 10/trunk/extensions/SemanticResultFormats/ (7 files in 3 dirs): Initial commit of the Google Charts visualizer and the graphs extension [21:53:14] Hi, how can I modify colors of tabs on monobook? [21:53:43] and how to modify hyperlinks? I don't need them to be blue... [21:54:28] !css | thegr0m [21:54:28] --mwbot-- thegr0m: 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 [21:55:35] !skins [21:55:35] --mwbot-- Overview: . Skin usage: . Gallery of CSS styles: . Writing your own: [22:00:04] 03dale * r39738 10/trunk/extensions/MetavidWiki/ (5 files in 5 dirs): [22:00:04] media search clean up & [22:00:04] fixes to advanced search [22:22:43] 03demon * r39739 10/trunk/phase3/includes/AjaxResponse.php: Document a bit. [22:23:25] brion: I'm changing the history page so that there are 2 compare buttons, one for HTML and one for Source [22:24:06] hmmk [22:25:18] I can't have 2 submit buttons :( [22:37:19] <^demon|away> guyvdb: Btw, I'm tweaking your HtmlDiff to use messages instead of the hardcoded $bundle in TagToString. [22:38:16] oh thanks [22:38:23] that was on my todo [22:39:15] <^demon> Yeah, should be done shortly. Just formatting them into MessagesEn then changing your getString() method into wfMsg calls, etc. [22:39:34] much obliged [22:40:27] there already is a section for the html diff in MessagesEn [22:41:46] <^demon> Yeah, putting them with that.