- 10https://bugzilla.wikimedia.org/show_bug.cgi?id=14921 (10N/A) [11:31:47] <kassoe> mwbot - thanks! I think it has what I need, I'm a newbee, didn't find this following the links on the edit help pages, nor did I find it using the search functions ... I wonder why ... [11:37:44] <wikibugs> 03(NEW) problems with forced PNG rendering - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=15057 normal; normal; MediaWiki extensions: texvc; (debjanux) [11:42:05] <CIA-55> 03raymond * r38705 10/trunk/phase3/includes/Linker.php: Tweak doc of link() to reflect the order in function. [11:57:52] <guyvdb> is there a way to get an HTML DOM tree or a stream from the parser? [11:58:34] <CIA-55> 03river * r38706 10/trunk/switchboard/ (Makefile Makefile.config.example swexec.c): swexec: fix indentation; need Solaris project support for stable toolserver [12:05:56] <Duesentrieb> guyvdb: it's not a parser. it's a munger. it applies convoluted regular expressions until the wiki text has been muddeled into something resembling html. [12:13:05] <RoanKattouw> !class Parser | guyvdb [12:13:05] --mwbot-- guyvdb: See http://svn.wikimedia.org/doc/classParser.html [12:14:33] <guyvdb> I'll have to parse the HTML string manually then :( [12:19:46] <RoanKattouw> guyvdb: Can't you use DomDocument or something to parse the HTML into a DOM tree? [12:20:27] <guyvdb> what's ???DomDocument? [12:21:26] <guyvdb> I see [12:22:21] *Aaron|away wonders if Nikerabbit is around [12:22:23] <guyvdb> Actually I want to define my own DOM with some extra metadata [12:23:07] <RoanKattouw> guyvdb: DOMDocument used to be a PHP extension and is built in since PHP5. It's documented on php.net [12:23:50] <RoanKattouw> AaronSchulz: I've reassigned bug 14345 to you [12:24:17] <AaronSchulz> heh, don't expect it to get done anytime ;) [12:25:50] <guyvdb> RoanKattouw: I prever a streaming approach because I will be building my own DOM tree, and I'd have the HTML string in memory, plus the DOMDocument, plus my own DOM [12:26:17] <guyvdb> is xml_parser_create() included in the default php? [12:26:50] <RoanKattouw> AaronSchulz: I know you're busy, just wanted to inform your TODO list has grown even further ;) [12:26:59] *Werdna physically abuses C++ memory leaks [12:27:29] *RoanKattouw finds a garbage man to do some garbage collection for Werdna [12:30:16] <Oberoten> Hola. [12:30:30] <Oberoten> Is there any way to extract JUST the NAME of a Subpage instead of the entire path? [12:31:01] <Oberoten> IE : I have A/7 B/7 C/7 and A/2 B/3 [12:31:27] <Oberoten> ... and if I am on A/2 or B/7 I want to be able to pull the 2 or the 7 respectively? [12:33:06] <Oberoten> Or would I be better of going with a namespace? [12:34:31] <RoanKattouw> I think it's possbile [12:34:35] <RoanKattouw> !magicwords [12:34:35] --mwbot-- For more information about creating magic words and their inner workings, see <http://www.mediawiki.org/wiki/Manual:Magic_words>. For a list of magic words, please see <http://meta.wikimedia.org/wiki/Help:Magic_words>. [12:35:11] <RoanKattouw> Oberoten: {{SUBPAGENAME}} [12:38:03] <Oberoten> Thanks. :) That works a charm. [12:40:51] <^demon> RoanKattouw: I did some tests last night. RefreshLinks runs redirects through Article::updateRedirectOn(), so no new maintenance script would be needed, just a running of refreshLinks. [12:41:01] <RoanKattouw> ^demon: Good [12:41:07] <RoanKattouw> I'm just adding a new patch to that bug [12:41:20] <RoanKattouw> Wait a few minutes and you'll see what it's about [12:41:28] <^demon> Okies. [12:42:22] <Oberoten> Eeeeh.... and if it under several pages? Seems Subpagename only returns the first one? [12:47:11] <^demon> RoanKattouw: I've got to be in a meeting in a few minutes. I'll check back when I get out of it. Also, new bug involving redirects: 15054. Might be worth looking into during our overhual-the-redirect-table-and-make-redirects-better work :-) [12:47:21] <RoanKattouw> !b 15054 [12:47:21] --mwbot-- https://bugzilla.wikimedia.org/show_bug.cgi?id=15054 [12:47:40] <RoanKattouw> Oberoten: I think you need to use {{#titleparts:}} then [12:48:00] <RoanKattouw> ^demon: Wrong bug number? [12:48:21] <^demon> Bah, -1. [12:48:24] <^demon> 15053 [12:48:27] <Oberoten> ... ookay? [12:48:32] <^demon> "Redirect parsing is over-permissive; syntactically broken redirects are parsed as redirects to the wrong place. " [12:48:40] <Oberoten> It'd be easier to make a new namespace instead? [12:49:31] <CIA-55> 03raymond * r38707 10/trunk/phase3/ (5 files in 3 dirs): [12:49:31] <CIA-55> * Consistency tweaks with other special pages like move/protect/delete: [12:49:31] <CIA-55> ** Add a backlink subtitle [12:49:31] <CIA-55> ** Remove 1 hardcoded '<' backlink and 1 '>' backlink [12:50:57] <wikibugs> 03(mod) Add rd_interwiki colum to redirect table - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=14418 +comment (10roan.kattouw) [12:59:16] <tusho> In a stylesheet how can I select html & body for printable? [12:59:28] <CIA-55> 03raymond * r38708 10/trunk/phase3/includes/specials/SpecialWhatlinkshere.php: [12:59:28] <CIA-55> Switch errormessage and filter to show 1) filters and 2) errormessage. [12:59:28] <CIA-55> Otherwise the filter hops around. [12:59:50] <wikibugs> 03(NEW) Entering Croatian characters produces errors - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=15058 major; normal; MediaWiki: Page editing; (radoslav.dejanovic) [13:22:45] <Ellywa> hi all, has anybody opportunity to give advice on a "special character" like ?? ? [13:23:32] <Ellywa> In a mediawiki installation I have severe problems, see this edit: [13:23:34] <Ellywa> http://www.gpps.nl/index.php?title=Inhoudsopgave&diff=118&oldid=50 [13:23:44] <Ellywa> this is caused by the ?? character. [13:24:06] <Ellywa> What is wrong in the installation? Anybody any idea? [13:24:10] <ilaiho> Ellywa: Is the UTF-8 or Unicode character set enabled? [13:24:32] <Ellywa> where should that be enabled? [13:25:23] <superdeluxe> Hi there; I've a confusing problem with mediawiki - it seems to have trouble saving some hex values I'm entering for a page on our internal wiki. [13:25:37] <ilaiho> Ellywa: Preferrably when you installed it [13:26:09] <ilaiho> if you want to migrate to UTF-8, you should ask someone wiser [13:26:29] <ilaiho> but I think it's possible [13:26:30] <Ellywa> my provider did the installation. I have access to the LocalSettingsPHP [13:26:38] <Ellywa> and everything else [13:32:11] <MaxSem|away> lol http://www.gpps.nl/index.php/Speciaal:Softwareversie [13:34:17] <Ellywa> yes indeed. :-(. Probably due toe ??var Arnfj??r?? Bjarmas etc.... [13:35:17] <Ellywa> in another installation of the same provider this is perfectly in order [13:35:19] <Ellywa> http://www.luchtnieuws.nl/index.php/Speciaal:Versie [13:39:11] <Ellywa> I found this in the LocalSettings: [13:39:14] <Ellywa> # Experimental charset support for MySQL 4.1/5.0. [13:39:17] <Ellywa> $wgDBmysql5 = false; [13:39:22] <Ellywa> might that be the cause? [13:39:39] <CIA-55> 03aaron * r38709 10/trunk/extensions/FlaggedRevs/specialpages/Stabilization_body.php: use wfMsgForContent() [13:39:43] <Werdna> Ellywa: yes. [13:39:57] <Ellywa> should it be true? [13:40:25] <Werdna> Ellywa: hmm, you may need to recreate your DB or something [13:40:29] <Werdna> it's hardly experimental now. [13:40:45] <Ellywa> what will that mean, that all is lost? (I hope not...) [13:41:21] <Ellywa> What will happen when I change it? [13:41:24] <Werdna> you can try turning it on :) [13:41:26] <Werdna> see what happens [13:41:31] <Werdna> it won't cause irreversible damage. [13:41:35] <Werdna> or, indeed, any damage at all [13:41:38] <Ellywa> sure? [13:41:39] <Werdna> you might get errors, though [13:42:05] <Ellywa> I will look in the other installation first... [13:43:22] <wikibugs> 03(mod) Add rd_interwiki colum to redirect table - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=14418 (10innocentkiller) [13:47:53] <Ellywa> Werdna: This cannot be the cause. In the other installation, this command in Local Settings is exactly the same [13:51:30] <CIA-55> 03tstarling * r38710 10/trunk/cortado/src/com/fluendo/jheora/FrArray.java: Workaround for bug 14610 (reproducible crash on 64-bit architectures) as suggested on http://bugs.sun.com/view_bug.do?bug_id=6732194 [13:53:08] <Werdna> TimStarling: you're up late. May I disable self-links for specialpages? This is the commit message I was about to use: Ignore self-links for special pages. They might be there for a good reason, and the current parsing of special-page output means that a link from any subpage of a special page to the special page itself is considered to be a self-link. This is evil, and doesn't take into account parameters added to other parts of the URL (e.g. ... [13:53:14] <Werdna> ... the unblock form on Special:Ipblocklist) [13:53:56] <Ellywa> who knows anybody who can help me? [13:54:16] <TimStarling> yes [13:54:27] <Werdna> Ellywa: I'm not familiar with migrating databases from ISO-8591 to UTF-8. [13:54:48] <CIA-55> 03werdna * r38711 10/trunk/phase3/includes/parser/Parser.php: [13:54:48] <CIA-55> Ignore self-links for special pages. They might be there for a good reason, and [13:54:48] <CIA-55> the current parsing of special-page output means that a link from any subpage of [13:54:48] <CIA-55> a special page to the special page itself is considered to be a self-link. This [13:54:48] <CIA-55> is evil, and doesn't take into account parameters added to other parts of the [13:54:50] <Ellywa> oka [13:54:51] <CIA-55> URL (e.g. the unblock form on Special:Ipblocklist) [13:55:49] <CIA-55> 03werdna * r38712 10/trunk/phase3/RELEASE-NOTES: Should probably update RELEASE-NOTES [13:57:20] <Werdna> TimStarling: If you have a moment to spare (if not, don't stress), is using a stringbuf over and over to process lots of data likely to cause a memory leak? I've got a memory leak in my native abuse filter parser which valgrind traces to the use of a stringbuf to process input. [13:58:21] <Toaster> Hey, I am editing my LocalSettings.php file and am trying to add a favicon. I am using this: $wgFavicon = "/favicon.ico"; [13:58:21] <Toaster> I have made sure I uploaded the favicon into the root folder where MediaWiki is installed, but no favicon is showing up. Can anyone help me? [13:58:40] <Werdna> !wg Favicon [13:58:40] --mwbot-- http://www.mediawiki.org/wiki/Manual:%24wgFavicon [13:59:06] <Werdna> Toaster: have you tried force-refreshing your site? [13:59:14] <Werdna> sometimes, it'll take a while to show in the browser, AFAIK [13:59:22] <Ellywa> my suggestion, try another browser.... [13:59:37] <Ellywa> or another PC [13:59:50] <Werdna> does this line appear in your HTML sauce? [13:59:50] <Werdna> <link rel="shortcut icon" href="/favicon.ico" /> [14:00:31] <TimStarling> Werdna: you wouldn't think so, no [14:00:49] <TimStarling> valgrind isn't necessarily accurate when it labels bad lines [14:00:59] <Toaster> Ah, thank you guys. IE is showing it just fine. :) [14:01:17] <Toaster> Stupid Fx... Been having issues with 3.0 since it came out... [14:02:50] <Toaster> Oh, do you guys know of a "tutorial" about creating the main page? [14:02:52] <Werdna> TimStarling: hmm.. If valgrind seems inaccurate, how should I find/plug the leak? [14:03:56] <TimStarling> ultimately by inspection, but you can narrow it down by commenting out parts of your code [14:04:12] <Werdna> ouch. [14:04:13] <TimStarling> or by doing things multiple times and seeing if it leaks more [14:04:27] <Werdna> 1700 lines of code to check :) [14:04:48] <Werdna> although presumably, there are two places it could be leaking. [14:04:57] <TimStarling> valgrind tends to get things roughly correct, it won't be totally wrong [14:05:06] <Werdna> hmmm... under what circumstances is an object's destructor automatically called? [14:05:07] <TimStarling> it might just, say, label two calls to MD_Update instead of one [14:05:43] <TimStarling> you can add debug output statements to the destructors to trace calls [14:05:43] <Simetrical> Werdna, if it's statically allocated and falls out of scope, or if delete is called on it, I think . . . [14:05:58] <TimStarling> that's sometimes useful to understand what's going on [14:05:59] <CIA-55> 03aaron * r38713 10/trunk/extensions/FlaggedRevs/specialpages/Stabilization_body.php: [14:05:59] <CIA-55> * Add null edit for changes [14:05:59] <CIA-55> * Bug fix if $row not present [14:06:22] <wikibugs> 03(FIXED) Changing Page configuration should trigger an entry in page history - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=14993 +comment (10JSchulz_4587) [14:07:10] <Werdna> TimStarling: yes, I did that a week or so ago when I learned the rule of 3 the hard way :) [14:08:28] <wikibugs> 03(FIXED) enable .ogv uploads - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=14315 +comment (10jeluf) [14:08:58] <Werdna> valgrind is pointing to completely the opposite end of code to where I'd expect memory leaks. [14:09:17] <Werdna> the code it's pointing to is a simple 20-line function to load input from stdin [14:09:21] <Simetrical> What about .oga? [14:09:24] <Toaster> Where can I learn the proper markup to make my main page fancy? [14:09:31] <Simetrical> !markup [14:09:31] --mwbot-- I don't know anything about "markup". You might try: !defaultsettings !parser !syntax !viewvc [14:09:37] <Simetrical> !syntax | Toaster [14:09:37] --mwbot-- Toaster: For help with MediaWiki's Wikitext syntax, please see <http://meta.wikimedia.org/wiki/Help:Editing>. For an (incomplete) formal specification, see <http://www.mediawiki.org/wiki/Markup_spec>. [14:09:41] <Simetrical> !markup alias syntax [14:09:41] --mwbot-- Successfully added alias: markup [14:10:05] <Werdna> http://pastebin.ca/1093811 <-- is there something obviously wrong with this function? [14:10:43] <Werdna> valgrind says this: [14:10:43] <Werdna> ==14225== by 0x806E67D: loadRequest() (main.cpp:51) [14:10:56] <Werdna> line 51 is this one [14:10:57] <Werdna> cin.get( codesb, '\0' ); [14:14:18] <Werdna> ==14225== by 0x41C9372: std::basic_stringbuf<char, std::char_traits<char>, std::allocator<char> >::overflow(int) (in /usr/lib/libstdc++.so.6.0.9) [14:14:21] <Werdna> hmm.. [14:15:02] <tusho> uhh, what's that got to do with MW? :\ [14:17:14] <CIA-55> 03tstarling * r38714 10/trunk/extensions/OggHandler/ (OggHandler.i18n.php OggHandler_body.php OggPlayer.js): Add support for mplayerplug-in, yet another browser plugin with a fake QuickTime. [14:18:03] <Werdna> tusho: it's an extension being developed for MW [14:18:08] <tusho> ah [14:18:55] <CIA-55> 03tstarling * r38715 10/trunk/extensions/OggHandler/ (OggHandler.php cortado-ovt-stripped-wm_r38710.jar): Update Cortado to r38710 to fix bug 14610 (crash on 64-bit) [14:19:38] <wikibugs> 03(mod) {{int:X}} respects user-defined interface language, breaking link tables etc. (aka {{USERIFCODE}} strikes back) - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=14404 +comment (10cbrown1023) [14:19:55] <RoanKattouw> ^demon: bug 15053 is related to redirect *parsing*, something I haven't touched. I don't intend to either, I guess Tim had best fix that [14:20:13] <TimStarling> Werdna: what is vars? [14:20:20] <Duesentrieb> redirect parsing? isn't that a single regex? [14:20:39] <Werdna> TimStarling: variables passed to the filter. [14:20:47] <Werdna> i.e. USER_NAME, MOVED_TO_TEXT, and so on. [14:21:01] <TimStarling> what type is it exactly? [14:21:09] <Werdna> those two are defined at global scope [14:21:15] <Werdna> string filter; [14:21:16] <Werdna> map<string,AFPData> vars; [14:21:48] <TimStarling> what's AFPData? is that your own class? [14:21:52] <Werdna> yes. [14:22:10] <RoanKattouw> Duesentrieb: Yes. But I'm not our in-house regex expert, and I think we can safely say Tim is ;) [14:22:23] <TimStarling> vars[keysb.str()] = AFPData( valsb.str() ); [14:22:29] <TimStarling> does it leak memory without that line? [14:23:02] <Werdna> yes [14:23:06] <Simetrical> RoanKattouw, there's no such thing as a regex expert. We aren't using, like, PCRE's recursion functions here or something. [14:23:17] <Simetrical> It's a one-line regex, IIRC. [14:23:28] <RoanKattouw> Simetrical: The redirect regex is yeah [14:24:04] <RoanKattouw> But since Tim wrote the parser which is really the ultimate regex nightmare, he probably knows a thing or two about regex pitfalls ;) [14:24:39] <Werdna> TimStarling: if helpful, full valgrind output is at http://pastebin.ca/1093820 [14:24:40] <Simetrical> . . . there's not much to know about regex. [14:25:01] <Simetrical> It's, like, regex. Finite-state machines. Easy stuff. No conditionals or recursion or anything. [14:25:03] <RoanKattouw> Simetrical: Of course anyone with sufficient regex skills could fix that bug. I just said I'm not that person [14:25:33] <Simetrical> if( preg_match( '!\[{2}(.*?)(?:\|.*?)?\]{2}!', $text, $m ) ) { [14:25:47] <RoanKattouw> Every time I use regexes (like once every three months) I end up looking over the entire regex tutorial *again* and again not memorizing it for the next time :P [14:25:52] <Simetrical> It should just avoid newlines. [14:26:15] <RoanKattouw> It does [14:26:16] <Simetrical> RoanKattouw, you really need to learn regexes properly to be a good programmer. I mean, how do you grep for stuff? [14:26:18] <TimStarling> it only says possibly lost [14:26:25] <Simetrical> Not according to the report . . . [14:26:25] <TimStarling> not definitely lost [14:26:37] <TimStarling> does it actually leak? [14:26:46] <RoanKattouw> Simetrical: I understand regexes well enough. I just don't use them often enough to remember everything [14:26:50] <Werdna> TimStarling: when I ran it with a few hundred thousand entries, mem usage slowly crept up. [14:26:51] <RoanKattouw> Anyway, the problem isn't newlines [14:27:13] <Werdna> so it does definitely leak. [14:27:17] <RoanKattouw> The problem is that if the first line doesn't match #REDIRECT [[Foo]], it should stop [14:27:26] <Werdna> I haven't run the same test with that line commented out, admittedly. [14:27:27] <Werdna> One moment. [14:27:41] <RoanKattouw> What it does is that #REDIRECT [[Foo] \n\n [[Category:Bar]] ends up being interpreted as a redirect to [[Category:Bar]] [14:27:44] <TimStarling> and it wasn't the size of that map increasing? [14:28:40] <Werdna> same vars are passed each time, but, for safety, I've since added vars.clear() beforehand. [14:28:55] <Werdna> There's some other in-process caching, but that has a limit of 100 entries each. [14:29:07] <TimStarling> what's line 55? [14:29:08] <Werdna> I commented out that other in-process caching, and it still leaked [14:29:26] <Werdna> cin.get( codesb, '\0' ); [14:29:45] <Werdna> commenting out the map set doesn't fix it. [14:32:14] <dabblz> Anyone know how I can make it so that searches look at the content of Category: pages as well? ie. Category:Foo ... has some content on it as well as the default list of articles [14:32:20] <TimStarling> what's the other one? [14:32:34] <RoanKattouw> dabblz: Go to Special:Search, check the "Category" box and start searching [14:32:35] <TimStarling> oh, static ofstream log [14:32:52] <TimStarling> what about the assignment to filter? [14:33:09] <TimStarling> filter = codesb.str(); [14:33:28] <Werdna> TimStarling: ah, perhaps this is a red herring. I figured I'd take that file out of the equation and try something like this: [14:33:30] <Simetrical> dabblz, change $wgContentNamespaces. [14:33:41] <Werdna> for(int i=0;i<=500000;i++) { result = e.evaluateFilter( "specialratio('foo;') == 0.25" ); } [14:33:51] <Werdna> TimStarling: and it's now at 20% memory usage and growing [14:33:56] <Werdna> so valgrind must have been lying to me. [14:34:20] <TimStarling> right [14:34:22] <Werdna> presumably, I'm not freeing the code once I'm done with it. [14:34:42] <Werdna> and that's why the error was detected back at the line where the code is first put into memory. [14:37:34] <wikibugs> 03(mod) Redirect parsing is over-permissive; syntactically broken redirects are parsed as redirects to the wrong place. - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=15053 (10Simetrical+wikibugs) [14:37:49] <CIA-55> 03raymond * r38716 10/trunk/extensions/Translate/groups/mediawiki-defines.txt: Update per r38714 [14:38:08] <Werdna> TimStarling: fortunately, the extra places where the code is kept per-evaluation is only 15 lines or so. [14:40:11] <Nikerabbit> who called [14:40:12] <dabblz> Anyone know how I can make it so that searches look at the content of Category: pages as well? ie. Category:Foo ... has some content on it as well as the default list of articles ... I added NS_CATEGORY => true to $wgNamespacesToBeSearchedDefault in includes/DefaultSettings.php ... and still doesn't set it by default on searchh [14:40:21] <Nikerabbit> AaronSchulz: wassup? [14:40:52] <AaronSchulz> nikerabbit: any way you could convert flaggedrevs to the new alias system? [14:41:15] <Nikerabbit> AaronSchulz: I could look at it tonight if I have some spare time [14:41:25] <AaronSchulz> thanks [14:41:46] <Simetrical> dabblz, it won't retroactively change logged-in users' preferences. You have to change your default search namespaces too. It will work for logged-out or new users. [14:42:16] <wikibugs> 03(mod) Redirect parsing is over-permissive; syntactically broken redirects are parsed as redirects to the wrong place. - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=15053 +comment (10roan.kattouw) [14:42:18] <dabblz> ahh okay thanks [14:42:25] <dabblz> need to edit User.php then [14:42:44] <RoanKattouw> dabblz: No need [14:43:00] <Nikerabbit> but now --> have to make some food [14:43:02] <RoanKattouw> Users' prefs are stored in the user table, you can change existing users' preferences there [14:43:09] <RoanKattouw> But make a backup first, of course [14:43:42] <Werdna> TimStarling: found one of them! [14:43:51] <Werdna> AFPData r = new AFPData(); [14:43:53] <Werdna> bad Andrew. [14:44:07] <dabblz> okay ty [14:44:15] <flyingparchment> that shouldn't even compile unless you're doing something odd [14:44:20] <RoanKattouw> Werdna: Isn't that illegal? [14:44:27] <TimStarling> #define AFPData void* [14:44:38] <TimStarling> that probably wouldn't work either though [14:44:43] <RoanKattouw> void *r = new void * (); [14:44:44] <RoanKattouw> Right... [14:44:53] <TimStarling> typedef maybe [14:45:01] <Werdna> flyingparchment: that stuff does compile, it's been driving me mad. [14:45:17] <RoanKattouw> Werdna: What is AFPData? A class? In that case it shouldn't compile [14:45:19] <flyingparchment> Werdna: how? do you have a ctor that takes an AFPData*? (why?) [14:45:36] <Werdna> flyingparchment: I have a copy constructor.. [14:45:45] <flyingparchment> that takes a refernce, not a pointer [14:45:56] <Werdna> AFPData::AFPData( const AFPData & oldData ) [14:45:58] <Werdna> mmm [14:46:29] <CIA-55> 03rotem * r38717 10/trunk/phase3/languages/messages/MessagesHe.php: Localization update for he. [14:46:48] <Werdna> anyway, now I have leaks in my assignment operator. [14:49:14] <Werdna> http://pastebin.ca/1093841 if anybody wants to take a look [14:50:02] <wikibugs> 03(FIXED) Cortado is broken on Java 6 with the server VM - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=14610 +comment (10tstarling) [14:51:39] <Werdna> anyway, I'm going to bed [14:51:47] <Werdna> it's like 1am and I hve to be up at 7 [14:52:06] <Simetrical> Bug ID 6732194? Sun has 6.7 million bugs? [14:52:56] <wikibugs> 03(mod) Redirect parsing is over-permissive; syntactically broken redirects are parsed as redirects to the wrong place. - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=15053 (10roan.kattouw) [14:53:20] <TimStarling> that's the implication [14:53:29] <RoanKattouw> Simetrical: Added patch to bug 15053 [14:54:01] <wikibugs> 03(mod) Redirect parsing is over-permissive; syntactically broken redirects are parsed as redirects to the wrong place. - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=15053 +patch (10roan.kattouw) [14:55:11] <RoanKattouw> Werdna: Isn't that function supposed to return a reference? Then why return *this; ? [14:56:31] <RoanKattouw> OK... flyingparchment, do you know why he's doing return *this; all over the place when the function says it returns a reference? [14:56:40] <flyingparchment> where? [14:56:49] <RoanKattouw> http://pastebin.ca/1093841 [14:57:07] <flyingparchment> what's wrong with that? [14:57:41] <RoanKattouw> Well according to the prototype, the function returns AFPData & (i.e. a reference), but he's returning a pointer to a reference if I read it correctly [14:57:50] <RoanKattouw> Or maybe even a reference to a pointer to a reference :O [14:57:54] <flyingparchment> no, 'this' is already a pointer [14:58:01] <RoanKattouw> Oh no wait [14:58:02] <flyingparchment> * dereferences it and turns it into an object (AFPData) [14:58:06] <RoanKattouw> * is the DEreference operator [14:58:11] <RoanKattouw> C is too long ago for me apparently [14:58:24] <RoanKattouw> Yeah, that makes sense [14:58:25] <TimStarling> that and the keyboard doesn't have enough punctuation characters [14:58:49] <RoanKattouw> Werdna: Never mind, I was being stupid. I've been away from C so long I don't even remember pointers properly any more [14:58:55] <flyingparchment> also, wow, that code is ugly [14:59:14] <flyingparchment> what's with all these pointers and casting and manual type management? no wonder it's full of leaks [14:59:56] <wikibugs> 03(mod) Skins should add their name as a class in <body> - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=15052 (10Simetrical+wikibugs) [15:00:28] <wikibugs> 03(mod) Skins should add their name as a class in <body> - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=15052 (10Simetrical+wikibugs) [15:00:31] <guyvdb> Can I search for an object in an array with array_search()? It's not clear to me if the comparis will be by fields or by instance [15:01:10] <Staeiou> Greetings, everyone. I unsuccessfully imported a lot of incomplete pages into my 1.13rc1 LAMP installation and then deleted them in the database. Now my page_ids start at about 1.2 million, even though I only have ~300 real pages. Is there any way to reset this? [15:02:17] <RoanKattouw> Staeiou: No. [15:02:20] <Simetrical> guyvdb, PHP array functions are mostly insane. A loop is probably more reliable. [15:02:23] <TimStarling> guyvdb: by instance, I think [15:02:44] <Simetrical> Staeiou, yes. Set AUTOINCREMENT in MySQL. Make sure you change your existing page_ids first. [15:02:49] <flyingparchment> Staeiou: why would you care what page_id is? [15:02:58] <Staeiou> refreshLinks.php takes forever [15:03:01] <guyvdb> thanks [15:03:05] <Simetrical> Staeiou, you never want anything in the table with page_id above AUTOINCREMENT. . . . but yeah, why do you care? [15:03:06] <RoanKattouw> Simetrical: I added a patch to that redirect bug that removes the #REDIRECT part and adds ^\s* to the beginning of the regex [15:03:09] <Simetrical> It *shouldn't* take longer. [15:03:14] <TimStarling> oh [15:03:19] <TimStarling> http://au.php.net/manual/en/language.oop5.object-comparison.php [15:03:47] <TimStarling> "Two object instances are equal if they have the same attributes and values, and are instances of the same class. " [15:03:49] <Staeiou> It takes a normal amount of time from 1-300, then flies through 300-1.2 million at about 1000 a second [15:03:52] <Simetrical> RoanKattouw, looks good, if we want that. [15:03:55] <TimStarling> "On the other hand, when using the identity operator (===), object variables are identical if and only if they refer to the same instance of the same class. " [15:04:05] <guyvdb> yes I read that [15:04:16] <TimStarling> so presumably with array_search(), if strict=false, it'll do it like ==, and if strict=true, it'll do it like === [15:04:18] <guyvdb> but it isn't specified which comparison operator array_search uses :( [15:04:21] <RoanKattouw> guyvdb: So I guess it depends on whether you set the strict parameter on array_search() [15:04:29] <Simetrical> Staeiou, hmm. That seems strange. It should just skip over the blank space. [15:04:43] <TimStarling> I'll check the code to confirm [15:04:48] <Simetrical> Oh, no, it doesn't. [15:04:50] <guyvdb> I'm not so sure, they might check for type seperately and then it's still not comparing instances [15:04:52] <Simetrical> That's stupidly written. [15:05:01] <RoanKattouw> I expected it to be [15:05:24] <RoanKattouw> It should really SELECT stuff FROM page WHERE page_id > 123 ORDER BY page_id LIMIT 1; [15:05:32] <Simetrical> LIMIT 1000 or something, not 1. [15:05:38] <RoanKattouw> Well ok [15:05:58] <RoanKattouw> Anyway, that should skin nonexistent page_ids [15:06:18] <Simetrical> The current code should skip non-existent id's. [15:06:20] <TimStarling> int (*is_equal_func)(zval *, zval *, zval * TSRMLS_DC) = is_equal_function; [15:06:26] <TimStarling> if (Z_LVAL_PP(strict)) { [15:06:26] <TimStarling> is_equal_func = is_identical_function; [15:06:28] <Staeiou> Roan: I'll change the SQL statement in the code and see what happens [15:06:28] <Simetrical> Staeiou, what exact command are you running? [15:06:37] <Simetrical> On the command line? [15:06:53] <Staeiou> Simetrical: php rebuildall.php [15:07:04] <Simetrical> Ah. [15:07:09] <Simetrical> So it's rebuildall.php that's stupid, then. [15:07:21] <Simetrical> refreshLinks.php is perfectly intelligent. [15:07:24] <Staeiou> Simetrical: Same thing with refreshLinks.php too [15:07:30] <Staeiou> Just tried it out [15:08:00] <Simetrical> Oh, geez, you're right. [15:08:02] <Simetrical> Ugh, stupid thing. [15:08:09] <RoanKattouw> Simetrical: rebuildall.php just includes refreshLinks.inc and some other stuff [15:08:14] <Simetrical> No, so it is refreshLinks.inc. [15:08:23] <Simetrical> RoanKattouw, I know, but it depends how it calls it. [15:09:01] <TimStarling> yes, it's all good [15:09:54] <TimStarling> with the interesting exception that the behaviour will change if ze1_compatibility_mode is set [15:10:05] <TimStarling> anyway, same as == and === [15:10:34] <RoanKattouw> Simetrical: Wow, rebuildLinks is seriously brain-damaged [15:10:46] <RoanKattouw> It gets MAX(page_id) and then runs a for loop from 1 to that value [15:11:20] <TimStarling> who are you calling brain-damaged? [15:12:04] <RoanKattouw> rebuildLinks.inc [15:12:16] <TimStarling> there's no such file [15:12:29] <RoanKattouw> refreshLinks.inc then [15:12:44] <RoanKattouw> It works, but in a very ugly way [15:12:52] <TimStarling> what's ugly about it? [15:13:14] <RoanKattouw> Well look what happened to the guy who has page IDs 30 - 12000 missing [15:13:26] <Staeiou> Roan: 300 to 1.2 million [15:13:36] <RoanKattouw> OK, add two zeros [15:13:40] <Staeiou> hehe [15:13:41] <TimStarling> better than trying to load a million page rows, running out of memory and crashing [15:14:02] <Simetrical> TimStarling, you could load 10,000 at a time. [15:14:17] <TimStarling> well, that guy with the broken database can do that [15:14:19] <RoanKattouw> Ideally, the main loop would fetch the next row (or the next 10K rows) from the DB and call its helpers with a row as parameter [15:14:22] <RoanKattouw> rather tahn a page ID [15:14:44] <Simetrical> It's not broken to have large gaps in your primary key. [15:14:59] <TimStarling> if you do that, you either get an old page row, or you have to lock that section of the database [15:15:03] <RoanKattouw> Which is better than running about a million SELECT stuff FROM page WHERE page_id=400 queries that don't return anything [15:15:38] <wikibugs> 03(mod) Cortado is broken on Java 6 with the server VM - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=14610 (10N/A) [15:16:52] <^demon> RoanKattouw: commented on your new patch, looks good :-) [15:16:57] <wikibugs> 03(mod) Exclude bot-generated spam reports on meta from indexing via robots.txt - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=14076 summary; +comment (10mike.lifeguard) [15:17:12] <RoanKattouw> ^demon: Yeah I saw the comment. Now I'm waiting for Brion to get here so I can poke him about ti [15:17:45] <RoanKattouw> TimStarling: Anyway, what's wrong with fetching a page row from the DB and passing that to a helper function instead of passing a page ID so the helper can fetch the page row himself? [15:17:45] <Simetrical> TimStarling, even if you only select one at a time, then? [15:18:17] <TimStarling> it sounds like you want to make it ugly for everyone so that it can be fast for that one guy [15:18:38] <wikibugs> 03(mod) Exclude bot-generated spam reports on meta from indexing via robots.txt - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=14076 (10mike.lifeguard) [15:18:48] <RoanKattouw> Not really. I want to throw out futile queries trying to get page rows that don't exist [15:19:26] <RoanKattouw> In this guy's situation, it happens to do a million of those [15:20:11] <RoanKattouw> But that's not really relevant. What is, is that the number of futile queries roughly equals the number of deleted pages, which is quite big on any sufficiently large and vandal-pestered wiki [15:21:50] <Simetrical> TimStarling, it's not really ugly. [15:21:59] <Simetrical> Marginally longer code, not by much. [15:22:51] <RoanKattouw> It's what we use in other scripts too, like the script that updates the category table [15:22:59] <RoanKattouw> I mean populates [15:23:19] <flyingparchment> Werdna: is the code you're having issues with in svn? [15:23:34] <RoanKattouw> flyingparchment: Werdna's asleep [15:27:24] <TimStarling> true, some scripts do batch queries [15:28:15] <TimStarling> I chose to do it by each page_id when I wrote refreshLinks because the processing time for each page was so long, and did so many queries, that I didn't think it would matter [15:29:09] <Simetrical> Well, you can do it per-row but make the next row number equal to one greater than the last Title's row number. [15:29:15] <Simetrical> Rather than just incrementing it by one. [15:31:44] <TimStarling> you mean do another query just to find the page_id? [15:31:57] <TimStarling> that sounds a bit sucky [15:32:09] <TimStarling> I think it's probably best to do it in *small* batches, of say 100 [15:32:33] <RoanKattouw> SELECT page_id FROM page WHERE page_id > 123 ORDER BY page LIMIT 100; [15:32:39] <TimStarling> yes [15:32:49] <RoanKattouw> Then work with those page ids [15:32:58] <TimStarling> then you get the right page_ids, with minimal overhead, the information is reasonably up-to-date, and you don't lock anything [15:33:03] <Simetrical> TimStarling, no, I mean do a query, then use Title::newFromRow() instead of newFromId(). [15:33:15] <RoanKattouw> Simetrical: Was just gonna say that [15:33:44] <TimStarling> I'm not sure if Title::newFromRow() would gain you a query... [15:33:46] <RoanKattouw> BTW, why are $wgTitle and $wgArticle set all the time in that code? Is that required or is it just a relic from the time we actually had to do that? [15:33:53] <TimStarling> but yes, you could do it that way [15:33:55] <Simetrical> I'm not sure either, actually. [15:34:01] <TimStarling> RoanKattouw: they aren't [15:34:03] <RoanKattouw> You'd have to select more that page_id in that case [15:34:12] <TimStarling> they're not set by Setup.php, they're set by the entry point [15:34:21] <RoanKattouw> $wgTitle = Title::newFromID( $id ); [15:34:22] <TimStarling> and not at all by commandLine.inc [15:34:38] <TimStarling> oh, you mean in refreshLinks.inc [15:34:41] <RoanKattouw> Yes [15:34:43] <TimStarling> because things break without that [15:34:44] <Simetrical> There's no speed difference with batches, though, you were right. [15:34:47] <Simetrical> The queries aren't the slow part. [15:35:06] <TimStarling> like {{int:}} [15:35:18] <RoanKattouw> TimStarling: Oh right, $wgParser->parse() [15:36:15] <RoanKattouw> Somewhere, deep down, the parser requires that $wgTitle be set for certain things [15:37:05] <TimStarling> like {{int:}} [15:37:26] <RoanKattouw> Simetrical: I though that too. Eliminating a few blazing-fast indexed queries doesn't help when your parsing lots of stuff, updating/deleting hundreds of rows, etc. [15:37:28] <TimStarling> because {{int:}} calls wfMsg(), which calls transformMsg(), which sets $parser->mTitle to $wgTitle [15:37:45] <TimStarling> I used to be interested in that kind of thing [15:37:51] <RoanKattouw> TimStarling: Why does the parser need a title to parse transformMsg() [15:37:58] <TimStarling> I've found other things to get excited about since then [15:38:02] <TimStarling> {{PAGETITLE}} [15:38:22] <TimStarling> and to determine whether a link is a self-link [15:38:34] <RoanKattouw> TimStarling: You might wanna get back to being interested in $wgTitle in the parser. It requires that e.g. ApiEdit has to set $wgTitle, which is kind of stupid [15:39:11] <TimStarling> seriously, there's a whole world out there of interesting mediawiki projects [15:39:20] <RoanKattouw> Indeed [15:39:25] <TimStarling> hundreds of features to write, not enough time to write them all [15:39:28] <Simetrical> RoanKattouw, how can you parse things like {{PAGENAME}} without $wgTitle set? [15:39:40] <RoanKattouw> Simetrical: By setting $parser->mTitle [15:39:58] <Simetrical> I see . . . [15:40:23] <RoanKattouw> Same for [[Foo]] though, the way that's parsed depends on whether $parser->mTitle is Foo [15:40:44] <TimStarling> the job queue used to use a FakeTitle for $wgTitle, which throws an exception when $wgTitle is used [15:41:07] <TimStarling> but I didn't bother fixing it and it was just throwing exceptions every time {{int:}} was used, for months [15:41:33] <TimStarling> I guess I stopped caring about $wgTitle around then [15:42:27] <RoanKattouw> TimStarling: Well like I said, the parser really should just take $parser->mTitle and use that all the time, so you can e.g. parse pages that aren't $wgTitle or even parse pages when $wgTitle is unset [15:42:59] <TimStarling> pass a title to wfMsg()? [15:43:12] <RoanKattouw> Why does wfMsg() need a title? [15:43:25] <RoanKattouw> I thought wfMsg() didn't parse wikitext [15:43:27] <TimStarling> I said why, because it calls transformMsg() [15:43:59] <Staeiou> Simetrical, Tim, and Roan: thanks for helping me out, but it looks like Semantic Mediawiki's maint functions do the same thing. *sigh* I guess the easiest thing to do is export via XML, reinstall, and import. [15:45:48] <RoanKattouw> Oh right but then wfMsgWikiHtml() and possibly others do allow wikitext in messages, and that needs to be parsed, and that needs a title, and that needs $wgTitle [15:45:51] <RoanKattouw> *sigh* [15:45:59] <Staeiou> Anyways, I've got to head to a meeting. Thanks for trying to help me out. [15:46:03] <RoanKattouw> I think I'm finally starting to understand the globals-are-evil people [15:46:31] <TimStarling> it's not wikitext exactly, just double-brace constructs like {{PLURAL}} [15:46:50] <Barnicle> is version 1.12 the latest version of mediawiki? [15:47:00] <RoanKattouw> Barnicle: Yes. Read the topic [15:47:14] <TimStarling> messages are preprocessed with the mediawiki preprocessor, then they can be written to HTML, escaped or not [15:47:46] <RoanKattouw> Well eventually something somewhere is gonna wfMsgWikiHtml() a message, which requires full parsing [15:48:17] <RoanKattouw> Oh yeah that's what you said ^^ never mind [15:49:35] <gegema> I am trying to copy contents of a template from WP [http://en.wikipedia.org/wiki/Template:!], how do I get the source of the template? [15:49:53] <RoanKattouw> gegema: By going to the template and editing it or clicking view source [15:50:33] <RoanKattouw> gegema: Or you can know that Template:! just contains | [15:51:29] <gegema> RoanKattouw: Alright... thanks >> all I saw was [|<noinclude>{{template doc}}</noinclude>] and didn't really get all I needed was | :) [15:51:48] <RoanKattouw> gegema: Well you'll lose the fancy documentation and all that [15:52:58] <RoanKattouw> TimStarling: Where do I find docs telling me what {{int:}} even does? [15:53:18] <gegema> RoanKattouw: I am trying to make an infobox template... following steps given on Mwusers forum... and I was stuck on this stage...The absence of documentation on infoboxes on the MW manual or the meta is a mystery to me though [15:54:45] <CIA-55> 03rotem * r38718 10/trunk/phase3/languages/messages/MessagesHe.php: essage=Localization fixes. [15:54:58] <RoanKattouw> TimStarling: Never mind, I got it [15:57:50] <wikibugs> 03(mod) $wgServer lacks brackets in IPv6 URLs - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=14977 (10roan.kattouw) [15:57:58] <Lady_Aleena> When using {{SUBJECTSPACE}} in an if statement and the subject space is main space, how do I denote that? {{#ifeq:{{SUBJECTSPACE}}|articlespace|good|bad}} [15:58:20] <RoanKattouw> Lady_Aleena: IIRC, {{SUBJECTSPACE}} is empty for the main namespace [15:58:34] <Lady_Aleena> RoanKattouw, thanks... [15:58:54] <RoanKattouw> So {{#ifeq:{{SUBJECTSPACE}}||good|bad}} should work, as should {{#if:{{SUBJECTSPACE}}|bad|good}} [15:59:07] <gegema> RoanKattouw: is it a requirement to enable tidy to be able to use Infobox templates? [15:59:10] <wikibugs> 03(mod) No line break after 'subpages' class in Modern skin - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=13185 (10innocentkiller) [15:59:32] <RoanKattouw> gegema: Not necessarily. But if you're stealing infobox templates from WP, be aware that some of them rely on Tidy [15:59:34] <RoanKattouw> !tidy [15:59:34] --mwbot-- Wikimedia uses HTML Tidy, which allows mixed html and wikitables inside parser functions and templates. This can often be confusing for people copying templates from Wikimedia projects. For more information about HTML Tidy, see <http://www.mediawiki.org/wiki/Manual:%24wgUseTidy>. [15:59:37] <CIA-55> 03raymond * r38719 10/trunk/phase3/ (3 files in 3 dirs): [15:59:37] <CIA-55> Reapply part of r38229 under consideration of Brions revert comment (r38249): [15:59:37] <CIA-55> * Add a nice fieldset for consistency with a better legend [15:59:37] <CIA-55> * Switch result line and input form in kind that the hotspot for "Show new images since..." link does not change when following new uploads [15:59:39] *Lady_Aleena plops on the sofa, glowering at this switch that isn't working in main space. [16:00:28] <^demon> RoanKattouw: Any reason you changed 14977 to NEW instead of REOPENED? [16:00:34] <RoanKattouw> Where's Brion? Surely it's like 10 PM in the US now [16:00:49] <RoanKattouw> ^demon: It was assigned to me, I unassigned it [16:00:58] <Simetrical> RoanKattouw, it's 12 PM on the East Coast, 9 AM on the West Coast, sure not 10 PM anywhere. :) [16:01:13] <TimStarling> that's a bit out-of-date, the effects of enabling tidy aren't so severe in 1.12+ [16:01:22] <RoanKattouw> Simetrical: It's 10 PM in Mountain time then [16:01:31] <gegema> RoanKattouw: Now "stealing" is a word that makes me feel guilty, am merely using it as a template to learn the basics of creating my own infobox [16:01:33] <gegema> :) [16:01:34] <RoanKattouw> Anyway, Brion lives on the West coast right? [16:01:42] <RoanKattouw> gegema: No insult intended ;) [16:01:51] <TimStarling> it's 10am, not 10pm [16:01:55] <Lady_Aleena> It is 9 am on the west coast. [16:01:57] <RoanKattouw> Oh AM right [16:02:07] <Simetrical> RoanKattouw, California is Pacific time, not Mountain time. [16:02:14] <TimStarling> our time system wasn't invented by computer programmers [16:02:26] <RoanKattouw> Simetrical: I know. Just wanted to point out that it was 10 AM (not PM, typo) *somewhere* in the US [16:02:27] <TimStarling> 12pm - 2 = 10pm ??? [16:02:39] <TimStarling> damn well should be [16:02:42] <Simetrical> RoanKattouw, my point was your AM vs. PM slip. [16:02:46] <RoanKattouw> I mixed up. Dutch doesn't define AM and PM [16:03:06] <Barnicle> is Brion the same person who work the AuthPlugin? [16:03:11] <Barnicle> wrote [16:03:16] <RoanKattouw> We just say it's 3 past 6 in the afternoon/evening or 18:03, never 6:03 PM [16:03:31] <RoanKattouw> Barnicle: Possibly. The Brion we're talking about is Brion Vibber [16:03:38] <Barnicle> yes, same [16:03:48] <TimStarling> maybe the dutch had more mathematically-inclined people inventing their time conventions [16:03:56] <Barnicle> AuthPlugin is like LDAP? [16:04:04] <RoanKattouw> TimStarling: More than the English, anywy [16:04:43] <RoanKattouw> I still don't know whether 12 PM is noon or midnight [16:05:04] <Simetrical> RoanKattouw, usually noon. But I've heard it varies by location. [16:05:11] <RoanKattouw> Right... that's useful [16:05:48] <RoanKattouw> But Tim is right, our date and time system is one of the most brain-damaged inventions ever [16:05:53] <wikibugs> 03(mod) $wgServer lacks brackets in IPv6 URLs - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=14977 +comment (10tstarling) [16:06:17] <Simetrical> There's a great quote on that. [16:06:30] <RoanKattouw> Yeah I'm struggling to remember it [16:06:35] <Simetrical> http://www.gnu.org/software/tar/manual/html_node/Date-input-formats.html [16:06:40] <Simetrical> Our units of temporal measurement, from seconds on up to months, are so complicated, asymmetrical and disjunctive so as to make coherent mental reckoning in time all but impossible. Indeed, had some tyrannical god contrived to enslave our minds to time, to make it all but impossible for us to escape subjection to sodden routines and unpleasant surprises, he could hardly have done better than handing down our present system. It is like a [16:06:40] <Simetrical> set of trapezoidal building blocks, with no vertical or horizontal surfaces, like a language in which the simplest thought demands ornate constructions, useless particles and lengthy circumlocutions. Unlike the more successful patterns of language and science, which enable us to face experience boldly or at least level-headedly, our system of temporal calculation silently and persistently encourages our terror of time. [16:06:41] <Simetrical> ??? It is as though architects had to measure length in feet, width in meters and height in ells; as though basic instruction manuals demanded a knowledge of five different languages. It is no wonder then that we often look into our own immediate past or future, last Tuesday or a week from Sunday, with feelings of helpless confusion. ??? [16:06:44] <RoanKattouw> Yeah the GNU manual [16:06:45] <Simetrical> ??? Robert Grudin, Time and the Art of Living. [16:06:47] <Simetrical> Wow, that was too long. [16:07:12] <TimStarling> 12pm = noon, FWIW [16:08:10] <TimStarling> perhaps because PM stands for post-meridian, and 12:01 is indeed post-meridian [16:08:41] <TimStarling> and it wouldn't make sense to make noon 12am but have one minute later be 12:01 pm [16:09:28] *Jack_Phoenix waves to SysRq868 [16:09:31] <RoanKattouw> Well since meridian is noon, noon itself is neither AM or PM by definition [16:09:46] *SysRq868 waves to Jack_Phoenix [16:09:47] <CIA-55> 03yaron * r38720 10/trunk/extensions/ReplaceText/SpecialReplaceText.php: [16:09:47] <CIA-55> New class for handling 'ReplaceText' special page; contains most of the [16:09:47] <CIA-55> extension's functionality [16:09:55] <TimStarling> it's just mnemonic [16:10:00] <RoanKattouw> (mathematically speaking, but then time is a mathematical nightmare) [16:10:09] <TimStarling> you can't actually expect to understand the mind of whatever genius made this stuff up [16:10:31] <CIA-55> 03yaron * r38721 10/trunk/extensions/ReplaceText/ReplaceText.i18n.php: Added 'replacetext_invertselections' value [16:11:07] <Barnicle> Is there a way to have mediawiki recognize a user logged into an internal system, using LDAP and their credentials, without having them to log into the wiki? So, once they log into their computers, the wiki already knows person X has read-only permissions, and what not. [16:11:37] <RoanKattouw> Barnicle: There is an extension on www.mediawiki.org that does that. Search that site for LDAP [16:11:48] <TimStarling> good night [16:11:51] <Barnicle> i'll check it out. [16:11:53] <Barnicle> thanks [16:12:08] <CIA-55> 03yaron * r38722 10/trunk/extensions/ReplaceText/ReplaceTextJob.php: Bug fix so that text replacement doesn't override the global $wgUser variable [16:13:38] <CIA-55> 03yaron * r38723 10/trunk/extensions/ReplaceText/ReplaceText.php: Most functionality moved to SpecialReplaceText.php [16:13:56] <CIA-55> 03yaron * r38724 10/trunk/extensions/ReplaceText/README: New version: 0.2.4 [16:23:52] <wikibugs> 03(NEW) Make browser headers information available through WikiStats - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=15059 15enhancement; normal; Wikimedia: Usage Statistics; (Wiki.Melancholie) [16:24:30] <wikibugs> 03(mod) Make browser headers information available through Wikistats - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=15059 summary (10Wiki.Melancholie) [16:24:53] <Simetrical> Damn, I missed Tim. [16:25:09] <Simetrical> I wanted him to review ObjectArray. http://svn.wikimedia.org/viewvc/mediawiki?view=rev&revision=38165 [16:27:57] <wikibugs> 03(mod) Redirect parsing is over-permissive; syntactically broken redirects are parsed as redirects to the wrong place. - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=15053 +comment (10Simetrical+wikibugs) [16:35:24] <wikibugs> 03(NEW) Make non-sensitive browser referrers available through Wikistats - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=15060 15enhancement; normal; Wikimedia: Usage Statistics; (Wiki.Melancholie) [16:36:47] *Lady_Aleena blurbles about misplaced end brackets. [16:37:14] <wikibugs> 03(mod) Make browser headers information available through Wikistats - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=15059 (10Wiki.Melancholie) [16:37:40] <Lady_Aleena> !bugzilla [16:37:40] --mwbot-- All bugs in MediaWiki should be reported at https://bugzilla.wikimedia.org. This is also the place to request site configuration changes, new features or enhancements to existing features, although bear the following points in mind before making a feature/enhancement request: 1) If the request is specific to a Wikimedia wiki, please discuss the issue on that wiki first. 2) Consider whether a custom extension would be more appropriate. [16:39:33] <Simetrical> Did you know that the word "chortle" was made up for Lewis Carroll's Jabberwocky? [16:41:15] <alnokta2> why not make the code guides people automatically? specific words -triggers-> specific messages. [16:42:11] <wikibugs> 14(INVALID) Disable FlaggedRevs custom configuration on Russian Wikipedia - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=15056 +comment (10vladimir.medeyko) [16:42:48] <Simetrical> alnokta2, what? [16:45:35] <alnokta3> Simetrical, mwbot [16:46:10] <Simetrical> You mean make mwbot trigger when someone randomly says a word in the middle of a sentence? [16:46:18] <Simetrical> Because it would be stabbed to death for annoyingness, that's why. [16:46:27] <Simetrical> An early attempt got banned by domas for that nonsense. [16:46:38] <Simetrical> It tends to be irrelevant half the time. [16:48:15] <wikibugs> 03(mod) default uselang problems - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=14802 (10brion) [16:51:31] <CIA-55> 03rotem * r38725 10/trunk/phase3/languages/messages/MessagesHe.php: Localization update for he. [16:58:25] <DMF> any mediawiki experts interested in a (quick?) job? i have 10 hours tops and willing to pay $50 via paypal and another $50 if you complete it (make sure you have a paypal account since that's the only way i can pay you). must be good at writing templates for mediawiki 1.11.0 -- i prefer if you evaluate the task first to see if it's do-able in the given timeframe. message me asap (only if you're serious about it). [17:11:21] <CIA-55> 03yaron * r38726 10/trunk/extensions/ReplaceText/SpecialReplaceText.php: Added title as a hidden field in the various forms [17:15:52] <DMF> anyone here? not sure if my last message was visible to the channel (i'm new to freenode.net) [17:21:45] <CIA-55> 03ialex * r38727 10/trunk/phase3/includes/Diff.php: For compatibility with PHP 5.0, it does not accept array as param type [17:23:07] <Dantman|afk> yes it was [17:23:31] <DMF> cool, thx, you wouldn't happen to be an expert in wiki code would you? [17:28:04] <Dantman> well ya, but I don't even know what you want [17:30:14] <ronparker1> Hello and help :D [17:33:16] <ialex> ronparker1: just ask your question, maybe somebody will answer it [17:35:31] <CIA-55> 03ialex * r38728 10/trunk/phase3/includes/specials/ (SpecialRecentchanges.php SpecialRecentchangeslinked.php): [17:35:31] <CIA-55> Fixes for r38707: [17:35:31] <CIA-55> * $wgOut is already passed to the function in the $out param [17:35:31] <CIA-55> * Don't throw a fatal error if the target param is set but to an invalid/empty [17:35:31] <CIA-55> title (Fatal error: Call to a member function getPrefixedText() on a non-object [17:35:33] <CIA-55> in ./includes/specials/SpecialRecentchangeslinked.php on line 169) [17:41:29] <CIA-55> 03demon * r38729 10/trunk/phase3/includes/ (Skin.php specials/SpecialPreferences.php): Logic for $wgSkipSkins should be in Skin.php, where it's usable by things other than preferences. (bug 6795) [17:41:41] <wikibugs> 03(FIXED) Usable Skin Names Logic Belongs in Skin Class - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=6795 +comment (10innocentkiller) [17:56:26] <wikibugs> 03(mod) Add variant of MediaWiki: Pagetitle for customizing the main page - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=15007 (10Simetrical+wikibugs) [17:56:38] <wikibugs> 03(mod) Add variant of MediaWiki: Pagetitle for customizing the main page - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=15007 (10Simetrical+wikibugs) [18:02:52] <wikibugs> 03(mod) Add variant of MediaWiki: Pagetitle for customizing the main page - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=15007 +comment (10Simetrical+wikibugs) [18:03:35] <ronparker1> Hello, Help, I've just installed the extension PDF, and i got this error message Warning: require_once(PdfExport.i18n.php) [function.require-once]: failed to open stream: , whats wrong? [18:04:46] <will_hough> You are missing the file [18:05:08] <ialex> or the file is not in the include_path [18:05:23] <Dolledre> stupid question: what's the mediawiki equivalent of $LOGNAME ? [18:07:00] <ialex> Dolledre: what is $LOGNAME? [18:07:13] <Dolledre> whoami [18:07:23] <wikibugs> 03(mod) Implementing "bigger" and "smaller" parameters for images will allow flexibility without overruling default thumb size preferences . - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=15054 +comment (10Eugene.Zelenko) [18:08:58] <Dolledre> i mean the same thing as wgUserName ialex, but without javascript [18:09:18] <ronparker1> will_hough: but its installed in the directory /extensions/dompdf [18:09:20] <ialex> $wgUser->getName(); [18:09:38] <ronparker1> will_hough: or where should i find that file PDFexport.i10n.php [18:09:39] <ronparker1> ? [18:09:44] <Dantman> be sure to 'global $wgUser;' [18:10:16] <Dolledre> ok tnx [18:10:45] <Lady_Aleena> Is there a way to shorten the following very long line? [18:10:46] <Lady_Aleena> {{#if:{{#ifeq:{{NAMESPACE}}|>2}}{{#ifeq:{{{class|}}}|Disambig}}{{#ifeq:{{{class|}}}|Merge}}{{#ifeq:{{{class|}}}|Redirect}}|do something}} [18:11:29] <Lady_Aleena> I feel I should know this, but for some reason it is eluding me. [18:11:50] <ialex> ronparker1: try to replace require_once( 'PdfExport.i18n.php' ); by require_once( dirname( __FILE__ ) . '/PdfExport.i18n.php' ); [18:13:53] <^demon> brion: Dunno if Roan poked you earlier or not (haven't read the scrollback :-p), but he posted a new patch on the redirect changes bug. [18:15:18] <wikibugs> 03(mod) Add variant of MediaWiki: Pagetitle for customizing the main page - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=15007 +comment (10Simetrical+wikibugs) [18:16:38] <Simetrical> brion, could you review this at some point for coding style? (And for whether we want it . . . but it makes a lot of sense to customize the main page in this way, IMO.) https://bugzilla.wikimedia.org/show_bug.cgi?id=15007#c7 [18:17:35] <brion> i'm leery of it on principle but in practice people seem to love that sort of thing, so prolly :) [18:21:19] *Lady_Aleena is getting a stress headache. [18:23:21] <alnokta2> That's not technically AJAX. And more to the point, JavaScript hacks [18:23:21] <alnokta2> don't need to be discussed on Wikitech-l. They're for people without [18:23:21] <alnokta2> commit access. [18:24:16] <Simetrical> Yes? [18:24:36] *Simetrical goes off to schedule a doctor's appointment and get some exercise [18:26:31] <c0daine> hi, I need a way to access my extension through api.php (or someway else) to handle POST requests, so that my extension can use MW libs to validate session and user. Can someone help me with this? [18:28:31] <Lady_Aleena> Are the regulars here getting sick of my constant questions about the use of parser functions? I know that there are times when I can be annoying. [18:30:24] <ronparker1> help , why the PDF extension drops this error exception 'DOMPDF_Exception' with message 'Box property calculation requires containing block width' ? [18:37:59] <cobalt027> hi all. [18:38:11] <ronparker1> does anyone how to make the pdf conversion work in shared hosting>? [18:38:19] <cobalt027> is there a tag that will display on a media wiki page the last updated time? [18:38:50] <cobalt027> something like ${LastUpdatedTime}. not a sig. something that will continue to change as the page is updated? [18:39:22] <Raymond_> cobalt027: it's in the footer of every page [18:39:31] <wikibugs> 03(mod) {{int:X}} respects user-defined interface language, breaking link tables etc. (aka {{USERIFCODE}} strikes back) - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=14404 (10mormegil) [18:40:11] <cobalt027> Raymond_: heh, thanks. i never bothered to scroll down that far to notice. Chalk one up for the ass factor. ;) [18:40:26] <Raymond_> :) [18:40:29] <alnokta2> how to take a screenshot like this? http://en.wikipedia.org/wiki/Image:EnglishWikipediaMainPage3rdAugust2007.PNG [18:43:11] <alnokta2> i mean a screenshot showing from top to button [18:44:01] <solifugus> In writing an extension, can I use the same database as mediawiki and just add a new table? [18:44:20] <FunPika> yes [18:44:40] <solifugus> ok [18:45:21] <wikibugs> 03(FIXED) Request to add contents of $wgAddGroups and $wgRemoveGroups to Special: ListGroupRights. - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=13846 +comment (10stevend811) [18:47:36] <Nikerabbit> VasilievVV: re r38424, group names are not i18ned [18:48:12] <c0daine> is there a way to change namespace delimiter? (say, ':' to '/' so that URLs would look like site.com/Special/Version) [18:48:17] <solifugus> I am making my extension use the wiki's database connection info by default, but can be overridden in LocalSettings.php to specify otherwise.. [18:48:41] <Nikerabbit> c0daine: nope [18:50:58] <Prodego> That would cause a problem too [18:51:06] <Prodego> you could create Special/Version potentially [18:51:19] <Prodego> where version is a subpage of special [18:53:05] <c0daine> :) I'd be willing to sacrifice subpage functionality [19:00:24] <Simetrical> Lady_Aleena, if we get sick of you we'll either tell you so or ignore you. Don't worry. :) [19:00:57] <Simetrical> c0daine, no. [19:01:03] <Simetrical> It's probably hardcoded all over the place. [19:01:14] <Lady_Aleena> Simetrical, sometimes I do, since it seems that I always get something wrong. [19:05:38] <Simetrical> brion, regarding the bug link from before (15007), do you think OutputPage::setRawPageTitle() would be appropriate to dodge the automatic wfMsg( 'pagetitle' ) wrapper? Or some other way? [19:10:08] <brion> Simetrical: i'm not quite sure in what context here [19:10:14] <c0daine> and, is there documentation about extending api.php? [19:10:52] <Simetrical> brion, basically, we want the main page to have its own <title>. This shouldn't have "- From Wikipedia, the free encyclopedia" or whatever appended to it. OutputPage has no method that allows you to set the title without that getting appended. [19:11:32] <brion> hmm [19:11:36] <Simetrical> brion, you could special-case wfMsg( 'mainpage' ) as input in OutputPage::setPagetitle(), which is what the patch does. But that sucks. So I was thinking setRawPageTitle would be better. [19:11:59] <brion> Simetrical: so Article::view would special-case which func it calls when doing the page view setup? [19:12:02] <brion> sounds sensible [19:12:08] <Simetrical> Right. [19:14:19] <Simetrical> Oh, wait, OutputPage::setHTMLTitle() already does this? [19:14:20] *Simetrical prods [19:15:54] <wikibugs> 03(mod) Cite anchors should be numbered starting at 1 to correspond with on-screen labels - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=10537 -easy -need-review +reviewed; +comment (10ssanbeg) [19:16:44] <solifugus> How can I get the manual for parser functions.. Manual:Parser_functions is empty.. [19:16:46] <wikibugs> 03(mod) Cite anchors should be numbered starting at 1 to correspond with on-screen labels - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=10537 -reviewed (10ssanbeg) [19:17:07] <wikibugs> 03(mod) Entering Croatian characters produces errors - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=15058 (10brion) [19:18:47] <Simetrical> !parserfunctions | solifugus [19:18:47] --mwbot-- solifugus: "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 <http://meta.wikimedia.org/wiki/Help:Parser_function>. For details about the extension, see <http://meta.wikimedia.org/wiki/Help:ParserFunctions>. [19:18:59] <Simetrical> solifugus, there's no manual page because it's not part of the core software. You want the extension page. [19:19:49] <solifugus> ok.. the documentation on the mediawiki website for writing extensions says to go to Manual:Parser functions page [19:20:52] <CIA-55> 03simetrical * r38730 10/trunk/phase3/ (6 files in 4 dirs): (bug 15007) New 'pagetitle-view-mainpage' message allows the HTML <title> of the main page to be customized. [19:20:55] <Simetrical> Oh, damn, I forgot to credit him in the commit message. [19:20:58] <Simetrical> Oh well, he's in CREDITS. [19:20:59] <Simetrical> :) [19:23:48] <Freddy1990> Hey, how do I enable svg uploads on my mediawiki? [19:24:45] <Reedy> http://www.mediawiki.org/wiki/SVG [19:26:37] <wikibugs> 03(FIXED) Add variant of MediaWiki: Pagetitle for customizing the main page - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=15007 +comment (10Simetrical+wikibugs) [19:27:26] <gegema> am working on an infobox template and just ran into this...one of my functions {{#if:{{foo|}}} has a list of variables that can correspond to it. bar1, bar2, bar3... so when doing my infbox, instead of typing foo = bar1, I wanted to do, foo = b1 and have that b1 correspond to bar1.. I know am not being clear enough..but I wanted to learn how to insert a template within a template [19:28:13] <wikibugs> 03(mod) Tries to send e-mails to users without an e-mail address - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=15055 (10brion) [19:31:06] <alnokta2> that is so stupid [19:32:12] <alnokta2> i uploaded a file without adding an extension to the second form for name, after it finished it told me that the file doesn;t have an extension and i have to reupload! [19:32:57] <alnokta2> shouldn't the warning appear before starting to upload? [19:34:04] <wikibugs> 03(mod) Skins should add their name as a class in <body> - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=15052 (10brion) [19:35:19] <tusho> can a skin have its own print stylesheet? [19:35:41] <wikibugs> 03(mod) Trailing slashes should be ignored when subpages are enabled - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=11747 (10brion) [19:36:08] <MZMcBride> tusho: Usually @media is used on the specific skin. [19:36:12] <MZMcBride> Specific skin page, rather. [19:36:17] <tusho> True. [19:36:25] <tusho> Then why are there printable version links? [19:37:41] <brion> tusho: it can if it likes, sure [19:37:55] <brion> however i generally prefer the print style to be the common one [19:38:04] <tusho> brion: my skin sets a background [19:38:11] <tusho> I don't wanna waste people's ink [19:38:12] <tusho> :) [19:38:14] <alnokta2> brion, what do you think, am i totally irrelevant? [19:38:23] <brion> alnokta2: an extension check could be done in JS before form submit, that would be handy [19:38:34] <brion> make sure it's in bz if not already [19:38:55] <AaronSchulz_> brion: how is that css fix going? :DD [19:39:01] <alnokta2> thx, good to know i am sane [19:39:09] <tusho> speaking of which how can you disable the tracking of page view count? [19:39:31] <brion> AaronSchulz_: you could give it a shot ;) [19:39:43] <AaronSchulz_> hey, I didn't break it [19:39:46] <brion> ;) [19:40:11] <brion> let's not assign svn blame now [19:42:05] <wikibugs> 14(INVALID) {{SUBPAGENAME}} and {{SUBPAGENAMEE}} don' t work on all namespaces - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=15051 +comment (10brion) [19:44:03] <wikibugs> 03(mod) Tries to send e-mails to users without an e-mail address - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=15055 (10lalinsky) [19:49:12] <wikibugs> 03(mod) searching CU log for usernames with a hyphen gives unspecific results - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=15049 (10brion) [19:49:21] <brion> AaronSchulz_: feel like poking that CU bug? looks like IP::parseRange() needs adjusting [19:49:42] *AaronSchulz_ is busy making a stats page [19:49:50] <AaronSchulz_> brion: are you doing that css fix? [19:50:05] <brion> not at this exact moment, no :) [19:51:05] <siebrand> AaronSchulz_: is message 'revreview-basic-i' in use? I couldn't find where the "*-i" messages are used. [19:53:24] <AaronSchulz_> doesn't seem to be [19:53:30] <AaronSchulz_> revreview-basic-i and revreview-quality-i [19:54:23] <siebrand> AaronSchulz_: oki, will remove those. [19:55:29] <tusho> What should I do in a skin to generate a link to the main page? [19:55:33] <tusho> With a specific text [19:57:39] <ialex> [[Main Page|text to display]] [19:58:02] <tusho> ialex: "In a skin" [19:58:18] <CIA-55> 03ialex * r38731 10/trunk/phase3/includes/specials/SpecialListgrouprights.php: [19:58:19] <CIA-55> Tweaks to Special:ListGroupRights: [19:58:19] <CIA-55> * Allow wikitext in messages [19:58:19] <CIA-55> * Per Nikerabbit: use User::makeGroupLinkWiki() for groups [19:58:43] <MrHappy> Hi, quick question, I have 1 wiki server, I would like another in a different location, in case the first one breaks, how do I got about doing that and making sure they are in sync [19:58:43] <Nikerabbit> :o [19:58:46] <Nikerabbit> I didn't say that [19:59:26] <ialex> tusho: $this->makeLinkObj( Title::newMainPage(), 'some text to display' ); [19:59:42] <Duesentrieb> MrHappy: read up on mysql replication [20:00:31] <Duesentrieb> MrHappy: mediawiki itself does not support hot switching, but changing the master db to another server is just a matter of chaning a config variable. [20:00:58] <CIA-55> 03siebrand * r38732 10/trunk/extensions/FlaggedRevs/language/FlaggedRevs.i18n.php: [20:00:58] <CIA-55> Localisation updates [20:00:59] <CIA-55> * remove 3 unused messages, rebuild file [20:01:01] <Duesentrieb> basically, mediawiki doesn't help you to do it, but is quite happy to live in a redundant environment. [20:01:14] <Duesentrieb> you can even use you backup as a read-only slave server, to improve performance [20:02:05] <wikibugs> 03(mod) $wgServer lacks brackets in IPv6 URLs - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=14977 (10gillou.ray) [20:02:26] <AaronSchulz_> siebrand: opps, seems they are [20:02:36] <siebrand> AaronSchulz: are? [20:02:40] <MrHappy> Duesentrieb, thanks. [20:02:53] <siebrand> AaronSchulz: all 3? [20:02:54] <alnokta2> mediawiki is a she [20:03:15] <AaronSchulz> $msg = $quality ? 'revreview-quality' : 'revreview-basic'; [20:03:16] <AaronSchulz> if( $synced ) { [20:03:18] <AaronSchulz> $msg .= '-same'; [20:03:20] <AaronSchulz> } else if( $revsSince == 0 ) { [20:03:21] <AaronSchulz> $msg .= '-i'; [20:03:22] <AaronSchulz> } [20:03:41] <CIA-55> 03river * r38733 10/trunk/switchboard/ (5 files): should have a way to limit max queued processes per user (incomplete fix; urgent for hemlock) [20:03:47] <siebrand> AaronSchulz: argh. I'll revert *and* add some comments to the code to ease grepping. [20:04:06] <AaronSchulz> yeah, the others too [20:04:29] <siebrand> AaronSchulz: well, you know better than me, so I'll stick to these 3 :) [20:06:35] <CIA-55> 03rotem * r38734 10/trunk/phase3/languages/Language.php: Fixing comment of Language::tsToHebrew: Month 1 is Tishrei. [20:19:29] <siebrand> Simetrical: re wikitech-l '': see extension BackAndForth. [20:19:43] <wikibugs> 03(mod) Entering Croatian characters produces errors - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=15058 (10radoslav.dejanovic) [20:19:49] <siebrand> Simetrical: ehr: "Re: [Wikitech-l] Next/previous articles" that is :) [20:21:50] <wikibugs> 03(mod) $wgServer lacks brackets in IPv6 URLs - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=14977 +comment (10dan_the_man) [20:21:54] <siebrand> AaronSchulz: done. [20:21:57] <CIA-55> 03siebrand * r38735 10/trunk/extensions/FlaggedRevs/ (3 files in 2 dirs): [20:21:57] <CIA-55> * Revert r38732. Aaron found they are used after all [20:21:57] <CIA-55> * Might as well add some comment to not have this happen again [20:21:57] <CIA-55> * Fix indentation and remove whitespace [20:24:07] <CIA-55> 03siebrand * r38736 10/trunk/extensions/Translate/Translate.i18n.php: [20:24:07] <CIA-55> Localisation updates. Fix typo. Thanks to Rotem Liss for noticing. [20:24:07] <CIA-55> No localisation updates needed. [20:28:21] <Lady_Aleena> Still wondering if there is a shorter way to do the following... [20:28:22] <Lady_Aleena> {{#if:{{#ifeq:{{NAMESPACE}}|>2}}{{#ifeq:{{{class|}}}|Disambig}}{{#ifeq:{{{class|}}}|Merge}}{{#ifeq:{{{class|}}}|Redirect}}|do something}} [20:33:57] <CIA-55> 03catrope * r38737 10/trunk/phase3/ (RELEASE-NOTES includes/Title.php): * (bug 15053) Syntactically incorrect redirects with another link in them no longer redirect to the second link [20:34:28] <wikibugs> 03(FIXED) Redirect parsing is over-permissive; syntactically broken redirects are parsed as redirects to the wrong place. - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=15053 +comment (10roan.kattouw) [20:38:15] <solifugus> A parser function calls my function and has me replace it with HTML text in its place, correct? [20:39:31] *Vandale_172 stabs his connection [20:40:33] <wikibugs> 03(mod) searching CU log for usernames with a hyphen gives unspecific results - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=15049 (10maxsem.wiki) [20:41:10] <solifugus> The documentation on writing an extension lies.. It says Manual:Parser_functions will gives specifics on how to build your hooks for a parser function.. but all I get is a blank page.. [20:41:16] <solifugus> empy page... wiki page [20:41:34] <solifugus> Help:Parser_function just tells me what one is and lists the build-in ones [20:42:09] <ialex> read docs/magicwords.txt in 1.13, it might help a bit [20:42:53] <Duesentrieb> the relation between magic words and parserfunctions is confusing. and the procedure to hook up a parser function is simply disgusting. [20:44:48] <solifugus> yes.. i am confused [20:44:48] <wikibugs> 03(mod) Enable FlaggedRevs custom configuration on en.wikibooks - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=14618 (10mike.lifeguard) [20:45:10] <solifugus> it's be nice if someone could just fix the documentation on how to write extensions [20:47:38] <Duesentrieb> solifugus: it depends on the type of extension, of course [20:48:29] <^demon> Duesentrieb: The "examples" extension could be updated. Some of that is rather outdated. [20:48:50] <Duesentrieb> i guess my knowledge might be outdated by now, too :) [20:48:59] <Duesentrieb> at least i hope so [20:49:02] <ialex> hmm, somebody can replace 'parse' by 'parseinline' that I introduced in includes/specials/SpecialListgrouprights.php [20:49:12] <ialex> I can't commit anything [20:49:22] <solifugus> Duesentrieb: I was using a printed version of the documentation.. I just noticed that the online version has a link to Manual:Parser_functions [20:50:10] <^demon> Does the description still point to meta, or has that finally been changed too? [20:50:17] <^demon> ialex: Updating my copy, then on it. [20:50:26] <Duesentrieb> *printed* manualy? omg havnt' used one of those since the 90s :P [20:50:32] <Duesentrieb> err, manuals [20:50:39] <solifugus> Duesentrieb: What approach is best--I've been tasked with producing an extension for adding a little feedback form in wiki pages.. a little javascript to collect some info and then save to database.. [20:50:57] <^demon> Who reads manuals? ;-) [20:51:00] <CIA-55> 03ialex * r38738 10/trunk/phase3/ (RELEASE-NOTES includes/specials/SpecialListgrouprights.php): [20:51:00] <CIA-55> * Tweak for my last commit: parse -> parseinline [20:51:00] <CIA-55> * Add bug no in RELEASE-NOTES [20:51:08] <^demon> Bah. [20:51:09] <ialex> ^demon: finally OK [20:51:16] <^demon> What happened to can't commit anything? [20:51:20] <ialex> tried at least ten times :( [20:51:38] <ialex> ^demon: 30 kpbs connection [20:51:46] <solifugus> Duesentrieb: it would be like: Helped? (Y)es or (N)o; if no then drop-down list of reasons why not.. of which "other" would present a textbox to enter a comment free hand. [20:51:48] <Duesentrieb> solifugus: well... js can't save to the database. to do that, you need some type of hanlder. it could be an ajax handler, and action handler, or a special page. [20:51:53] <^demon> Ouch :( [20:52:30] <Duesentrieb> solifugus: to generate the form on page, you can use a parser hook or parser function. old-style parser hooks are a bit simpler to write but less flexible. but there's the generic #tag function to get around that. [20:52:53] <ialex> ssh doesn't like it :) [20:53:11] <Duesentrieb> solifugus: what the form looks like, and how it works internally with js or whatever, doesn't matter really to how you structure the extension [20:53:12] <^demon> ialex: More bandwith for Christmas? [20:53:46] <ialex> ^demon: normally I have a 3500 kpbs connection, but I'm not at home [20:54:12] <solifugus> Duesentrieb: and to bring the data back, you suggest I just write a separate little php file that does its work and maybe redirects back to mediawiki? [20:54:41] <CIA-55> 03brion * r38739 10/trunk/phase3/ (RELEASE-NOTES includes/IP.php): [20:54:41] <CIA-55> * (bug 15049) Fix for CheckUser extension's log search: usernames containing [20:54:42] <CIA-55> a "-" were incorrectly turned into bogus IP range searches. [20:54:42] <CIA-55> Patch by Max Semenik. [20:54:43] <Duesentrieb> no. i wrote "to do that, you need some type of hanlder. it could be an ajax handler, and action handler, or a special page." [20:55:16] <Duesentrieb> you yould use a separate php file, but then you need to deal with db connections and all that again. [20:55:17] <solifugus> Duesentrieb: ok.. so you mean that's possible with mediawiki somehow. [20:55:25] <Duesentrieb> sure [20:55:39] <wikibugs> 03(FIXED) searching CU log for usernames with a hyphen gives unspecific results - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=15049 +comment (10brion) [20:56:16] <wikibugs> 03(NEW) $wgGroupsAddToSelf and $wgGroupsRemoveFromSelf missing in Special: Listgrouprights - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=15061 15enhancement; normal; MediaWiki: Special pages; (robin_1273) [20:56:24] <solifugus> Duesentrieb: what's the easiest approach among those three? I don't have a clue (yet) how any of them work. [20:57:12] <Duesentrieb> using the ajax approach is probably nicest, if you are confortable dealing with that on the JS side. [20:57:27] <^demon> Duesentrieb: Seeing as the examples don't even mention wfLoadExtensionMessages() (using the older method of hooking into LoadAllMessages). [20:57:30] <^demon> Back shortly. [20:57:40] <Duesentrieb> ^demon: sofixit :) [20:57:51] <solifugus> Duesentrieb: i feel fine with ajax.. where can I find info on how to make an ajax handler? [20:57:58] <Duesentrieb> !ajax [20:57:58] --mwbot-- http://www.mediawiki.org/wiki/Manual:Ajax [20:58:06] <solifugus> owesome [20:58:26] <CIA-55> 03purodha * r38740 10/trunk/phase3/languages/messages/MessagesNah.php: [20:58:26] <CIA-55> Added $namespaceAliases to new Nahuatl localization reflecting the [20:58:26] <CIA-55> previous Spanish names which may still be present in Nahuatl pages. [20:58:34] <Duesentrieb> that leaves the part of generating the actual form on the wiki page [20:59:02] <Duesentrieb> write a parser hook for that. it's a bit simpler, and it's easy to convert to a parser function later if you want. [20:59:13] <Simetrical> Lady_Aleena, how would {{#ifeq:{{NAMESPACE}}|<2}} ever be true? [20:59:18] <Duesentrieb> unless you want this thing on every page. that gets tricky, because the skins are nasty to hook into. [20:59:39] <Lady_Aleena> Simetrical, if the item were in article or article talk space? [20:59:49] <solifugus> Duesentrieb: no only certain pages.. They are help pages for a software application we publish. [21:00:02] <Lady_Aleena> Actually, just article space, because {{NAMESPACE}} 2 is talk space for articles. [21:00:15] <Duesentrieb> so write a tag (aka parser hook) for that. [21:00:17] <Lady_Aleena> Also, special pages are negative numbers. [21:00:33] <Duesentrieb> Lady_Aleena: but {{NAMESPACE}} gives a name, not a number [21:00:57] <CIA-55> 03river * r38741 10/trunk/switchboard/ (6 files): instead of letting the web server return FastCGI Protocol Error when something bad happens, print an error message to the user. [21:01:32] <Lady_Aleena> Duesentrieb, so, how would you do what I am attempting to do? I am really interested. [21:01:48] <Duesentrieb> flyingparchment: is that optional? sounds like it should be (but then, i don't know much about fastcgi internals) [21:01:52] <solifugus> Duesentrieb: ok.. I will start reading/working.. I am thinking maybe something to the effect of {{#helped: was irrelevent | too little detail | was incorrect }} -- where the arguments define the dropdown list. [21:01:55] <flyingparchment> Duesentrieb: no [21:01:59] <flyingparchment> why would it be? [21:02:37] <Duesentrieb> flyingparchment: i dunno - giving a choice between "convenient" vs "compliant" behaviour? or humand readable vs. machine readable? [21:02:50] <flyingparchment> Duesentrieb: i don't think this is any less compliant [21:02:53] <Duesentrieb> flyingparchment: no big deal, it just ocurred to me. [21:03:02] <flyingparchment> Duesentrieb: the previous behaviour was to immediately drop the server connection, which if anything is less compliant [21:03:06] <Simetrical> Lady_Aleena, um, #ifeq tests equality. It doesn't accept comparison operators. [21:03:10] <flyingparchment> (that's what made the server generate its own error) [21:03:12] <Duesentrieb> flyingparchment: ugh :P [21:03:26] <Simetrical> Lady_Aleena, {{#ifeq:{{NAMESPACE}}|<2}} would be true if {{NAMESPACE}} evaluated to the literal name "<2". [21:03:40] <Duesentrieb> Lady_Aleena: i dunno what you are trying to do... and maybe there's a magic word for the namespace number? {{NAMESPACENUM}} or something? no idea [21:03:51] <Simetrical> Lady_Aleena, you want {{#ifexpr:{{NAMESPACE}} < 2}}, if {{NAMESPACE}} returned a number. [21:04:05] <Lady_Aleena> I am trying to make the following short... [21:04:06] <Lady_Aleena> {{#if:{{#ifeq:{{NAMESPACE}}|>2}}{{#ifeq:{{{class|}}}|Disambig}}{{#ifeq:{{{class|}}}|Merge}}{{#ifeq:{{{class|}}}|Redirect}}|do something}} [21:04:33] <Lady_Aleena> I was hoping the last three...(phone call) [21:04:37] <Simetrical> Lady_Aleena, you can just delete the whole thing, and it would have the same effect. What you wrote is never going to be true in any reasonable scenario. [21:05:55] <flyingparchment> Duesentrieb: fwiw, the 'error' is just a fake fastcgi response (as if the PHP script had been <? header("HTTP/1.0 500 Internal server error"); echo "something bad happened." ?>). the other option is to return a non-zero exit code, but i don't see any situation that'd be preferable [21:06:41] <Duesentrieb> flyingparchment: you are the expert :) [21:06:51] <wikibugs> 03(mod) $wgGroupsAddToSelf and $wgGroupsRemoveFromSelf missing in Special: Listgrouprights - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=15061 (10siebrand) [21:07:07] *Duesentrieb goes to implement a multimap for the 100th time [21:07:45] <Raymond_> Duesentrieb: whats a multimap? [21:08:11] <Duesentrieb> a map where there may be more than one value to a key [21:08:24] <Raymond_> ok [21:13:40] <CIA-55> 03siebrand * r38742 10/trunk/extensions/ (47 files in 37 dirs): Localisation updates for extension messages from Betawiki (2008-08-06 23:05 CEST) [21:27:10] <aib_> is there a way to set article background? [21:27:14] <CIA-55> 03siebrand * r38743 10/trunk/phase3/languages/messages/ (174 files): Localisation updates for core messages from Betawiki (2008-08-06 23:05 CEST) [21:27:16] <aib_> set an article background image?\ [21:33:41] <wikibugs> 03(NEW) Don't show diff on edit page if user can't review changes - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=15062 15enhancement; normal; MediaWiki extensions: FlaggedRevs; (vasilvv) [21:39:00] <wikibugs> 03(mod) Don't show diff on edit page if user can't review changes - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=15062 (10stasielvova) [21:43:38] <Simetrical> Hmm. Why might a call to xcache_set() fail? [21:44:19] <Duesentrieb> out of space? [21:47:42] <Simetrical> Plenty of space. [21:48:04] <Simetrical> Anyway, then it uses LRU to kick out old cached items, doesn't it? [21:48:39] <Simetrical> Oh, wait. [21:48:44] <Simetrical> I was reading the little charts wrong. [21:48:47] <Simetrical> It's horribly out of space. [21:48:56] <Simetrical> I assumed that the bar would be filled in blue if it's *used*. [21:48:57] <Simetrical> Ahahaha. [21:49:02] <Simetrical> Let's try this again. [21:49:06] <TigerTails> would it be possible to join MediaWiki accounts with say... SimpleMachinesForums accounts? [21:50:21] <metanil> how hard is it to upgrade from 1.12 to 1.13rc? [21:50:57] <Simetrical> metanil, very easy if you have shell access. Not hard if you only have FTP access, either. [21:51:06] <Simetrical> TigerTails, if someone's written an appropriate extension, yes. [21:51:43] <metanil> ???Simetrical: any links for how tos? [21:52:58] <CIA-55> 03brion * r38744 10/trunk/extensions/Collection/ (Collection.body.php Collection.php): (log message trimmed) [21:52:58] <CIA-55> Couple of changes: [21:53:00] <CIA-55> * Support the sidebar extension hooks in current MediaWiki for the little Collections portlet. [21:53:04] <CIA-55> ** NOTE: We seem to end up with a phantom empty portlet for non-article views if a 'COLLECTION' line is explicitly added to MediaWiki:Sidebar [21:53:06] <Simetrical> !upgrade | metanil [21:53:06] --mwbot-- metanil: http://www.mediawiki.org/wiki/Manual:Upgrading [21:53:07] <CIA-55> ** It probably doesn't play very well with sidebar caching [21:53:09] <CIA-55> ** I disabled this for anonymous visitors to ensure that weird unexpected things don't creep into the general HTTP caches. Not sure how best to handle this? [21:53:12] <CIA-55> * Changed the cache invalidation on add/remove/etc from the page to the user [21:55:17] <Simetrical> Dammit, I *meant* to say month 1 is Tishrei. Somehow I wrote Nissan. [21:55:28] *Simetrical grumbles at giving Rotem all these opportunities to fix his errors lately [22:01:06] <CIA-55> 03siebrand * r38745 10/trunk/phase3/languages/messages/ (54 files): Localisation updates. Formatting, some author updates. [22:02:31] <siebrand> Simetrical: well, even NewUserMessage, with is 5 times smaller than BackAndForth is not being taken care of, unfortunately ;( [22:02:48] <Simetrical> siebrand, because nobody reads the bug reports. [22:04:04] <metanil> does running the update.php script is enough to upgrade database? [22:04:15] <siebrand> Simetrical: https://bugzilla.wikimedia.org/show_bug.cgi?id=14311#c1 ... [22:04:35] <solifugus> Duesentrieb: is setFunctionHook($id, $callback, $flags) a global function--I should just call it in my setup.php file? [22:04:46] <Simetrical> siebrand, yeah, like I said. [22:05:02] <siebrand> Simetrical: I discussed it with brion around that time and got the advice to put it on the review list. [22:06:09] <siebrand> Simetrical: guess I have to get noisier, then :( [22:06:16] <Duesentrieb> solifugus: no and no. look at the example. [22:06:52] <Duesentrieb> it's not even a standalong function [22:07:05] <Duesentrieb> it's a method of the parser object [22:07:31] <Simetrical> siebrand, Tim is better about this, in my experience. [22:07:47] <Emufarmers> Are there any developers free to look at my little patches? [22:07:57] <siebrand> Simetrical: I'll try to persuade Tim then. Thanks. [22:08:00] <solifugus> Duesentrieb: i don't see anything in the example that calls this.. oh wait.. i see it now.. missed that! [22:08:36] <siebrand> Simetrical: where *did* Rob go anyway? Didn't read any goodbyes. He just disappeared. [22:09:45] <ffm|sh> Is there a magic word for the current minute? [22:10:20] <Simetrical> ffm|sh, if it's not {{CURRENTMINUTE}}, I guess not. [22:10:36] <Simetrical> siebrand, on the other hand, he had said goodbye a few times before and then *not* disappeared. ;) [22:12:29] <ffm|sh> thx [22:12:52] <bazz> so, i'm using the dynamicarticlelist extension. i had the problem where the list wasn't refreshing unless i edited the page it was on. i found the solution on the discussion page of putting: $wgCacheEpoch = gmdate( 'YmdHis' ); in LocalSettings, which works, but i'm wondering what that actually does and if it will have any other side effects [22:16:21] <leonardo> hi, on mediawiki template is it possible add for example a city name and automatically show the country? [22:18:31] <Lady_Aleena> I'm back from my phone call. I hate when I am asked to search for something that I know nothing about. [22:18:38] <Duesentrieb> leonardo: no. a template is a reusable piece of text, not some magic oracle [22:19:20] <leonardo> any idea how have something like this? [22:19:25] <Lady_Aleena> Simetrical, is there a way to combine the following (part of the bigger thing, but this is the part that I want to make smaller). [22:19:35] <Lady_Aleena> {{#ifeq:{{{class|}}}|Disambig}}{{#ifeq:{{{class|}}}|Merge}}{{#ifeq:{{{class|}}}|Redirect}} [22:20:31] <leonardo> ? [22:20:37] <Simetrical> Lady_Aleena, what does {{#ifeq:a|b}} even do? Don't you need a third parameter? [22:21:02] <Lady_Aleena> Simetrical, if any of those apply, I want it to do something. [22:21:08] <Simetrical> The answer is no, though, you probably can't combine them. There are no arrays here. [22:22:01] <Lady_Aleena> So, no using Disambig or Merge or Redirect then? [22:22:53] <Lady_Aleena> {{#ifeq:{{{class|}}}|Disambig or Merge or Redirect}} won't work. Darn it. [22:23:23] <Simetrical> I suggest you make sure you understand exactly what #ifeq does before you proceed. It's very simple, but you don't seem to quite get it. [22:23:29] <Simetrical> It compares the first and second arguments, as strings. [22:23:33] <Simetrical> Nothing more, nothing less. [22:23:56] <Lady_Aleena> Ah, strings.. [22:23:57] <Simetrical> For more complicated logic, use #expr or #ifexpr (but those only support numbers, not strings). [22:24:42] <Lady_Aleena> Okay then, I can't combine those three things. Now, how to add if SUBJECTSPACE is not null. [22:26:02] <solifugus> Duesentrieb: ok.. does this look right? http://pastebin.ca/1094225 [22:27:08] <Duesentrieb> solifugus: move the config variable initialization out of the setup function, into global scope. [22:27:08] <solifugus> Duesentrieb: (presuming I didn't leave off the end quote on line 34) [22:27:16] <Lady_Aleena> leonardo, where are you building this template? [22:28:02] <Duesentrieb> solifugus: setting the credits can also be done in global scope. i see no reason to do it in the init function [22:28:19] <leonardo> Lady_Aleena: I figured out something, i can have this using SF SMW [22:28:32] <Lady_Aleena> leonardo, oh, okay... [22:28:59] <leonardo> But i even don't know create advanced templates [22:29:08] <leonardo> just wanted to catalog a city [22:29:12] <leonardo> using mediawiki [22:29:24] <Lady_Aleena> leonardo, a city or any city? [22:29:29] <Duesentrieb> solifugus: otherwise it looks ok on first glance, although you never register that ajax handler [22:29:30] <leonardo> a city [22:29:31] <solifugus> Duesentrieb: ok.. done.. (I should've know better about the config variables) [22:29:54] <solifugus> Duesentrieb: I thought I'd get it working like this first.. then work on the ajax handler part [22:29:57] <leonardo> http://www.gizmogarden.com/index.php/City_Documentation_Project [22:30:08] <leonardo> Lady_aleena: http://www.gizmogarden.com/index.php/City_Documentation_Project [22:30:12] <Duesentrieb> solifugus: sure, no problem, it's just something i noticed. [22:30:42] <leonardo> Lady_Aleena: this extension will be part http://www.gizmogarden.com/index.php/Ajax_whos_online [22:31:23] <Lady_Aleena> Oh, then that is really out of my league. [22:31:41] <leonardo> why? [22:31:43] <leonardo> hehe [22:31:53] <Emufarmers> leonardo: just because she happens to be saying something in the channel doesn't mean she has the interest or expertise to help you [22:32:34] <leonardo> thanks to intermediate [22:32:41] <solifugus> Duesentrieb: should the parser renderer function and the ajax handler be put in the body file while all the rest of this in the setup file? (to be proper) [22:32:42] <leonardo> sooy for my bad english [22:33:36] <Duesentrieb> solifugus: yes. and they should be static methods of a class, and that class should be registered to auto-load the file you need. [22:34:21] <Duesentrieb> solifugus: you can even do two classes in two files, because needing one doesn't imply needing the other. actually, they are probably never used together at all. [22:35:08] <leonardo> if some crazy person would like to help me on a city mediawiki project for free [22:35:09] <solifugus> Duesentrieb: kinda od making one class for one function.. i suppose that's for future expandability.. [22:35:17] <leonardo> i'll be crazy too [22:35:26] <leonardo> a skin for example [22:35:57] <Duesentrieb> solifugus: it's an alternative to prefixes for avoiding name conflicts. and it'S for the autoloader - i don't know if it works on function level. [22:36:11] <solifugus> i c [22:36:42] <bazz> so, i'm using the dynamicarticlelist extension. i had the problem where the list wasn't refreshing unless i edited the page it was on. i found the solution on the discussion page of putting: $wgCacheEpoch = gmdate( 'YmdHis' ); in LocalSettings, which works, but i'm wondering what that actually does and if it will have any other side effects [22:39:47] <leonardo> anyone would like to help in development of this ajax whos online extension: http://www.gizmogarden.com/index.php/Ajax_whos_online [22:45:33] <Dantman|afk> bazz: By the looks of that it's setting the cache epoch to the current time... in other word each and every time the page loads the new cache epoch is the current time and so things get cached, but they are all ignored 100% of the time, so your server is re-parsing and redoing every last thing it can do on every one of your page loads... ie: worst performance you can find [22:45:54] <siebrand> leonardo: YAWO? (Yet Anoter Who's Online) [22:46:16] <Dantman|afk> ^_^ No YAHO [22:46:16] <leonardo> siebrand? [22:46:18] <leonardo> what? [22:47:07] <leonardo> siebrand: this one has a good todo list and it's ajax [22:47:31] <bazz> Dantman|afk: that's what i thought, thanks. i'll hack up something different :) [22:47:50] <Dantman|afk> The default one used in LocalSettings.php is pretty good [22:48:18] <Dantman|afk> Basically it uses the file's time... in other words every time you change the wiki's settings the cache epoch is updated [22:49:47] <siebrand> leonardo: Firbug is giving errors on that page. [22:50:16] <leonardo> really? [22:50:25] <leonardo> http://www.gizmogarden.com/index.php/Ajax_whos_online [22:50:30] <leonardo> try again [22:51:45] <leonardo> siebrand: which errors? [22:52:04] <solifugus> Duesentrieb: I put the require_once("$IP/extensions/useful/useful.setup.php"); in LocalSettings.php (at the bottom) and it doesn't seem to be working.. {{#useful: one | two| three }} just shows up as literally.. [22:52:28] <siebrand> leonardo: hmm, not anymore. Apologies for the false report. [22:52:42] <Lady_Aleena> What is the max nesting again? To do what I want, I will have to nest 11 #ifeq statements.) [22:52:44] <Duesentrieb> solifugus: well, then, debug time :) [22:53:10] <leonardo> siebrand: what did you think about the extension? are you a developer? wants to help in development? [22:53:13] <Duesentrieb> solifugus: sprinkle echo() around, or, nicer, use wfDebug and set up a debug log file [22:53:15] <Duesentrieb> !debug [22:53:15] --mwbot-- For information on debugging (including viewing errors), see <http://www.mediawiki.org/wiki/How_to_debug>. [22:53:24] <leonardo> i help giving suggestions [22:53:36] <metanil> where is the content of page is save in database?? is it in ***_page table? [22:53:39] <siebrand> leonardo: too deep into i18n/L10n here... [22:53:42] <solifugus> Duesentrieb: ok.. thank you.. what joy. [22:53:52] <Duesentrieb> solifugus: so, see what gets called and when. also, always test on a preview page, it's never cached. [22:54:01] <Duesentrieb> seeing broken stuff from cache can get confusing [22:54:07] <siebrand> leonardo: I help run http://translatewiki.net [22:54:15] <leonardo> ha ok [22:58:19] <Duesentrieb> bwahaha, one more reason to rewrite texvc: http://www.langpop.com/ [23:11:04] <Lady_Aleena> I can't figure this out. [23:12:00] <Platonides> VB the fourth language... [23:12:54] <Lady_Aleena> I have gotten as far as {{#ifeq:{{SUBJECTSPACE}}|||do something}}, but how to I add the rest of the stuff that I want this template to do the same thing with? [23:16:31] <solifugus> Duesentrieb: somehow its trying to declare the classes in useful/useful.body.php twice.. My error: Fatal error: Cannot redeclare class usefulfrontend in [23:18:45] <alnokta2> how to remedy that ? http://alnokta.arablug.org/omkalthom/index.php?title=Test [23:20:04] <solifugus> alnokta2: that looks all backwords.. and funny writing, too.. [23:21:32] <LennySh> how can i set the default page to use index.php instead of index.php5? [23:21:34] <alnokta2> solifugus, i'm talking about the failing in creating the thumbnail [23:22:07] <alnokta2> i get this error http://pastebin.com/m10896606 [23:22:27] <solifugus> alnokta2: i know.. i was just being silly.. it's been a long day and I am tired. [23:22:53] <alnokta2> solifugus, :) [23:23:30] <LennySh> anybody? [23:25:54] <exobuzz> brion: there are some encoding issues with the doxygen interface (with < showing as <).. not sure but it might be fixed in newer versions. Possibly to upgrade to the latest ? [23:26:34] <brion> exobuzz: well i suspect it's supposed to actually support some html or some such silly things. might want to check that :) [23:28:09] <exobuzz> its difficult when the documentation is specifically telling you to use < or something but it shows as < [23:28:13] <exobuzz> like use < instead of < [23:28:15] <exobuzz> :-) [23:28:46] <Lady_Aleena> Is there an a or an switch thingy? [23:29:09] <exobuzz> brion: does $wgRequest->getText sanitize /clean the input from certain characters ? [23:29:47] <exobuzz> i would read the docs. but this bug strikes again and it says : [23:29:48] <exobuzz> Fetch a text string from the given array or return $default if it's not set. [23:29:49] <exobuzz> is stripped [23:29:54] <exobuzz> .. is stripped.. [23:29:55] <exobuzz> :) [23:29:57] <brion> all $wgRequest inputs do UTF-8 validation and unicode normalization, and perhaps strip a few chars (nulls etc?) [23:30:15] <brion> getText() also optionally does language-specific transforms (esperanto x-coding) [23:30:34] <brion> getText() should always return a string, i think? (eg not an array by surprise) [23:30:38] <exobuzz> k [23:31:15] <wikibugs> 03(mod) Redirect parsing is over-permissive; syntactically broken redirects are parsed as redirects to the wrong place. - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=15053 (10snottygobble) [23:31:25] <alnokta2> brion, would installing a custom ffmpeg fix that breakage ? [23:31:26] <exobuzz> btw is mysql_real_escape_string completely safe? or should i generally clean the string also first ? [23:32:06] <brion> alnokta2: dunno, it might [23:32:14] <brion> ffmpeg is pretty fragile is what i hear [23:32:32] <brion> exobuzz: depends what you're trying to do with it i guess ;) [23:33:17] <brion> in general in mediawiki code you shouldn't use mysql_real_escape_string yourself, though [23:33:30] <brion> you should be passing the strings into the Database class query abstraction functions [23:33:42] <brion> or if constructing your own SQL, use the database class's addQuotes() func [23:33:55] <brion> which will call the appropriate function for its backend [23:34:03] <exobuzz> brion: im talking to another database in a custom extension so not using any of your db stuff. sorry it was kinda mediawiki unrelated question [23:34:09] <brion> no prob :) [23:34:40] <exobuzz> i could probably use mediawikis db class to talk to my other database, but it would take a lot longer to learn ;-) [23:36:20] <exobuzz> btw. one part of me really wants to write "<div class="ml_nav">" instead of $wgOut->addHTML(Xml::openElement('div', array('class' => 'ml_nav')));.. is that bad| :) [23:41:24] <Simetrical> exobuzz, if you're using fixed attributes, just output the string directly. You only want Xml::openElement and so forth if you have some attributes that you want to be auto-escaped. [23:41:30] <LennySh> can i not get any help? [23:41:48] <wikibugs> 03(mod) Variable {{USERIFCODE}} (user interface language code) - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=2085 (10joystick.pl) [23:41:56] <Simetrical> exobuzz, so don't use '<div class="' . htmlspecialchars( $myclass ) . '">'. In that case use openElement(). [23:42:09] <Simetrical> Because sticking htmlspecialchars everywhere is ugly and error-prone. [23:42:18] <exobuzz> Simetrical: aah k [23:42:21] <brion> LennySh: check with your hosting provider on how to configure php5 as default [23:42:23] <exobuzz> yeh [23:43:30] <CIA-55> 03brion * r38746 10/trunk/phase3/ (RELEASE-NOTES includes/UserMailer.php): [23:43:30] <CIA-55> * (bug 15055) Talk page notifications no longer attempt to send mail when [23:43:30] <CIA-55> user's e-mail address is invalid or unconfirmed [23:43:39] <wikibugs> 03(FIXED) Tries to send e-mails to users without an e-mail address - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=15055 +comment (10brion) [23:46:54] <exobuzz> is it bad to have some "const whatever" in myClass extends SpecialPage - and also i have some private variables for the class which functions share (to avoid having huge parameter lists) [23:46:55] <exobuzz> ? [23:47:03] <Teckie> hey [23:47:49] <Teckie> i was told this is a place to ask questions [23:48:48] <LennySh> brion, php5 is enabled..but it will not go into my page when the end of the url ends in/ - i have to type /index.php5 [23:50:01] <exobuzz> Teckie: best just to ask, and see [23:50:20] <Teckie> alright here goes: [23:50:29] <Teckie> i have a wiki setup on my own webserver > and i want to use infoboxes [23:50:38] <Teckie> confusion: i've installed both ParserFunctions and ExpandTemplates, but i havent been able to get a properly formatted infobox [23:50:43] <Platonides> LennySh, change your directoryindex to index.php5 [23:51:25] <exobuzz> Teckie: for a simple infobox you just need standard built in template functionality.. experiment learning templates and you will get the hang of it [23:51:31] <Teckie> okay [23:51:59] <exobuzz> try making a really simple template with one parameter. and go from there [23:52:44] <alnokta2> brion, nope, new svn build of ffmpeg didn't fix it :(... [23:53:03] <Teckie> so infobox is essentially a subtemplate [23:53:10] <LennySh> Platonides, how do i do that, i do not see that option or anything similar under the home directory settings for this subdomain [23:53:15] <Teckie> sorry not subtemplate [23:53:26] <Teckie> its an actual template, but can be made into a template category? [23:54:37] <exobuzz> i really really hate how php treats false in regards to returning a failure from a function and having to use === false because 0 == false and '' == false [23:54:38] <exobuzz> etc [23:54:47] <Simetrical> exobuzz, we all hate it. [23:54:53] <Simetrical> PHP's weak typing is horrible. [23:55:02] <exobuzz> php coders need to die. and that stupid hebrew error i got the other day.. wtf [23:55:08] <Simetrical> exobuzz, it's not as bad as ((bool)'0') === false, though. [23:55:18] <Simetrical> PAAMAYIM_NEKUDOTAYIM or something? [23:55:19] <exobuzz> Simetrical: argh :/ [23:55:20] <exobuzz> yes [23:55:21] <Simetrical> "two colons" [23:55:35] <exobuzz> yes.. but saying two colons is more helpful [23:55:40] <exobuzz> i had to use google [23:55:50] <Teckie> can i post a link to this page, it might be more helpful [23:55:59] <Teckie> if you saw what i was trying to do [23:56:32] <exobuzz> Simetrical: even perl works so much better .. return undef on failure.. and everything just works [23:57:13] <Teckie> you have to remember why php does this though [23:57:16] <exobuzz> also if you need to use === false.. you have to add extra brackets like if ( ( $something = function() ) === false ) [23:57:22] <LennySh> ah [23:57:26] <LennySh> figured it out i think [23:57:43] <exobuzz> how unreadable is that [23:57:47] <Teckie> if it used a different style, they'd have to redo all their OOp [23:58:14] <brion> LennySh: is it going to index.php (which will possibly error out) or just sitting at the directory? [23:58:26] <brion> See if you can change the directory index file option in web server config. check w/ your hosting provider [23:58:35] <LennySh> i think i got it..had to make a .htcaccess file make it use .php files as php5 [23:58:39] <LennySh> seems to work [23:58:43] <brion> yeah [23:58:44] <brion> super :D [23:58:48] <LennySh> ty! [23:58:57] <Emufarmers> Any developers have a moment to look at some little patches I submitted? [23:59:10] <exobuzz> i just really hope that one day.. say a php 10 will come out.. and it wont be shit. [23:59:12] <exobuzz> :) [23:59:29] <LennySh> lol [23:59:50] <exobuzz> id use python but i need my { } :-) [23:59:58] <exobuzz> at least python is thread safe.

[00:11:54] 03(mod) Global group membership isn't listed on Special:ListUsers/group - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=15047 +comment (10jhsoby) [00:15:33] Hi, I can't create a new header in my sidebar via Mediawiki:sidebar... It doesn't appear at all. Can someone help me? [00:16:07] !sidebar [00:16:07] --mwbot-- To edit the navigation menu on the left, edit [[MediaWiki:Sidebar]] using its special syntax. For more details, see . [00:16:33] I read that, it doesnt work. [00:16:43] It's why I'm here. [00:18:26] proletaire, url? [00:18:37] http://www.cinema-quebecois.com [00:18:45] still working on it [00:19:48] use messages instead [00:19:57] What do you mean? [00:20:13] directly edit monobook.php? [00:20:35] instead of "Filmographie compl??te", use "filmographie-url|filmographie" and then add "Filmographie compl??te" to MediaWiki:Filmographie-url and MediaWiki:Filmographie [00:20:43] same for the other stuff [00:20:59] i'm not sure it will work, but it has a higher chance of working than what you have now [00:21:15] And what about the heading title itself? [00:22:00] use "nav-categories" and add Cat??gories to that message [00:22:17] i won't guarantee it will work though, but i think it will [00:22:21] Ok thank you I'll look for this [00:22:24] Thanks a lot [00:22:49] np :) [00:58:01] 03(mod) Skins should add their name as a class in - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=15052 +comment (10innocentkiller) [01:08:11] 03ashley * r38675 10/trunk/phase3/ (4 files in 3 dirs): (bug 15052) Skins should add their name as a class in [01:09:01] 03(FIXED) Wikitext in badaccess-group1 messages not parsed - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=14807 +comment (10innocentkiller) [01:09:24] 03(FIXED) Skins should add their name as a class in - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=15052 +comment (10wikia.jack) [01:27:46] can i update a certain page without editing it? (like making some kind of direct request or sth) [01:35:24] metanil, what do you mean by update exactly? i.e. re-render the page? [01:35:48] 03(NEW) Redirect parsing is over-permissive; syntactically broken redirects are parsed as redirects to the wrong place. - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=15053 15enhancement; normal; MediaWiki: Redirects; (snottygobble) [01:36:03] nono.. i mean editing a page .. (updating the content of a page) directly .. without actually going to the browser. [01:37:16] ah [01:37:20] Krimpet: the normal process is to click on 'edit' link, then type the whatever content we have .. then click 'save page'..... <---- could this process done directly with a single request. [01:37:50] yes, though it usually involves an external tool doing the edit for you [01:38:00] external tool?> [01:38:18] you mean some sort of third party plugin ? [01:38:21] like the Pywikipedia framework [01:38:44] you may want to check out the http://www.mediawiki.org/wiki/API:Changing_wiki_content if you've got a recent version of MW. :) [01:39:18] yes i do have recent wiki version [01:39:18] :D [01:40:05] but ???Pywikipedia seems also interesting .it says we can do all mediawiki stuff using python shell.... sweet [01:42:05] woot!!! API at mediawiki! [01:43:09] ???Krimpet: i think it doesn't work with 1.12.. it says 1.13 :( [01:43:47] ah, yeah :/ [02:04:50] :/ Do I have to kill jack... [02:08:10] 03dale * r38676 10/branches/MetavidWiki-exp/MetavidWiki/ (6 files in 6 dirs): updates for person display [02:18:27] 03dantman * r38677 10/trunk/phase3/ (4 files in 3 dirs): [02:18:27] Revert r38675: [02:18:27] This commit was clearly not thought out and poorly implemented. [02:18:27] * The Sanitizer has not been used [02:18:27] * Proper implementation of this would follow the same convention as the other classNames and have a 'skin-' prefix [02:18:29] Consistency in the code organization wasn't even kept, a bit of code was just lazily tacked onto the end of another line. [02:24:00] dan: that wasn't even applied to nostalgia/standard/cologneblue was it? [02:24:12] And it didn't prefix skin- [02:24:26] As well... it was using get_class( $this ); to find out the name of the skin [02:24:41] I doubt it even outputted a valid skin name [02:24:53] gonna reopen the bug? [02:24:54] Probably just threw 'MonoBookTemplate' as cruft into the body tag [02:25:02] Making a comment right now [02:25:27] can you add [02:25:30] > Another use for this will be when MediaWiki:Print.css and user/print.css are implemented, allowing skin-specific print rules without a separate /PrintSkinname.css for each. [02:25:38] then I don't have to log in to say it, heh [02:25:51] (just tack it in somewhere if ya can) [02:29:47] 04(REOPENED) Skins should add their name as a class in - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=15052 +comment (10dan_the_man) [02:30:57] bleh... I wish bugzilla had an 'edit comment' feature [02:31:26] or preview [02:31:26] ta [02:31:36] mhmm [02:32:21] hey, fix the bug, no food for you! [02:32:22] heh [02:32:25] "Hmm... I'll just writeup this code..." *click* "oh wait, fuck... that's not properly formatted" [02:32:47] ^_^ Actually that should be discussed [02:32:59] It's been WONTFIXed plenty of times in the past [02:33:13] And there are some strong opinions against it [02:33:17] so why is there a new bug? shouldn't this be DUP? [02:33:37] ^_^ I have no clue how to track down old bugs in bugzilla... [02:33:49] advanced search! it does works [02:33:52] Why do you think things I fix never get tagged as fixed? [02:33:56] sometimes [02:34:05] heh [02:35:01] https://bugzilla.wikimedia.org/show_bug.cgi?id=6547 preview class [02:35:10] https://bugzilla.wikimedia.org/show_bug.cgi?id=9758 history class [02:35:18] https://bugzilla.wikimedia.org/show_bug.cgi?id=13242 anon/logged in class [02:35:32] heh [02:35:50] ^_^ And here I am thinking... "Hmm, that was a pretty brionesque move" [02:36:14] https://bugzilla.wikimedia.org/show_bug.cgi?id=4438 actions [02:36:38] I never saw one for skin specific names in body [02:36:48] but maybe there are some [02:37:55] I remember it at least from a old mailing list discussion I think [02:38:05] right [02:38:17] One or two devs has strong feelings against 'body classname cruft' [02:38:25] and I've been trying to rouse discussion here about it since brion refactored the head css [02:38:41] namespace, direction, and skin name, should be good enough for anyone [02:39:02] and then of course: action, software, page name, user name, user age, parser functions used on the page... [02:39:22] whether the page contains images or not, whether the page is a stub, redirect, disambiguation, "good" article.. [02:39:23] bwahaha [02:39:36] put the whole content into the tag as parameters [02:40:42] https://bugzilla.wikimedia.org/show_bug.cgi?id=5874 hah, not even sure what this is [02:41:23] ? [02:44:02] https://bugzilla.wikimedia.org/show_bug.cgi?id=9174 haha, pageparent-Foo_Bar [02:44:34] so sure, many good arguments against adding another body class, especially with all the cruft to follow [02:45:04] <^demon> I'd say wontfix them all. It's stupid. [02:45:06] can tags have a class=" [02:45:08] ? [02:45:20] it is, but for Print.css, a skin specific one would be highly useful [02:45:31] maybe not quite required but... [02:46:04] <^demon> Too bad @media print {} doesn't apply to printable=yes. [02:46:25] indeed [02:52:18] damn, class/id/style/title "Not valid in base, head, html, meta, param, script, style, and title elements." [02:53:09] 03mfarag * r38678 10/trunk/extensions/ (4 files in 2 dirs): Add arz translation [03:01:18] 03shinjiman * r38679 10/trunk/phase3/languages/messages/ (5 files): Localisation updates Cantonese, Chinese and Old/Late Time Chinese [03:02:46] since I'm still new to committing, I'm curious - I noticed bug 9858 for the CheckUser extension's been open for a while, and I'm thinking of tackling it, but is there etiquette regarding touching "someone else's" extensions? [03:03:15] improvement is never detrimental, in my opinion anyway [03:03:40] 03mfarag * r38680 10/branches/REL1_13/extensions/Makebot/ (Makebot.alias.php Makebot.i18n.php): Add arz translation [03:03:59] 03mfarag * r38681 10/branches/REL1_13/extensions/Makesysop/ (SpecialMakesysop.alias.php SpecialMakesysop.i18n.php): Add arz translation [03:04:02] hmm, it looks like Aaron and Tim said they'd be open to anyone else working on it, so I guess that'd be OK. :) in general, though, are devs expected to stay in certain areas? [03:05:29] just want to make sure I don't run into any faux pas. :) [03:06:51] whats the worst they can do, ask you not to help? :) [03:07:08] good point. :) [03:07:09] If somebody had "taken" it, he probably would have assigned it to himself [03:07:21] That field is there for a reason, after all ;o [03:07:59] on another note, BTW, doxygen segfaults if I try to generate docs from SVN. :p [03:08:04] On a related note, what's a good way to get somebody to look at a patch once I've submitted it? [03:08:57] Emufarmers, poke a dev I guess - which bug#? [03:09:10] https://bugzilla.wikimedia.org/show_bug.cgi?id=12142 [03:10:17] I'm embarrassed how long it took me to write that patch...especially since it mostly involved copying and pasting existing code [03:10:36] hmm, I'm a bit bamboozled... shouldn't 'skipcaptcha' just disable the captcha altogether? o_O [03:10:55] For that group, yes; [03:11:14] But it doesn't, probably because ConfirmEdit was originally designed to, you know, confirm edits [03:11:15] Emufarmers . . [03:11:18] 03(mod) Add variant of MediaWiki: Pagetitle for customizing the main page - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=15007 (10N/A) [03:11:31] will_hough; don't I know you from phpBB? [03:11:36] Yes you do [03:11:42] howdy [03:11:46] Hello ;o [03:11:55] T_T not that app... [03:12:14] Although I'm surprised I remembered your name [03:13:46] 03(mod) Add variant of MediaWiki: Pagetitle for customizing the main page - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=15007 (10N/A) [03:20:20] Dantman|FS: What was that comment for? [03:20:53] oh heh... I actually took a look at the code a few days ago for once... and ended up thourougly disturbed at it... [03:21:16] at phpBB? [03:21:34] Don't be a hata [03:21:42] T_T Such an unreliable method of preventing SQL injection attacks... one minor slip up and the software is easy to crack [03:22:18] Dantman|FS: I invite you to give it your best shot [03:23:15] No... I mean, if you accidentally use the wrong type for the 'default' value of a query parameter... ie: a string when you use ints... the query won't be escaped right and is an easy injection vector [03:23:35] 03mfarag * r38682 10/branches/REL1_13/phase3/languages/messages/MessagesAr.php: remove dupe link, it was removed from en file [03:23:50] One little slip up when programming is all it takes [03:32:30] 03(NEW) Implementing "bigger" and "smaller" parameters for images will allow flexibility without overruling default thumb size preferences . - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=15054 15enhancement; normal; MediaWiki: Images; (snottygobble) [03:32:49] 03(mod) Redirect parsing is over-permissive; syntactically broken redirects are parsed as redirects to the wrong place. - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=15053 15enhancement->minor (10snottygobble) [03:41:16] 03(mod) Add variant of MediaWiki: Pagetitle for customizing the main page - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=15007 (10dan_the_man) [03:56:04] 03krimpet * r38683 10/trunk/phase3/includes/ (3 files in 2 dirs): [03:56:04] Wrapped certain tricky constructs in @cond/@endcond to prevent Doxygen [03:56:04] from choking on them and segfaulting. [03:56:19] *Marshalrusty slaps Dantman|FS with a refrigerator [03:56:29] ;o [03:57:04] *Dantman|FS steals the fridge and stores it so he has one when he moves out... [03:57:05] Is will gathering up a posse to tar and feather Dantman for daring to insult phpBB? [03:57:23] I am already boiling the oil [03:57:43] How dare thou biteth thou thumb at thy [03:58:02] *thee [03:58:11] That'll work [03:59:49] Thy challengst thou to a glove slap off [03:59:57] I'll be here when you're ready [04:00:24] I hope you're not mad at me for cheating on phpBB with other software [04:00:52] Your name shall be stricken from the phpBB book :P [04:01:29] How to Generate Generate the HTML code for the Wikitext $text inside a Extension ? [04:02:25] 03mfarag * r38684 10/branches/REL1_13/extensions/ (3 files in 3 dirs): Add arz translation [04:02:34] Aghghhghg [04:02:37] HELLO MEDIAWIKI PEOPLE [04:02:47] Extension:FramedVideo works and it works like happiness and joy! YAY!!! [04:02:47] o_O [04:03:20] Me love you long time? [04:03:24] sure! [04:03:45] I cannot believe how nice it is. [04:03:49] YAY! I am so excited [04:04:21] for extra fun, tweak it to enable autoplay. ;D [04:04:40] Hahah, I don't hate my users -that- much [04:04:47] $5 GI Joe [04:05:49] it'd make a nice block message, I think. get blocked, and be greeted by the Flea Market Rap at full blast [04:06:21] never gonna give you up... [04:06:53] or http://www.youtube.com/watch?v=U4CWl-o388A :p [04:07:10] Krimpet: this is an excellent idea [04:08:38] You could actually respond to what he said; we could use a good flamewar in here [04:08:40] T_T Such an unreliable method of preventing SQL injection attacks... one minor slip up and the software is easy to crack No... I mean, if you accidentally use the wrong type for the 'default' value of a query parameter... ie: a string when you use ints... the query won't be escaped right and is an easy injection vector [04:09:01] I would love to, but he seems to have gone back to his corner :P [04:09:38] ((Or is watching but doesn't care)) [04:09:38] Ah, why not? [04:10:11] Specifying the wrong type would be silly. It would be like casting an int to a string and then blaming php [04:11:10] There's really no excuse for incorrect using a function, and it's definitely not "easy" to do. You need to know what you're asking of the function, and then everything works [04:11:11] yknow, Datrio once integrated phpbb into mediawiki (for Wikia), it parsed wikicode and share a user database, even gave it a monobook-ish skin. [04:11:57] Ack yuck... FramedVideo... it's just another crappy video embedding implementation... :/ 38 hosts? Why not do it right and make it easy to add new ones [04:12:01] Well, that's my part. I await rebuttals [04:12:42] Splarka: Is there a punch [04:12:43] Splarka: I gave mediawiki a phpBB skin :) [04:12:44] line? [04:12:45] the Dantocratic delegate, your response? you have 2 minutes to answer [04:13:27] Emufarmer: just the suggestion that we can "all just get along", heh [04:13:34] Oh [04:14:09] To be honest, none of the attempts at integration I've seen have ever had much use; I don't know if it's a matter of quality or visibility [04:14:35] I think Wikia gave up on it from lack of use, but it was an interesting experiment [04:15:04] the two types of users, forum users, vs wiki users, are probably just too divergent [04:15:28] Datrio dissapeared, and the actual implementation had many bugs... so it's not in wide use [04:15:51] And even if the systems were integrated, talk pages are still more convenient because they're properly tied to articles [04:15:51] ^_^ Splarka, not if you get into the anime wiki [04:16:04] http://rankings.big-boards.com/?sort=week phpbb does run the busiest and spammiest message board on the internets [04:16:21] (and it is an anime board, heh) [04:25:14] sherrod, here's a patch to do that. :) http://rafb.net/p/vBA9VD66.html [04:25:57] if a YouTube video is embedded in a user's block message (or in a blocking template), when they get the block message, it autoplays. >_> [04:26:34] That's nasty [04:26:41] "Nuh-uh-uh!" [04:26:41] see, this is why I don't have flash installed [04:27:43] I've made more diabolical extensions in the past, for a friend's private wiki ;) [04:28:16] Special:FixRfA? [04:28:24] I always wanted to have a "pocket universe" extension [04:28:50] you fork-ban someone, for vandalism or such, and all their contributions are silently undone [04:29:07] but a temporary database is created of their changes, that only they see, and continue to see as they make more changes [04:29:48] so they just keep happily vandalizing, thinking they're able to blank the whole wiki, and after ~24hr the temp data is quietly deleted [04:30:26] and the "modified" data and vandalish recent changes is only sent to their session [04:30:50] I would settle for the option to undo all a vandal's edits when blocking [04:30:52] Splarka: That's how we deal with serial problem users at ACC. Block them by a combination of useragent, regex'ed request, and, IP, and let them quietly think they're getting giggles. [04:33:03] 03(mod) {{int:X}} respects user-defined interface language, breaking link tables etc. (aka {{USERIFCODE}} strikes back) - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=14404 (10brad9626) [04:35:10] 03mfarag * r38685 10/trunk/phase3/languages/messages/MessagesEn_gb.php: Add specialPageAliases [04:36:31] SXT40: what is ACC? [04:37:06] sherrod: it's the system we use at enwiki (on the toolserver) for handling requests for assistance in creating accounts (usually due to the captcha being to hard, antispoof, or etc) [04:37:21] *too hard [04:37:34] oh. [04:37:38] enwiki.net? [04:37:46] en.wikipedia.org , sorry [04:38:04] ahh [04:38:10] oh wow you work on wikipedia? [04:38:28] millions of people work on wikipedia [04:38:29] I edit there occasionally :P [04:38:33] tsh [04:38:35] thousands do [04:38:40] millions are Grawp clones [04:38:45] Splarka: heh [04:39:00] flyingparchment: probably not a whole million yet, I forget what the exact statistic was for non-blocked accounts with > 1 edit [04:39:14] probably 3 [04:39:23] IIRC it was something surprisingly low for 6M registered accoutns [04:39:26] I am trying to move a wiki from one host to another following the docs "Manual:Backing up a wiki" and "Manual:Moving a wiki". Everything seems to have got moved over - But only a part of the pages are displayed. What to check? [04:39:27] *accoutns [04:39:36] what are there, like 8 "active" users on commons? [04:39:38] SXT40: why only consider accounts? [04:39:40] sherrod: /almost/ everyone here edits wikipedia [04:39:45] i have only edited wikipedia about 3-5ish times [04:40:03] IPs aren't people, they are subhumans [04:40:09] spelling errors - I'm one of "those" people [04:40:11] sherrod: that's probably about on par with the number of '''useful''' edits I've made :P [04:40:17] they need a transparent star on their shirt sleeves [04:40:34] sherrod: thats the type of reader we really want [04:40:43] http://s23.org/wikistats/largest_html.php - does anyone know much about this list? [04:40:45] Splarka: a target barnstar [04:41:14] like, is it subject to the MW blacklist? [04:41:16] "barnstar of david" (if one wants to be offensive, heh) [04:41:31] what is wiki.cn...? [04:41:52] ahh, the great firewall of china [04:45:49] sherrod, I don't believe so [04:46:03] it seems to rank wikis based on "good," whatever that means... [04:46:15] 03(FIXED) Change time zone in Korean Wikisource - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=14849 +comment (10jeluf) [04:46:30] more come up if you set it to a lower threshold, http://s23.org/wikistats/mediawikis_html.php?sort=good_desc,total_desc&th=1000&lines=1527 [04:49:20] if (wikihasuserProdego() == 1) { $good = $good * 1000} else { die (not enough Prodego)} [04:57:31] Krimpet: oh, cool, thank you. [04:59:43] I am trying to move a wiki from one host to another following the docs "Manual:Backing up a wiki" and "Manual:Moving a wiki". Everything seems to have got moved over - But only a part of the pages are displayed. What to check? [04:59:45] Krimpet: "Good articles" [05:00:01] When you say, "a part", what part? [05:16:27] 03(FIXED) Set the logo on Hungarian Wikinews - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=14753 +comment (10jeluf) [05:17:36] 03(mod) Add variant of MediaWiki: Pagetitle for customizing the main page - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=15007 (10N/A) [05:21:53] 03(FIXED) New import sources for it.wikinews - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=14844 +comment (10jeluf) [05:22:38] http://i194.photobucket.com/albums/z293/eyehaverisen/rodney.jpg [05:25:10] 03krimpet * r38686 10/trunk/phase3/ (includes/User.php maintenance/Doxyfile): Fix Doxyfile aliases, now that doxygen has wisen fwom its gwave [05:26:26] "be vewwy vewwy qwiet, Im hunting comments, huh huh huh huh huh"? [05:26:57] [[w:Altered Beast|WISE FWOM YOUR GWAVE.]] [05:27:52] pow-er-up... [05:28:04] the voice synth wasn't *that* bad ^_^ [05:42:16] 03(FIXED) "Portal" namespace for Nynorsk Wikipedia - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=14846 +comment (10jeluf) [05:42:57] *GreenReaper has a better portal: http://en.wikipedia.org/wiki/Portal:Portal [05:43:47] Admittedly, it is not in other languages. [05:44:09] 03(NEW) Tries to send e-mails to users without an e-mail address - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=15055 normal; normal; MediaWiki: Email; (lalinsky) [05:44:31] But who uses Nynorsk, anyway? Bokm??l for the win! [05:57:33] 03tstarling * r38687 10/trunk/extensions/OggHandler/ (4 files): (log message trimmed) [05:57:33] Fixes for all the browsers that broke themselves since I wrote this extension: [05:57:33] * IE7: Use fully qualified .jar URL to avoid total failure to load applet [05:57:33] * Opera 9.5: Don't use iframes for Java, it's now broken. Revamped the whole detection system to deal with its terrible navigator.mimeTypes behaviour. [05:57:33] * Konqueror: Blacklisted the totally broken VLC and Java plugins. Added support for Kaffeine and KMPlayer which kind of work. Also benefits from new mimeTypes stuff. [05:57:38] Also: [05:57:40] * Added explicit player selection for Totem [06:02:59] *Splarka stabs his ISP for forcing opendns on him [06:03:08] *Splarka adds guide.opendns.com to his hosts file [06:15:26] Splarka: I tried to get one of them but the tld I had in mind to use had reverted to central ownership or something. [06:18:45] hmm? [06:20:05] i've set $wgLogo = 'skins/common/images/imagename.png'; and the resulting main page gives a style="background-image: url(skins/common/images/imagename.png);" tag, yet I see no image. The image is 125x125 (less than 135 max) [06:20:57] am i missing something? [06:21:07] Jinx, is this site available publically? [06:21:39] Jinx: is skins/ relative? you may wish to use an absolute path [06:21:39] (if so I can fire up Firebug and take a look-see) [06:21:43] i can access public_html/wiki/skins/common/images/imagename.png yes [06:21:52] eg /public_html/wiki/skins/common/images/imagename.png [06:21:54] ok, i'll give you the real url then :) [06:21:57] ah, Splarka has a good point [06:22:12] it says it supports a relative path [06:22:14] if you use short URLs, the /wiki/ will confuse browsers [06:22:15] i tried public_html... [06:22:20] but not /public_html [06:22:40] well, a server-absolute path, not a web-absolute path [06:22:47] which I guess would also be relative [06:22:51] but relativity is relative [06:22:56] $wgLogo = "$wgScriptPath/skins/common/images/imagename.png"; [06:22:57] *Jinx nods. [06:22:58] try that :) [06:23:10] oh can i put variables directly in the string? [06:23:12] (note, double quotes) [06:23:18] ah, crazy php [06:23:20] yes, only with double quotes though [06:23:26] yahm, heh, use the skin path [06:23:53] $wgStylePath ? [06:24:37] yeah, either would work. :) [06:25:01] no go [06:25:06] i wonder if my image name is bad? [06:25:21] does it have anything other than alphanumerics in it? [06:25:21] just has alphanumeric, _ and - [06:25:28] is _ and - no good? [06:25:35] nah, those should be fine... [06:25:49] is your site live? [06:25:52] yep [06:26:09] oh, you don't want the server side relative path [06:26:17] you want the web-side relative path, which is not /public_html [06:26:20] Splarka, yeah, client-side [06:26:29] $wgScriptPath should be, though [06:26:38] $wgLogo = '/skins/common/images/imagename.png'; [06:26:45] scriptpath is not always stylepath [06:26:58] Jinx, the ()'s are what's killing it [06:27:06] nah, i tried renaming it :) [06:27:11] ohh [06:27:19] sorry for the mis-copy [06:27:50] I'm loading it, though, and it's still loading from skins/ absolutely; did you set $wgLogo to "$wgScriptPath/skins/common/images/imagename.png" ? [06:28:02] yep [06:28:15] oh wait a second here [06:28:16] ... [06:28:28] it's acting as though you didn't :/ [06:29:01] i didn't :( [06:29:05] "{$wgStylePath}/common/images/wiki.png" [06:29:15] and there ya go ;) [06:29:17] is the default for $wgLogo [06:29:21] loaded localsettings.php into the darn image folder [06:29:26] because i was renaming the images like mad [06:29:32] sheesh, heh [06:29:45] $wgScriptPath are for scripts, index.php, api.php, thumb.php, etc [06:29:51] (only did that this once, to be fair... so this wasn't all just due to that :) ) [06:29:56] $wgStylePath are for skins css/js/images [06:30:16] shouldn't $wgStylePath."/common/images/imagename.png"; work? [06:30:21] i tried that first [06:30:29] since that was in the default [06:30:34] have you linked to your site yet? just do that and it'll be easier ^_^ [06:30:53] Splarka, he did to me, and it's fixed : [06:30:56] :)* [06:31:00] :) [06:31:07] didn't want to spam links publicly, sorry [06:31:16] you two get a room ^_^ [06:31:28] anyway, Jinx [06:31:39] i'm just curious though... shouldn't "$wgStylePath/common" be the same as $wgStylePath."/common" ? [06:31:42] what many do, is link to the path of an uploaded image [06:31:50] like /images/b/bc/Wiki.png [06:32:02] then protect it? [06:32:28] yah [06:32:43] then you can upload new images for april fools jokes and other stupid things ^_^ [06:32:43] not that i expect significant abuse, but i'd rather not have some kid upload a nsfw image over our logo :) [06:32:55] true [06:33:56] interesting [06:34:02] $wgStylePath must not be defined [06:34:04] for my wiki [06:34:19] *GreenSleeper chuckles. [06:34:22] because $wgScriptPath/common works, but $wgStylePath/common yields blankness again [06:34:52] why on earth wouldn't it work just doing "skins/common" or "public_html/wiki/skins/common" ? [06:35:06] so define it? [06:35:15] $wgStylePath should, in almost all setups, be equal to "$wgScriptPath/skins" [06:35:18] i inherited this wiki [06:35:23] i don't claim to be an expert :) [06:35:27] wtf [06:35:32] Krimpet: not everyone uses null scriptpath! [06:35:47] wikimedia for example, /w/ and /skins-1.5/ [06:36:01] Splarka, anyone that does, at least knows what they've changed. :) [06:36:07] var stylepath = "/skins-1.5"; [06:36:08] var wgScriptPath = "/w"; [06:36:10] wlel [06:36:29] should i set $wgStylePath = "skins" ? [06:36:34] or /skins ? or what? [06:36:37] also, wikimedia has to have them dynamic and not absolute, because they have multiple URI schema for access (secure.wikimedia.org) [06:37:00] Jinx: whatever you need for it to work, is yourserver.org/skins/ the proper stylepath? [06:37:18] Jinx, just leave it the way it worked, and it should be fine. :) [06:37:20] and is your scriptpath null? [06:37:25] it starts from my server root? no then. server.org/public_html/wiki/skins [06:37:27] 03tstarling * r38688 10/trunk/extensions/OggHandler/OggPlayer.js: Debugging code accidentally left in [06:37:49] heh [06:37:52] Krimpet: true... but what if i change scriptpath? :) [06:37:55] "if it ain't broke, don't break it more" [06:37:57] not that i really pan to [06:38:02] plan even [06:38:05] Splarka, the MW installer automagically sets $wgScriptPath [06:38:14] Jinx, I doubt you will. :) [06:38:35] yes yes, but if you're changing the stylepath you might as well not have it rely on a possible breakage of scriptpath [06:38:36] so $wgStylePath = "public_html/wiki/skins" ? [06:38:59] or do i need a / suffix again? [06:39:00] Jinx: that would break with short URLs if you plan to use them [06:39:10] short urls? [06:39:17] /wiki/ for example [06:39:23] it would try to load /wiki/public_html/wiki/skins [06:39:32] ah [06:39:37] that's no good then [06:39:51] should be /public_html/wiki/skins or such, if it works that is [06:39:54]