[00:08:17] brion: so are bug wednesdays on or not? [00:09:54] the new schedule is "any day ending in Y is a bug-____day" [00:10:00] welp it's a holiday [00:10:07] so that's kinda up to y'all this week :D [00:14:24] 03(mod) CSS tweak for Safari - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=1941 (10michael) [00:18:20] 03(mod) Other Input for arraymaps - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=16780 (10yaron57) [00:21:34] hrm.... PdfHandler totally not working for me. :( debug time! [00:22:21] 03aaron * r45010 10/trunk/extensions/FlaggedRevs/FlaggedArticle.php: (bug 16734) '#ca-stable and #ca-current tab shown on rollback incorrectly' [00:22:40] bah [00:22:46] 03(FIXED) #ca-stable and #ca-current tab shown on rollback destination page, although there wasn't any unsighted revision - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=16734 +comment (10JSchulz_4587) [00:22:46] just didn't set the damn settings [00:25:54] 03(FIXED) exclamation mark on RC - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=16724 +comment (10JSchulz_4587) [00:29:58] 03brion * r45011 10/trunk/extensions/PdfHandler/ (PdfHandler.php PdfHandler_body.php): (log message trimmed) [00:29:58] Some cleanup... [00:29:58] * remove no longer accurate code comment [00:29:58] * set default values for the processor commands, otherwise it's totally confusing that not only doesn't it work, but it looks like it's just not installed [00:30:01] * added a comment about adding to $wgFileExtensions -- not doing it [00:30:03] automatically since some wikis might want to *use* PDFs from a shared server [00:30:05] without locally uploading them.... mebbe... dunno... keeping this one open for [00:31:30] *AaronSchulz floats https://bugzilla.wikimedia.org/show_bug.cgi?id=16640 to brion [00:32:53] *brion peeks [00:33:04] let's try it [00:33:40] *brion pokes off [00:34:56] *brion tests [00:35:15] 03aaron * r45012 10/trunk/extensions/CodeReview/Subversion.php: Set locale in case of UTF-8 in commit message [00:36:08] seems to help yes [00:36:26] you can run svnImport MediaWiki 0 to fix the oldies [00:36:35] that'll redo the comment? good [00:37:07] and it will leave the status :) [00:37:57] :D [00:38:04] 03brion * r45013 10/trunk/tools/codereview-proxy/index.php: add LC_ALL=en_US.utf-8 on svn calls for proper output encoding [00:38:48] *brion tests it offline [00:39:50] "checks, ?\195?\160 la LinkBatch," [00:39:52] hmmmmm [00:40:04] nope didn't do it :( [00:44:20] *brion slaps self [00:44:25] it helps if i install the updated code i bet [00:46:19] i'm still seeing the bad output [00:49:50] dude it looks fine in the terminal but it's spitting the other crap back at me live :( [00:51:16] what do you mean? [00:53:10] LC_ALL=en_US.utf-8 svn log -v -r44559:44560 --non-interactive http://svn.wikimedia.org/svnroot/mediawiki/ [00:53:11] ^ good [00:53:14] http://codereview-proxy.wikimedia.org/?action=log&path=&start=44559&end=44560 [00:53:15] ^ bad [00:53:31] the latter runs the former from php [00:55:54] lemme try LC_CTYPE [00:56:49] no change [00:58:11] capitalizing UTF doesn't change either [01:00:04] http://codereview-proxy.wikimedia.org/?action=log&path=&start=45001&end=45001 [01:00:37] http://www.mediawiki.org/wiki/Special:Code/MediaWiki/45001 [01:00:41] ugh [01:01:28] brion: what is the php locale? [01:02:12] attempting to check. moment [01:02:20] http://www.php.net/setlocale [01:03:44] hmmmmmmm [01:03:50] ok something's not working right here [01:05:03] URRGGHHHH [01:05:12] i had a /etc/hosts entry hardcoded :) [01:06:11] ok, it is in fact working. yay! [01:06:26] *sigh* [01:09:06] AaronSchulz: re-running import :D [01:09:09] thx! [01:09:55] 03(FIXED) UTF-8 in commit message confuses CodeReview - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=16640 +comment (10brion) [01:10:13] (locale was "C" btw) [01:10:21] (before being fixed :D) [01:10:47] heh [03:05:55] ho ho ho [03:07:30] 03(NEW) Restore "whatlinkshere" and "related changes" links in sidebar for classic skin - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=16786 minor; normal; MediaWiki: User interface; (paul.copperman) [03:09:18] is there a way to disable templates wiki wide [04:31:58] 03aaron * r45014 10/trunk/phase3/maintenance/renameDbPrefix.php: Add a little script I use to change db prefix [04:41:00] Can I override $wgUseCommaCount in a expr that can't interpret commas? [06:38:54] 03(NEW) Wikipedia page does not render.. rather IE7 attempts to save the page as a file - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=16787 normal; normal; MediaWiki: Page rendering; (dchapiesky) [06:41:34] 03(mod) Wikipedia page does not render.. rather IE7 attempts to save the page as a file - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=16787 +comment (10dchapiesky) [06:51:45] 03(mod) Wikipedia page does not render.. rather IE7 attempts to save the page as a file - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=16787 +comment (10overlordq) [08:35:21] werdnum: are you interested in https://bugzilla.wikimedia.org/show_bug.cgi?id=16721 ? [08:56:41] 03aaron * r45015 10/trunk/extensions/FlaggedRevs/specialpages/UnreviewedPages_body.php: [08:56:41] *querycache mode tweaks [08:56:41] **Make new redirect filter work [08:56:41] **Show page size [08:56:41] *Broke some long lines [08:56:41] *Removed duplicate global declaration [09:19:45] 03dale * r45016 10/trunk/extensions/MetavidWiki/skins/ (2 files in 2 dirs): domain detection for content providers to determine insert mode [09:46:11] 03siebrand * r45017 10/trunk/extensions/NewUserNotif/ (NewUserNotif.class.php NewUserNotif.i18n.php): Allow to use date and time separately in 'newusernotifbody' [10:03:35] 03dale * r45018 10/trunk/extensions/MetavidWiki/skins/mv_embed/libRemoteMediaSearch/mv_remote_media_search.js: fixes to support multiple "local" domains for a given content provider: ie commons is local for wikipeida, wikibooks, wikimedia etc [10:20:27] sneak peek of (very early version) of an add media wizard: importScriptURI('http://mvbox2.cse.ucsc.edu/w/extensions/MetavidWiki/skins/external_media_wizard.js'); [10:31:00] 03siebrand * r45019 10/trunk/phase3/languages/messages/ (13 files): Localisation updates for core messages from Betawiki (2008-12-25 11:22 CET) [10:37:48] 03siebrand * r45020 10/trunk/extensions/ (130 files in 124 dirs): Localisation updates for extension messages from Betawiki (2008-12-25 11:22 CET) [12:12:04] 03danny_b * r45021 10/trunk/phase3/includes/Linker.php: * reverting r44896 - it causes massive invalidity of pages. See http://www.w3.org/TR/html401/types.html#type-name and http://www.w3.org/TR/xhtml1/#C_8 for definitions of "id" and "name" content. [12:21:40] 03danny_b * r45022 10/trunk/phase3/maintenance/parserTests.txt: * reverting r44927 because of r45021 [12:23:19] 03(mod) Unicode combining characters are difficult to edit in some browsers - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=16697 (10Gerard.meijssen) [12:41:12] 03danny_b * r45023 10/trunk/phase3/RELEASE-NOTES: * reverting r44896 because of r45021 [12:59:58] Danny_B: can you be more specific? [13:01:27] in what? [13:02:09] what exactly is causing invalidity? id's that don't begin with a letter? [13:02:15] http://lists.wikimedia.org/pipermail/wikitech-l/2008-December/040728.html [13:03:02] yeah but you don't clearly mention the answer to that question there [13:03:13] s/clearly/in a clear way/ [13:03:24] to which? [13:03:39] "what exactly is causing the invalidity" [13:03:55] ids not beginning with letter [13:04:14] thank you [13:04:30] it's in [1] and [2] [13:06:25] Danny_B: well yes and no [13:06:29] only if you know what to look for [13:06:50] yes and no what? [13:41:09] 03ialex * r45024 10/trunk/phase3/ (2 files in 2 dirs): PHP 5.3 really doesn't like Namespace:: :) [13:47:39] hi [13:47:43] how to enable [13:47:49] URL friendly links>? [13:49:21] S.O.S. [13:49:28] !shorturl [13:49:28] --mwbot-- To create simple URLs (such as the /wiki/PAGENAME style URLs on Wikimedia sites), follow the instructions at . There are instructions for most different webserver setups. If you have problems getting the rewrite rules to work, see !rewriteproblem [13:49:41] im using... wiki.domain.com [13:49:58] ialex: can I just replace all occurances of Namespace:: with MWNamespace:: ? [13:50:16] siebrand: yes [13:50:35] ialex: are there any incompatibilities with earlier MediaWiki versions? [13:50:48] siebrand: 1.13+ iirc [13:50:55] ok. [13:51:42] introduced in r32309, so it's 1.13 [13:54:18] happy [13:54:28] Merry Xmas! [13:59:14] 03siebrand * r45025 10/trunk/extensions/ (12 files in 12 dirs): [13:59:14] Some PHP 5.3 compatibility fixes for extensions [13:59:14] * replaced Namespace:: by MWNamespace:: [13:59:14] These changes are incompatible with MediaWiki versions below 1.13 r32309 [14:49:01] 03shinjiman * r45026 10/trunk/phase3/languages/messages/ (5 files): Localisation updates Cantonese, Chinese and Old/Late Time Chinese [14:54:31] body { font-family: serif; font-size: 30%; } [14:54:50] why this doesn't change anything? [14:54:57] http://www.wikilivres.info/wiki/User:Yann/monobook.css [14:55:52] hi, how long does it normally take for mediawiki to upload [14:56:29] 03(NEW) HTML codes visible on Meta - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=16788 trivial; normal; Wikimedia: General/Unknown; (nurtsch-cv60) [14:57:05] yannf: cache? or there is more specific css rule overriding it? [14:58:35] Nikerabbit, i don't have any other css rule [14:58:58] shift + reload should clear the cache, isn't? [14:59:31] it might not be enouch [14:59:42] the FAQ has probably something about it [14:59:57] 14(INVALID) HTML codes visible on Meta - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=16788 +comment (10niklas.laxstrom) [15:01:52] Nikerabbit, then editing a page should clear the cache of that page? [15:02:21] http://www.wikilivres.info/wiki/MediaWiki:Monobook.css [15:04:03] http://meta.wikimedia.org/wiki/MediaWiki_FAQ#I_tried_that_but_it_didn.27t_work [15:06:38] I emptied the cache of the browser (FF), but it doesn't change anything [15:08:47] don't you have web developer toolbar extension for firefox? You can use it to edit css to test if the rule is correct [15:09:39] no toolbar extension now [15:13:02] I use the same rule on Wikisource, and it works [15:13:29] alos I tried with another browser (Konq) and it is the same [15:15:37] so it works then [15:16:39] no, I mean, it doesn't change anything in Konq too [15:17:41] well, then there is some problem, obviously :) [15:17:42] with font-size: 30%; the font should be very small [15:18:00] do you even have a firebug? [15:18:06] the error console might give some clues [15:22:17] 03ialex * r45027 10/trunk/phase3/ (4 files in 3 dirs): (log message trimmed) [15:22:17] * summary and subject messages now uses wiki text rather than raw HTML [15:22:17] * Added the colon in summary, subject and subject-preview messages rather than using colon-separator [15:22:17] * Changed the note in RELEASE-NOTES to also point to bug 212 [15:22:17] * Also escaped other messages that weren't escaped [15:22:18] * Added new $attribs option to Xml::label(), using it in Xml::inputLabel(), Xml::inputLabelSep(), Xml::checkLabel() and Xml::radioLabel() [15:22:23] * $fname -> __METHOD__ [15:25:53] no error related to that in the error console [15:30:10] yannf: is user CSS/JS enabled on wikilivres.info ? [15:32:07] ah, that might the pb [15:32:20] default is disabled [15:32:23] i didn't know it needs to be enabled [15:33:19] what's the var to add? [15:33:46] *Simetrical points Danny_B to wikitech-l [15:34:09] $wgAllowUserCss = true; [15:36:18] yes: that's it! [15:37:08] omg [15:37:28] why people don't read documentations and standards? :-( [15:38:23] 03ialex * r45028 10/trunk/phase3/ (includes/EditPage.php languages/messages/MessagesEn.php): oops, forgot summary-preview :( [15:38:59] Danny_B, ? [15:40:21] Simetrical: sections beginning with numbers will create invalid id, sections beginning with non-ascii char will be converted to .XY.XY style thus begin with period thus invalid; sections beginning with basic punctation will have invalid id because of beginning with that punctation etc etc [15:40:42] id MUST begin with [a-zA-Z] and nothing else [15:41:02] name can begin with whatever char is allowed to be in name [15:41:12] see NMTOKEN definition [15:42:33] So the prose is lying? [15:42:58] Laaknor, Nikerabbit thanks a lot [15:43:09] *Danny_B is reaching for dictionary,mmt [15:43:33] Danny_B, the prose bans names that aren't valid id's. Not all NMTOKENs are acceptable. [15:43:56] whatis proce? [15:43:59] prose [15:44:15] The actual text of the standard, as opposed to things like BNF and DTDs. [15:44:49] which text? [15:44:54] The part I quoted. [15:45:14] do you mean [1]? [15:45:14] "This attribute is constrained such that it can only have the same values as type ID, or as the Name production in XML 1.0 Section 2.3, production 5." [15:45:16] yannf: no problem ;) [15:45:42] The latter refers to: [5] Name ::= NameStartChar (NameChar)* [15:45:57] Names in XHTML 1 *must* begin with a NameStartChar, but this constraint is not part of the DTD. [15:46:01] Er, wait. [15:46:09] No, this is confusing me. [15:46:20] Things seem to contradict each other. [15:46:44] Yes, right, that's the same product as IDs must match. [15:46:45] Simetrical: nope, it's for backwards compatibility for transfers html<->xhtml [15:46:51] What is? [15:46:58] C.8 [15:47:12] Hmm, you're right, that appendix is informative. [15:47:18] When defining fragment identifiers to be backward-compatible, only strings matching the pattern [A-Za-z][A-Za-z0-9:_.-]* should be used. [15:47:29] Okay, now I'm really confused. [15:47:41] *Danny_B is always right about web standards ;-) [15:48:14] Danny_B, well, it doesn't matter, we're in violation of the standard either way. Look at 4.10: "In order to ensure that XHTML 1.0 documents are well-structured XML documents, XHTML 1.0 documents MUST use the id attribute when defining fragment identifiers on the elements listed above." [15:48:30] So is already invalid XHTML. [15:48:37] Even if the W3C's validator doesn't complain. [15:48:54] Simetrical: no it is not INVALID [15:49:07] it complies to DTD [15:49:16] Danny_B, it's non-conformant, then. [15:49:35] Danny_B, okay, now wait. Why is it that id's must start with a letter in XHTML 1? [15:49:50] [1] [15:49:56] as well as [2] [15:50:07] id is of type id [15:50:19] Danny_B, [1] applies to both id *and* name attributes. [15:50:25] which in [XML] is defined as the regexp i'vequoted [15:50:39] Simetrical: yes, that was in HTML4 [15:50:45] [2] is in an informative appendix, it's not normative. [15:50:49] XHTML made name less restrictive [15:51:11] Danny_B, where does it say in XML that IDs must start with a letter? http://www.w3.org/TR/REC-xml/#id [15:51:21] [5] Name ::= NameStartChar (NameChar)* [15:51:25] [4] NameStartChar ::= ":" | [A-Z] | "_" | [a-z] | [#xC0-#xD6] | [#xD8-#xF6] | [#xF8-#x2FF] | [#x370-#x37D] | [#x37F-#x1FFF] | [#x200C-#x200D] | [#x2070-#x218F] | [#x2C00-#x2FEF] | [#x3001-#xD7FF] | [#xF900-#xFDCF] | [#xFDF0-#xFFFD] | [#x10000-#xEFFFF] [15:51:39] Simetrical: see first three errors (ignore the latter) http://validator.w3.org/check?verbose=1&ss=1&uri=http%3A%2F%2Ftest.wikipedia.org%2Fwiki%2FUser%3ADanny_B.%2FBugs%2Fid_vs._name [15:51:47] The W3C Validator is not normative. [15:51:58] But if it is, it says that names are subject to the same restriction. [15:52:12] "For example, id and name attributes must begin with a letter, not a digit. " [15:52:15] *AND NAME* [15:52:27] You haven't given me anything that says id must start with a letter but name doesn't also have to. [15:52:37] Either both do or neither does. [15:52:40] I believe "both do" is correct. [15:53:39] i did [15:53:40] 04(REOPENED) HTML codes visible on Meta - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=16788 +comment (10nurtsch-cv60) [15:53:45] Danny_B, which one? [15:54:17] html 4: ID and NAME tokens must begin with a letter ([A-Za-z]) and may be followed by any number of letters, digits ([0-9]), hyphens ("-"), underscores ("_"), colons (":"), and periods ("."). [15:54:30] Danny_B, ID and NAME. Both of them must. [15:54:37] mmt, let me finish [15:54:49] moin [15:57:19] http://www.w3.org/TR/xhtml1/dtds.html#dtdentry_xhtml1-transitional.dtd_a [15:57:26] name is NMTOKEN [15:57:29] id is ID [15:57:54] 03shinjiman * r45029 10/trunk/phase3/languages/messages/ (4 files): Localisation updates Cantonese, Chinese and Old/Late Time Chinese [15:59:19] Simetrical: the quoted definition is for html4. xhtml took html and made some changes. one of them was thatname content has been changed to nmtoken [16:00:42] Danny_B, so where in XHTML1 does it say that *id*s must start with a letter? [16:01:10] XML doesn't require that. HTML4 requires it for id's but also for names. [16:01:26] The W3C validator says it's required for both id's and names. [16:03:05] 14(INVALID) HTML codes visible on Meta - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=16788 +comment (10alex.emsenhuber) [16:04:35] how do I see what are the namespaces already defined? [16:04:59] Simetrical: xhtml is xml reformulation of html. thus uses the definitions from html unless they're overriden [16:05:03] yannf, go to any page with a namespace dropdown, like Special:AllPages. [16:05:15] Danny_B, says who? [16:05:26] w3c [16:05:33] Danny_B, give me a normative source. [16:05:47] I mean, the W3C also says (in its validator) that names must begin with a letter. [16:05:49] give me normative source for your change ;-) [16:05:59] and the namespace numbers? [16:06:09] yannf, use the API if you want, or look at the HTML source. [16:06:29] Simetrical: where validator says names must begin with letter? [16:06:37] Danny_B, your own link. [16:06:40] I quoted it about four times. [16:06:43] where? [16:06:54] "For example, id and name attributes must begin with a letter, not a digit. " [16:07:40] thanks for founding a bug, i'lll report it. however as you can see id is marked as error but name is not [16:08:00] i mean finding [16:08:35] run your output through tidy [16:08:49] Danny_B, Tidy ADDS THE ID ATTRIBUTES THAT YOU REMOVED. [16:09:17] it never did [16:09:18] Danny_B, you're claiming that XHTML overrides the HTML4 restrictions on the name attribute, but not on the id attribute. What's your basis for this? The DTD gives name as type NMTOKEN, but it also gives id as type ID, which can start with non-letters. [16:09:54] id can not start with non-letter [16:10:08] Yes it can. Look at the blasted productions in the XML standard that I've already quoted several times. [16:10:21] i saw it [16:10:34] [5] Name ::= NameStartChar (NameChar)* [16:10:38] [4] NameStartChar ::= ":" | [A-Z] | "_" | [a-z] | [#xC0-#xD6] | [#xD8-#xF6] | [#xF8-#x2FF] | [#x370-#x37D] | [#x37F-#x1FFF] | [#x200C-#x200D] | [#x2070-#x218F] | [#x2C00-#x2FEF] | [#x3001-#xD7FF] | [#xF900-#xFDCF] | [#xFDF0-#xFFFD] | [#x10000-#xEFFFF] [16:10:44] "Values of type ID MUST match the Name production." [16:11:00] That allows XML IDs to start with characters like : and _. [16:11:08] hello who has called? [16:12:30] however look at "Note that the collection of legal values in XML 1.0 Section 2.3, production 5 is much larger than that permitted to be used in the ID and NAME types defined in HTML 4. " [16:12:31] nobody, 17:42:58 « yannf» Laaknor, Nikerabbit thanks a lot [16:12:44] Danny_B, ID and NAME. Both. [16:12:45] Together. [16:12:46] Same. [16:12:47] No difference. [16:12:59] __as defined in html 4__ [16:13:07] Forget it, I'm done with this argument. [16:13:24] while xhtml has override name with nmtoken [16:13:26] You cite HTML4 when it suits you but refuse to listen to it otherwise. [16:14:05] XHTML didn't override NAME with NMTOKEN for name anymore than it overrode NAME with ID for id. [16:14:09] Any more or any less. [16:14:34] It didn't override either. DTDs can't express the desired restrictions, so it picked the best fit and noted it in an informative section on compatibility. [16:15:02] bah,i can say the same - you cite xml when it suits you but refuse to listen to it otherwise ;-) validator marks ids starting with non-letters as errors while names not [16:15:09] oh [16:15:16] I let you two to figure this mess out :) [16:15:31] But the validator *says* that it's an error for both. [16:15:35] So which is the bug? [16:16:09] You can't cite anything other than your own belief that some provisions of XHTML1 override some of HTML4 and some don't. [16:16:16] You've given no rational reason why the DTD should overrule HTML4 for name but not for id. [16:20:51] validator's message under the error is informative and the current text is wrong [16:20:58] as i said [16:21:09] The validator is purely informative to begin with. [16:21:19] if name beginning with nonletter was error it would mark it as error [16:21:30] That's not true. The validator has plenty of bugs. [16:21:35] It's *not* a normative reference. [16:21:39] anyway [16:22:15] why the hell we would have

foo

? [16:22:32] why then not remove the entire empty ? [16:23:05] if we pretend that it was like you are saying [16:23:33] We don't have that, that's totally invalid. [16:23:44] The

doesn't and can't have the same id as the . [16:23:52] Or the same id as the 's name. [16:23:58] Since id and name share the same namespace. [16:24:01] i know it cant [16:24:11] we used to have this: [16:24:13] The reason we still use is because some ancient browsers don't support id. [16:24:16] For anchors. [16:24:41] If we don't care about them, we could move the id to the

. [16:24:47] And drop the tag altogether. [16:28:42] i wonder how long and where everywhere we're going to support ancient browsers. besides we don't do that globally ;-) we don't support lot of stuff from ancient browsers. we should say this and this version of such browser is the oldest supported. older version than these may have displaying problems [16:29:09] The current policy is to not drop support for older browsers unless there's a good reason, or unless they're completely hopeless (e.g., no virtual host support). [16:35:26] you asked about the "xhtml is reformulation of html": [16:35:38] This specification defines the Second Edition of XHTML 1.0, a reformulation of HTML 4 as an XML 1.0 application,... [16:35:43] it's in abstract [16:36:05] Which is not normative, but even if it were, you would have no reason to inherit the restriction on id but not on name. [16:36:23] They're given in the same breath. Either keep both or drop both, you can't have it both ways. [16:37:55] what you say is not written anywhere in xhtml definition [16:38:56] xhtml definition clearly stands that name is now nmtoken while id (sincenot being overriden by new definition) stays of the typedefined in html4 [16:39:33] id *is* overridden in the DTD. [16:39:37] id is defined to be of type ID. [16:39:46] Which is broader than in HTML4. [16:50:43] nope it is of type ID defined in html4 [16:51:05] anyway [16:51:15] let's do it this way: [16:51:55] What does "it is of type ID defined in html4" mean? The DTD defines it as type ID, which is defined in the XML standard. [16:52:07] HTML4 defines is as type NAME in the HTML standard, which is much more restrictive. [16:52:13] let's write to w3c ask them to give us the explanation [16:52:15] But I've wasted way too long on this bickering anyway. [16:52:24] so did i [16:52:33] It will be decided by Brion, and I'm content with that. [16:52:38] Go ahead and write to www-html if you like. [16:52:43] I'd be interested to hear the response. [16:53:03] :DD [16:53:13] standards are horrible, aren't they? [16:53:30] Only if they're confusing and poorly-written. [16:58:36] We should switch to HTML5. It's vastly more sensible. It appears to get rid of the name attribute entirely, and permit anything in id that doesn't contain spaces. [16:59:25] i agree there is not purely clear definition of the content of id and name attributes [17:00:37] any html is horibly parseable. html5 is nice concept and i agree with lot of newstuff there, but it should go the xhtml (resp. xml) way [17:02:01] XHTML5 is defined by the standard, but browsers aren't likely to implement it, and you MUST NOT serve it under a text/html MIME type. [17:02:19] The fact of the matter is we don't really serve XML anyway. XML is too unforgiving, IMO, in its well-formedness requirements. [17:02:49] Browsers are the primary user-agent we care about here, and they've made it clear that they aren't going to sacrifice compatibility and reject malformed pages. [17:04:30] So writing a standard that requires that is pointless. [17:04:42] Standards should never have requirements that implementers aren't willing to accept. [17:09:38] well-formedness is the key for simple further working with data [17:10:05] not well formed documents have many exceptions to deal with thus lot of additional work [17:10:23] unnecessary and worthless work [17:10:45] It depends on what sort of work you're doing. Browsers do all the work anyway, the standards need to acknowledge that. [17:11:05] browsers don't do the work the same way [17:11:05] And since web pages are written for browsers, there's no real point in pretending that things are stricter than what browsers accept. [17:11:31] browsers are trying to GUESS [17:11:42] browsers do not follow standards [17:11:51] neither even html standrads [17:12:10] HTML5 is designed to standardize that behavior. [17:12:21] So that they guess in a well-defined, standard fashion. [17:12:35] Because the browser implementers are clear that "don't guess" is something they're not willing to do. [17:13:43] vicious circle [17:13:57] 03siebrand * r45030 10/trunk/phase3/languages/messages/ (231 files): Localisation updates for core messages from Betawiki (2008-12-25 17:54 CET) [17:14:13] <^demon> standardized guessing? fun... [17:14:29] browser programmers are programming their own behavior and don't care about standards [17:17:18] anyway,well-formed document is much easier readable and understandable than html-style doc with bunch of exceptions where you have to think about every single unclosed tag if you are supposed to close it now or later [17:17:35] closed tags are logical [17:17:44] something startshere and eds up here [17:18:12] <^demon> You know...I have an ssh client, irc client, text editor...all on Android. Now all I need is svn :) [17:21:05] (that's also eg. why xhtml2 introduced to replace
) [17:21:33] On the other hand, as soon as closing tags are mandatory, you open up more room for misnesting. [17:21:48] We have several known XHTML validity errors solely because of that. [17:26:31] what is ? [17:37:21] Nikerabbit: line [17:37:50]

foo
bar

-->

foobar

[17:38:57] looks unuseful at first sight [17:39:29] Simetrical: however misnesting is very easily findable. you need just one trivial stack for this kind of check but for exceptions in html you need much much more coding [17:40:06] Danny_B, the whole point is that HTML5 is easier to author. You need many fewer checks on the author side. Simple string concatenation is much more likely to work. [17:40:24] Danny_B, and if it's so trivial to fix, I invite you to try your hand at doing so in Parser.php. [17:40:25] Nikerabbit: it's logical - rather than saying "line breaks here" you say "line starts and ends here"such as you do for all other elements [17:41:25] Simetrical: wikitext isn't html nor xhtml thus it's not so comparable [17:44:40] but it's very illogilan for runs of texts should wrap automatically [17:45:52] html5 may be easier to author but moreimportant is to have the web language easy to parse which will (or at least should) lead to thepoint where all browsers will display the similarpage in similar way and not like now [17:47:46] imageine math - you also pair parenthesis (is that right spelling?) and not only open them [17:48:39] (5 + (4 * 2 - (5 + 7 - would you know where to close them? [17:49:12] in math there is many other ways to mark expressions [17:55:14] 03(NEW) Image not shown in stable versions after move to Commons - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=16789 minor; normal; MediaWiki extensions: FlaggedRevs; (x00000000) [18:00:40] 03ialex * r45031 10/trunk/phase3/ (6 files in 4 dirs): [18:00:40] Per Nikerabbit and Siebrand: removed $skinNames from localisation files and replaced them with messages. Messages names are "skinname-" where is the internal name of the skin. [18:00:40] Updates for other languages should come soon :) [18:00:58] *siebrand makes a little dance with ialex. [18:01:08] :) [18:10:06] 03ialex * r45032 10/trunk/phase3/ (2 files in 2 dirs): Per siebrand: typo [18:12:32] Nikerabbit: is very useful and much better than
- you can eg. easily do css line-numbering there without luttering the source code with those numbers - thus it's easy to copy and past the code [18:14:46] 03rotem * r45033 10/trunk/phase3/maintenance/language/checkLanguage.inc: Localization script: Adding the possibility of checking all FlaggedRevs messages. [18:17:20] 03rotem * r45034 10/trunk/extensions/FlaggedRevs/language/ (OldReviewedPages.i18n.php RatingHistory.i18n.php): Localization update for he. [18:21:11] Danny_B, I agree, is nice and
should almost never be used (it's almost always presentational). [18:21:58] yay! we do agree with each other, where is the champagne? ;-) [18:22:34] 03ialex * r45035 10/trunk/extensions/Translate/ (SpecialMagic.php Translate.i18n.php): Per r45031: removed skin names localisation from Special:Magic [18:22:51] Simetrical: well, what would be the semantic then to force line break before some specifict statement or something [18:24:56] Nikerabbit, poems and so forth. [18:25:01] Where "line" is a logical, semantic unit. [18:25:04] Or code. [18:25:30] Simetrical: yes, but how about in running text which should autowrap? [18:26:19] Eh? Ordinary text wraps automatically, you don't need
or . [18:26:33] Simetrical's got the point [18:30:17] Simetrical: but if one wants to have forced line break, before image for example [18:30:32] Nikerabbit, the image should logically be in its own paragraph or something in that case, probably. [18:30:39] 03siebrand * r45036 10/trunk/phase3/languages/messages/ (91 files): Convert $skinNames to messages. Incorrect entries have been skipped (bs, sr-el, sr-ec) [18:30:49] Simetrical: nope, I have book right there which does that [18:31:29] "In the equation
x=y*bax
we solve the problem." [18:32:27] certainly it is wrong semantically to put paragraph break there, isn't it [18:34:30] It's still a presentational issue. You could do something like
if you liked. [18:34:42] I mean, really you can replace all
with
without adding or removing any semantic value. [18:34:55] But the class adds more semantics. [18:35:13] You could then style it inline if you liked without changing the source. [18:37:38] 03siebrand * r45037 10/trunk/extensions/ (135 files in 121 dirs): Localisation updates for extension messages from Betawiki (2008-12-25 17:54 CET) [18:40:14] 03(NEW) Add $wgDefaultSkinForAnons as a complement to $wgDefaultSkin - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=16790 15enhancement; normal; MediaWiki: User preferences; (danny.b) [18:41:28] Simetrical: mmm that's true [18:42:32] Nikerabbit: in your example the equation is emphasized by being standalone on line which means that it is not a part of running text [18:44:45] in fact in the regular book it would be printed as paragraph [18:46:27] maybe visually, but not semantically [18:47:04] *al_tally has installed his first wiki. Plz congratulate me :D [18:47:57] 03ashley * r45038 10/trunk/phase3/includes/Skin.php: [18:47:57] only call pageStats function in Skin::printFooter if $wgArticle is an object. [18:47:57] this might not always be the case (tested with a custom skin extending the Skin [18:47:57] class) and if so, a fatal error is shown on the page (tested with Special: [18:47:57] pages) and footer won't be rendered [18:49:19] Nikerabbit: well equation itself is semantically not part of running text [18:49:31] imagine more complex equation [18:49:35] say matrix [18:49:50] it's not being read as plain text [18:50:00] it must be treated different way than plaintext [18:50:39] Danny_B: but a sentence can't be split into multiple paragraphs [18:51:11] then you'll use [18:51:44] Danny_B: but the text should autowrap [18:52:12] yes [18:53:35] so far i remember (not saying it's 100% this way) does not behave like
 (no wraps) - if the width of page is smaller than length of text in  it should normally wrap
[19:00:43] 	That's a stylesheet issue.
[19:01:24] 	correct
[19:02:31] 	Nikerabbit: because you can have virtually infinite width of display
[19:02:54] 	I love infinite
[19:03:23] 	it's line by semantic not by presentation
[19:03:55] 	whatever, i guess xhtml2 is unfortunately further than html5
[19:04:09] 	they should join and take the best of both
[19:05:33] 	either the path goes by HTML or XHTML
[19:05:39] 	advancing by both is... odd
[19:06:18] 	03(NEW) move=autoconfirmed - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=16791 normal; normal; Wikimedia: General/Unknown; (cenarium.sysop)
[19:06:28] 	i updated the Polyglot extension and now I have
[19:06:30] 	Warning:  array_keys() [function.array-keys]: The first argument should be an array in /var/www/html/wikilivres.info/mediawiki-1.13.2/extensions/Polyglot/Polyglot.php on line 78
[19:07:03] *Danny_B 	would appreciate if html5 features were in  xhtml2 and xhtml2 was released asap. but it seems nearly dead :-(
[19:08:27] 	line 78 is: $wgPolyglotLanguages = array_keys( $GLOBALS['wgLanguageNames'] );
[19:08:43] 	Nikerabbit: in fact in your case you should wrap the equation in  with display: block
[19:08:50] 	hello, brion 
[19:09:14] 	merry xmas to y'all
[19:09:31] 	Merry Christmas, brion
[19:09:32] 	brion: did you get any new apple product? ;-)
[19:09:46] 	hehe not yet ;)
[19:20:05] 	any idea about my pb please?
[19:34:11] 	03(NEW) Edits of global bots are not autosighted - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=16792 15enhancement; normal; MediaWiki extensions: FlaggedRevs; (Ahonc.ua)
[19:38:06] 	14(INVALID) Image not shown in stable versions after move to Commons - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=16789  +comment (10JSchulz_4587)
[19:44:56] 	03pinky * r45039 10/trunk/extensions/GlobalUserrights/ (GlobalUserrights.php GlobalUserrights_body.php): Minor fix to GlobalUserrights: fixed bug where global rights wouldn't show up in special:listusers if the user only had one right because of a change in SpecialListusers.php (r41350)
[20:11:44] 	03aaron * r45041 10/trunk/extensions/FlaggedRevs/ (FlaggedRevs.hooks.php FlaggedRevs.php): (bug 16792) Improve global bot handling
[20:11:50] 	03(mod) Don't offer meaningless protection levels (e.g., move= autoconfirmed with moves restricted to autoconfirmed) - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=16791  summary; +comment (10Simetrical+wikibugs)
[20:12:24] 	03(FIXED) Edits of global bots are not autosighted - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=16792  +comment (10JSchulz_4587)
[20:15:36] 	03(mod) Need a method to refresh category member counts - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=16765  +comment (10matthew.britton)
[20:25:11] 	03(mod) Save CodeReview diffs to database instead of transitory cache - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=16774  (10JSchulz_4587)
[20:29:32] 	03(mod) API and UI limits are inconsistent - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=16781  (10roan.kattouw)
[20:30:37] 	03(mod) HTML codes visible on Meta - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=16788  +comment (10mikelifeguard)
[20:34:38] 	Hello there
[20:35:13] 	03(mod) API and UI limits are inconsistent - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=16781  (10matthew.britton)
[20:35:28] 	Is there someone quite skilled to help me to solve a complex issue with ParserFunctions on a highly-used Wikimedia Commons template?
[20:38:27] 	The problem is the following: let's look at a random featured picture, let's say http://commons.wikimedia.org/wiki/Image:Diving_emperor_penguin.jpg
[20:38:44] 	03(mod) Don't offer meaningless protection levels (e.g., move= autoconfirmed with moves restricted to autoconfirmed) - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=16791  +comment (10mrzmanwiki)
[20:39:33] 	Hmm, wait
[21:16:17] 	I'm having a problem with deletion
[21:16:27] 	my wiki won't let me delete
[21:17:14] 	can anyone help
[21:17:22] 	Majorly_, wont as in? permissions wont?
[21:17:24] 	or doesnt delete?
[21:17:45] 	it just doesn't delete. Comes up with an error page. I can protect and move pages, and edit
[21:18:00] 	error page says?
[21:18:11] 	Sorry! This site is experiencing technical difficulties.
[21:18:11] 	Try waiting a few minutes and reloading.
[21:18:11] 	(Can't contact the database server: Unknown error (mysql1.000webhost.com))
[21:18:25] 	1.13.3?
[21:18:38] 	yes, i only created it today
[21:20:35] 	lol
[21:21:19] 	it's annoying...
[21:21:34] 	hmm
[21:22:17] 	Presumably if it was a database permission issue it'd tell you..
[21:25:13] 	anyone else...?
[21:25:50] 	sounds like missing DELETE perms, but i'm unsure of that.
[21:26:14] 	what i was thinking
[21:26:33] 	Majorly_, do you have phpmyadmin or similar?
[21:26:42] 	yes
[21:27:04] 	"Can't contact the database server" doesn't sound like a db permission problem...
[21:27:13] 	sure that only happens when you try to delete?
[21:27:18] 	and always if you try to delete=
[21:27:22] 	have you tried editing?
[21:27:32] 	yes
[21:27:37] 	i also made myself a bot
[21:27:47] 	it may just be that the db is temporarily overloaded and doesn't let you connect because of that
[21:28:18] 	but just for delete?
[21:29:56] 	03(mod) Image not shown in stable versions after move to Commons - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=16789  (10x00000000)
[21:31:18] 	oh crap
[21:32:29] 	03(NEW) text entered into the log entry should remain when an error is returned - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=16793 15enhancement; normal; MediaWiki: Export/Import; (mikelifeguard)
[21:32:55] 	?
[21:33:18] 	i just deleted the page table
[21:33:25] 	\o/
[21:33:57] 	oops
[21:37:38] 	i updated the Polyglot extension and now I have
[21:37:41] 	Warning:  array_keys() [function.array-keys]: The first argument should be an array in /var/www/html/wikilivres.info/mediawiki-1.13.2/extensions/Polyglot/Polyglot.php on line 78
[21:37:44] 	line 78 is: $wgPolyglotLanguages = array_keys( $GLOBALS['wgLanguageNames'] );
[21:37:58] 	I have a little problem with using pretty urls (mod_rewrite). I don't have access to the apache.conf but I can use .htaccess files. However i'am unable to use the Alias statement ... suggestions ?
[21:38:33] 	it goes error 500 ... and the log file tells me ".htaccess: Alias not allowed here"
[21:38:40] 	analog_: if you can't use alias, use rewrite rules.
[21:39:17] 	see the instructions for the preferred method without root access
[21:39:39] 	Duesentrieb: just figured it out reading the apache mod_alias docs :) thanks
[21:40:11] 	how to revert to a previous version of the extension?
[21:40:47] 	These changes are incompatible with MediaWiki versions below 1.13 r32309 <- http://svn.wikimedia.org/viewvc/mediawiki/trunk/extensions/Polyglot/Polyglot.php?sortby=rev&r1=45025&view=log
[21:41:13] 	Duesentrieb: thank you, found it :)
[21:42:21] 	yannf: delete, get an older version and install it.
[21:44:40] 	Reedy, I'm starting over...
[21:44:46] 	sounds a good idea
[21:45:15] 	nuke the tables and run the setup script again..
[21:54:34] 	Duesentrieb, I installed Revision 20722 from march 2007, but still the same pb
[21:55:43] 	yannf: so maybe you don't need an older version, but a newer version... except ther eisn't any :) I have not tested the extension with the latest mw.
[21:55:57] 	though i htough it should work with 1.13. not sure about 1.14
[21:56:59] 	Duesentrieb, well, the redirect according to language works but there is an error message
[21:57:50] 	but it seems the pb is elsewhere
[21:58:30] 	the line where there is an error did not change in the old version
[22:03:24] 	Anybody know of any auth plugins to easily share logins among a wikifarm?
[22:03:50] 	CentralAuth?
[22:03:54] 	!centralauth
[22:03:54] --mwbot--	CentralAuth extension allows global/shared accounts between projects.  For more information, see 
[22:04:02] 	(what wmf uses)
[22:04:30] 	saw that, but then there's the big scary 'dont use this or a grue might eat your grandma' warning on it
[22:04:37] 	lol
[22:04:57] 	OverlordQ: use the shared user table thingy.
[22:05:05] 	it's pretty streight forward
[22:05:55] 	only if you're using MySQL
[22:07:15] 	right
[22:07:41] 	well, it *should* also work with PG.
[22:07:46] 	but no ideaif it actually does
[22:08:38] 	It'd be nice, but doesn't
[22:08:57] 	PG doesn't do cross DB queries
[22:16:14] 	ooooo, idea.
[22:18:31] 	is there any problem with me manually updating the current revision of an article using an insert statement?
[22:18:58] 	i need to make a lot of updates and pywikipediabo is sloooooooooooow
[22:19:50] 	that sounds liable to break something in a bad way
[22:26:01] 	aib: Write a quick script using the API?
[22:26:13] 	Manual insertions are nearly certain to break something.
[22:31:07] 	03(NEW) $wgSharedDB PostgreSQL support - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=16794 15enhancement; normal; MediaWiki: Database; (overlordq)
[22:31:08] 	03(mod) PostgreSQL/pgsql support (tracking) - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=384  (10overlordq)
[22:36:10] 	it has always struck me that MySQL's idea of "databases" pretty much matches PG's schemas
[22:37:43] 	yea
[22:38:11] 	03(mod) Access QueryPage-based special pages via API - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=14869  (10sam)
[22:41:01] 	Marybelle, thanks for some reason I always overlook the api
[22:44:35] 	but doesn't pywikipediabot use the api nowadays anyway?
[22:45:57] 	hello
[22:46:09] 	can i you your parser to my own website?
[22:47:21] 	aib: Not yet. It's being worked on, slowly.
[22:47:39] 	Chommik: You're free to use any part of MediaWiki as long as you abide by the license.
[22:47:57] 	GNU GPL, am I right?
[22:48:02] 	http://mediawiki.org/wiki/Special:Version <-- License for MediaWiki.
[22:48:17] 	(Yes.)
[22:49:21] 	then which files i need to copy to get only parset with e.g '''bold'''
[22:50:29] 	Chommik: Parser.php ?
[22:51:12] 	because in includes/parser there are some other files, do i need it?
[22:52:21] 	Parser.php is in that folder, no?
[22:52:29] 	yes
[22:52:57] 	Then you'll definitely need it. Though I'm not sure why you'd ever want to use MediaWiki's parser. It's a complete mess. ;-)
[22:53:45] 	because i can't find as expanded parser as is in MediaWiki
[22:54:15] 	And syntax is easy to understand
[22:54:21] 	That's true.
[22:54:47] 	But the belly of the beast is an induplicable mess. :P
[22:55:32] 	if I have function such as e.g ParseItToHtml("somestring") it will be good :)
[23:04:12] 	is there any documentation?
[23:05:39] 	hurm, got it kinda working
[23:07:18] 	right, there is very big mess in parser
[23:58:17] 	14(DUP) Update.php is broken when Postgres DB is < 8.2 - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=16711  +comment (10greg)
[23:58:18] 	03(mod) Incorrect PostgreSQL version requirements - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=15896  +comment (10greg)
[23:58:49] 	03(mod) Incorrect PostgreSQL version requirements (IF EXISTS) - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=15896  summary (10greg)