[00:00:15] Nowait, did the WRT do B? [00:00:30] WRT54G? [00:00:40] No, the N model [00:00:43] 03mrzman * r42964 10/trunk/phase3/includes/api/ApiProtect.php: (bug 16207) - Fix title protection in the API. $expiry is cast to an array near the start and doesn't have the correct timestamp format here anyway. [00:00:57] Dantman, mhm [00:00:59] My father has a router that supports N, but I'm really only using the G [00:01:03] 03(mod) Protecting against creation protects expiring on the epoch - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=16207 +comment (10mrzmanwiki) [00:01:13] Most WiFi routers support at least 2 versions of WiFi [00:01:17] Most commonly b/g [00:01:20] Updated to revision 42963. [00:01:21] _svnup [00:01:21] Updated to revision 42964. [00:01:27] The N's normally support g/n [00:01:29] 2 svnups in 5 minutes O_O [00:01:38] Though some might handle b/g/n [00:02:12] The crappy WiFi we have here supports only an old b [00:02:13] *gewt has a php bot updatign mediawiki to the latets SVN versiob for him [00:02:26] Our Wii works with it [00:02:43] So the Wii should be supporting b/g [00:03:05] But N is to new for the Wii/DS which was released awhile ago [00:03:10] woo [00:03:11] fixed [00:03:14] wrogn window [00:03:35] Apple hardware should be fine since they often create this bleeding edge tech [00:03:46] aye! [00:04:57] Or at least use it [00:04:57] So any new Apple hardware has a good chance of supporting b, g, and N [00:09:08] 03(mod) Protection log should display the protected as the first portion - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=16164 +comment (10mrzmanwiki) [00:21:23] 03dantman * r42965 10/branches/titlerewrite/phase3/: [00:21:24] Deleting failure... Yet again ;) [00:21:24] trying to handle a svn branch failed twice. Ideally now that I'm on [00:21:24] Ubuntu instead of Windows and I have svnmerge already through an apt [00:21:24] package I should be able to actually branch properly. [00:25:07] ^_^ Ok, guessthe connection did fail [00:25:38] mmmm... server proxy buffered irc [00:28:32] 03dantman * r42966 10/branches/titlerewrite/phase3/: Branching /trunk/phase3 to /branches/titlerewrite/phase3? [00:28:41] ^_^ Now to see if my svn branching utterly failed [00:29:13] whohoo, it worked [00:38:31] 03(mod) Protection log should display the protected as the first portion - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=16164 (10N/A) [01:00:06] Hey I've narrowed my issue with installing mediawiki down some, it seems that no matter what I put into the installation field it ignores it. [01:00:21] 64 Connect Access denied for user 'httpuser'@'localhost [01:00:21] ' (using password: NO) [01:01:08] any ideas why it might be attempting to connect as the apache user? [01:02:22] 03tomasz * r42967 10/branches/CentralNotice-SpecialPage-Integration/ (CentralNotice.i18n.php SpecialNoticeCentral.php): collapsing listNoticeDetail into one form. not listing templates that are already assigned [01:02:51] Toxic_: What version are you using? And what have you put into the superuser/database user fields? [01:03:46] v 1.13.2 (latest stable) and I put in 'wikiuser' and 'wikidb' both of which I created. [01:04:54] Are their any fields at all with a value of httpuser? [01:05:15] nope. [01:05:32] httpuser is the user the apache process runs as. [01:06:42] 03(mod) Create proofread page namespaces on Chinese Wikisource - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=15722 +comment (10jus168jih) [01:07:49] http://rafb.net/p/wEF4sA20.html [01:07:56] SQL Information [01:15:32] so given that MW distributions don't come with a LocalSettings.php in the tarball, how do you easily check when you're upgrading whether there are new things you ought to add to your LocalSettings? [01:15:49] diff the code in config/index.php? not worry about it? [01:15:56] check the Release Notes [01:15:56] reading the release-notes? [01:16:07] hmm, that's a radical idea. [01:16:13] if you were to diff anything, you should diff DefaultSettings, not index.php [01:16:41] well, DefaultSettings is going to get replaced in the update, so I'm not worried about that. [01:16:53] ... [01:16:53] but if any changes are made to the default LocalSettings, I wouldn't see those. [01:17:02] you diff DefaultSettings against your settings [01:17:08] new variables will be added there [01:17:20] also changes to default values [01:19:47] I see, each release notes has a "configuration changes" discussion. cool. [01:22:19] No other ideas why it ignores the values I have in the fields? [01:23:53] Toxic_, maybe it's forced by some other module to have the same mysql user as system user? [01:26:37] Toxic_: that info has nothing to do with your error [01:27:04] "Access denied for user 'httpuser'@'localhost' (using password: NO)" [01:27:10] Yes.. [01:27:21] GRANTS don't matter [01:27:23] MediaWiki is trying to connect as httpuser. [01:27:40] yes, so the relevant information, is what config do you have in your LocalSettings.php file [01:27:48] The GRANTS show that the user exists and has the permissions needed. [01:27:55] None I haven't installed this yet! [01:29:05] How did the user get created in the first place? [01:29:12] ..By me? [01:29:44] You have a superuser account? [01:30:02] yes. [01:30:07] Ok, then delete that user [01:30:13] Tried putting that in and a non-existant user. [01:30:13] 03tomasz * r42968 10/branches/CentralNotice-SpecialPage-Integration/ (SpecialNoticeCentral.php SpecialNoticeTemplate.php): adding inline previews for notice lists [01:30:19] It still tries to connect as httpuser. [01:30:48] Ya, one thing at a time, but I was just noting that, cause you may run into another issue if you try to use that existing user [01:31:05] It has permissions it doesn't need, and is missing a permission that may be needed [01:31:09] I've done that already.. [01:31:21] Best is normally to have the config installer create it on it's own [01:31:30] Then I put in the root user, and mysql root and it still trys to connect as httpuser. [01:31:35] Ok [01:32:20] How was PHP installed on your system? [01:32:27] from source. [01:32:38] Ok [01:33:01] Then perhaps we should look at the [MySQL] section of your php.ini file [01:35:03] sql.safe_mode [01:35:11] It sounds like you may have that set [01:35:20] http://rafb.net/p/L1gYPP79.html [01:35:27] http://www.php.net/manual/en/ini.core.php#ini.sql.safe-mode [01:35:52] yep I do [01:35:58] Ok, then that's the issue [01:36:43] sql.safe_mode makes it so that the arguments for connecting to the database are overridden with the default config values... because you have no default config user it uses the http user [01:36:59] Turn svn.safe_mode off and things should start working [01:37:03] sql* [01:37:55] And it works, thanks [01:41:01] Any common reason it would show a blank page after putting the localsettings file in the parent dir and removing the original? =p [01:42:56] Toxic_, the reason will be at the error log [01:47:13] And it works.. file was chowned www:www not www:httpuser. Alright one thing on my to do list down. Thanks guys. =) [01:50:15] *Dantman wonders if he should release his bash development script for other people to use [01:50:38] ^_^ I'm extending it to better handle branches [01:51:20] mmmm.... [01:51:44] svn info | grep 'Revision:' | awk '{print $2}' [01:59:26] 03dantman * r42969 10/branches/titlerewrite/phase3/: [01:59:26] Initialized merge tracking via "svnmerge" with revisions "1-42965" from [01:59:26] svn+ssh://wikimedia/svnroot/mediawiki/trunk/phase3 [02:00:19] ^_^ gotta love ssh config [02:03:30] 03tomasz * r42970 10/branches/CentralNotice-SpecialPage-Integration/ (CentralNotice.i18n.php SpecialNoticeCentral.php): restructure to vertical listing to accomodate long drop downs [02:05:51] Dantman: Branching is even harder on windows :p [02:06:09] ^_^ I know [02:06:21] I failed twice on my old system [02:06:38] Now that I'm on Ubuntu, it was quite easy [02:06:46] svnmerge even came in an apt package [02:18:46] 03tomasz * r42971 10/branches/CentralNotice-SpecialPage-Integration/SpecialNoticeCentral.php: adding ability to adjust project & language in listNoticeDetail [02:49:27] 03nad * r42972 10/trunk/extensions/PdfBook/PdfBook.php: don't remove toc for single page exports [02:51:55] 03(mod) Tidy sucks, won't leave stuff alone - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=2542 +comment (10innocentkiller) [03:29:16] 03(mod) Create proofread page namespaces on Chinese Wikisource - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=15722 +comment (10zy26) [03:42:04] 03(mod) query-continue incorrect for usercontribs - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=16195 (10thecentercannothold) [03:49:16] How do I make a skin respect the stuff in MediaWiki:Sidebar like "SEARCH" and "TOOLBOX"? [03:55:08] And is there a way to disable user's styles? [04:06:23] What was the "davinci" skin? [04:11:26] Anyone?... [04:11:52] Why does MediaWiki use BLOB all the time >< [04:13:37] In honor of the classic film? [04:22:44] 03brion * r42973 10/branches/CentralNotice-SpecialPage-Integration/CentralNotice.php: fixup defaults for loader calls -- $wgArticlePath may be unset at config time [04:39:44] 03charlie * r42974 10/trunk/phase3/ (RELEASE-NOTES includes/DefaultSettings.php includes/User.php): Added a new configuration option: $wgAllowUserSkin [04:41:41] brion's still up :/ [04:45:32] Wow, the daddio skin looks really awesome, especially with a logo: http://charlie.yourwiki.net/wiki/Home [04:45:46] Looks like its modern based [04:46:37] 03brion * r42975 10/branches/CentralNotice-SpecialPage-Integration/SpecialNoticeCentral.php: [04:46:37] Fix for r42971 -- make project/language selectors in notice detail form work correctly. [04:46:37] They worked for saving, but weren't setting the correct option as selected when re-displaying. [04:50:49] chuck: wow, that *is* sexy [04:56:04] 03brion * r42976 10/branches/CentralNotice-SpecialPage-Integration/SpecialNoticeCentral.php: some whitespace cleanup [04:57:32] chuck: that's awesome [04:57:35] 03brion * r42977 10/branches/CentralNotice-SpecialPage-Integration/ (3 files): save my sanity by putting Special:CentralNotice in SpecialCentralNotice.php :) [04:58:46] Yeah I'll definitely have to make the main page look better, but it's a good start [04:58:50] what is this special page/central notice thing we've been seeing so much of lately? [04:59:16] Mike_lifeguard: My guess is that Brion is cleaning it up for this year's fundraiser, that's coming up soon, isn't it? [04:59:28] yes, in like 2 days [04:59:31] but what is it? [04:59:48] *werdnus tickles Mike_lifeguard [04:59:48] Mike_lifeguard: That's the thing that shows that progress bar [04:59:55] on every wikimedia wiki [04:59:58] boring :P [05:00:17] werdnus: could you take a look at r42974 and tell me if its something that could make brion mad, please? ;-) [05:00:45] link to Special:Code? :) [05:01:16] http://www.mediawiki.org/wiki/Special:Code/MediaWiki/42974 [05:01:59] *chuck thinks someone should make a thing so that mwbot will format codereview links like !revision 12345, !author charlie, etc [05:03:01] I think it does !r - but probably links to viewvc or whatever it's called [05:03:08] !r 42974 [05:03:08] --mwbot-- http://svn.wikimedia.org/viewvc/mediawiki?view=rev&revision=42974 [05:03:25] meh [05:03:52] chuck: see comments. [05:05:09] replied [05:07:47] replied [05:09:26] 03brion * r42978 10/branches/CentralNotice-SpecialPage-Integration/ (SpecialCentralNotice.php SpecialNoticeTemplate.php): cleanup manual 'a href' => link() [05:14:42] 03brion * r42979 10/branches/CentralNotice-SpecialPage-Integration/ (SpecialCentralNotice.php SpecialNoticeTemplate.php): fix tab highlighting [05:17:12] Hmmmm [05:17:36] I'm starting to debate whether I should save _title_ui as page_title_ui or rev_title_ui... [05:34:57] revf [05:35:00] rev* [05:35:11] at first glance, that seems most sensible. [05:35:17] chuck: have you looked at the comments again? [05:36:09] Ya, after awhile I'm starting to think that [05:36:19] then you can track history of it. [05:36:27] :/ Though that's probably going to make this even more complex to program [05:36:51] meh, good programming is tough [05:37:06] Though, it'll be easier to handle the deletion code [05:37:28] ^_^ Time to kill my db [05:39:54] Title changing might even get a bit messy [05:40:26] Originally I was going to have page moves change that... [05:41:01] But now I might need to take a better look at the movepage backend [05:42:31] hi, i'm looking for an extension, but i dont remember the name of it [05:42:59] it prevents accounts from being created that are the same as another account that are only different by letter case [05:43:11] Mmmm... [05:43:23] ^_^ My title rewrite could make that possible [05:43:58] ;) ie: Make the User: namespace completely case-insensitive, and suddenly User:Dantman and User:DANTMAN are the exact same person [05:44:49] uberfuzzy: I think you're probably looking for http://svn.wikimedia.org/viewvc/mediawiki/trunk/extensions/AntiSpoof/ [05:45:04] yeah! [05:45:10] :) [05:48:26] werdnus: Sorry, my laptop is really sucky and it only charges while it's off, I'm back on now. So what exactly do you think I should fix/change? Disabling useskin? [05:49:10] chuck: well, it depends on what you actually want to use it for. [05:49:39] HOWEVER, I don't see it as sensible to be able to disable skins without also disabling useskin... [05:49:47] werdnus: I was also thinking adding a notice to the change skin page saying like "The wiki will currently override your skin settings with it's default skin (monobook), so the settings below won't have any visible affect. If the wiki does allow users to select different skins in the future, the one set below will be used." or something? [05:49:50] somebody could just append useskin to the URL parameters with greasemonkey or something. [05:50:00] Just gray it out. [05:50:10] *werdnus wanders off to study [06:05:37] Hi TimStarling. Happy Tim Starling Day for yesterday (since I didnt see you on then) [06:09:22] 03charlie * r42980 10/trunk/phase3/ (3 files in 3 dirs): [06:09:23] Made it so that users can't test out skins if user skins are disabled (per [06:09:23] Werdna), and user skin selection has now been disabled on the preferences [06:09:23] special page, with a notice on top telling why that happened. [06:11:53] 03charlie * r42981 10/trunk/phase3/includes/specials/SpecialPreferences.php: Let's hide the (preview) link, as it won't do anything now. [06:12:02] werdnus: All done ^ [06:14:33] hi [06:25:34] Tim: ^_^ Do you welcome the Apocalypse Title rewrite... heh [06:31:31] 03brion * r42982 10/branches/CentralNotice-SpecialPage-Integration/SpecialCentralNotice.php: [06:31:31] Far too much refactoring for this hour of the night. ;) [06:31:31] Removed lots of dupe code for generating table rows and date/time selectors. [06:31:31] Switched from 15-minute increments to just showing all available minutes, why not eh? [06:39:09] TimStarling: could you look at 42974, 42980, and 42981 and post a comment on my talk page on mw (user Cmelbye) about the quality, etc of them, please? only if you have some extra time though [06:39:18] s/at/at revisions/ [06:43:25] what's the point of it? [06:43:41] easier configuration? [06:45:56] chuck: are you svn-charlie? [06:49:14] yes, he is. [06:49:31] ok :) [06:49:32] chuck: good-o [06:50:05] chuck: you have to add new message names into messages.inc too [06:50:16] you could use $wgSkipSkins to achieve the same result [06:52:20] TimStarling: by the way, has the configuration for centralauth UDP notification been done yet? I couldn't see mention of it in *Settings.php (on noc.wikimedia.org) [06:55:04] probably not [06:55:54] agh... something is screwing with whitespace [06:56:12] I think the configuration is almost the same as core. [07:02:08] TimStarling: Looking at the code, I think adding $wgCentralAuthUDPAddress = $wgRC2UDPAddress; $wgCentralAuthNew2UDPPrefix = 'centralauth: '; to CommonSettings.php should do the trick. [07:02:12] Might want to ask Aaron, though. [07:06:01] what about the port? [07:06:21] I can't imagine that being unsafe, because private wikis have $wgRC2UDPAddress unset. [07:06:26] um, let me look at Aaron's code. [07:06:44] damn... explodes on save... need to fix [07:07:15] TimStarling: it looks like the function sendToUDP uses $wgRC2UDPPort as the destination port, and doesn't allow overriding. [07:07:32] I don't know why Aaron has overriding for the address but not the port. [07:07:38] http://svn.wikimedia.org/viewvc/mediawiki/trunk/phase3/includes/RecentChange.php?view=markup [07:07:43] search for public static function sendToUDP( $line, $address = '', $prefix = '' ) { [07:09:42] # Low-level sanity check\nif( $this->mTitle->getText() == '' ) { [07:09:50] Agh... this isn't low level enough [07:10:01] >:/ db key! [07:11:11] TimStarling: ah, wait a sec - I think I made an error in $wgCentralAuthNew2UDPPrefix -- should be "#centralauth\t" [07:11:23] sorry - I forgot we had a format for that. [07:11:25] I set it to #central\t [07:11:29] okay, good-o [07:12:30] We've probably hit the channel limit. [07:12:59] 18:12 [wikimedia] -!- #central: No such channel [07:16:21] no, it's just the same old bug [07:16:46] yeah, that's what I was talking about. [07:20:00] Whohoo... rev_title_ui is preserved on edits [07:21:23] what is that? [07:22:07] I think it's in his branch. [07:22:12] http://d.imagehost.org/view/0511/Screenshottitle_Development_Environment_titlerewrite_Mozilla_Firefox.png [07:22:50] Used to be page_title_ui, but a very short discussion proved enough to switch to rev_title_ui [07:22:58] why do you need to rewrite anything to achieve that diff? [07:23:06] it was done years ago with displaytitle [07:23:26] that screenshot, I mean [07:23:31] *TimStarling has diffs on the brain [07:24:02] heh, what have you been doing with diffs? [07:24:06] http://news.bbc.co.uk/2/hi/uk_news/wales/7702913.stm <-- Heh [07:24:28] (almost as good as http://farm4.static.flickr.com/3165/2691947209_4e4fe0d260_o.jpg ) [07:24:29] *werdnus has "the impact on society and the environment of the use of Sodium-24 in industry" on the brain. [07:24:42] HistoryBlob.php [07:24:46] http://d.imagehost.org/view/0768/Screenshot-All_pages_Development_Environment_titlerewrite_Mozilla_Firefox.png [07:25:06] big deal [07:25:43] that's hardly the reason displaytitle was crippled [07:26:02] Only a small subset of what this is about [07:26:15] it was crippled because brion felt that it was too confusing to have a different title for display and for linking [07:26:26] and too much clutter to have a subtitle [07:32:08] maybe you should send a post to wikitech-l [07:32:35] I have better things to do than stare at IRC for 5 minutes waiting for an answer [07:32:52] Title rewrite has existed since March. [07:33:04] how many posts have you sent? [07:33:05] It was already discussed in wikitech [07:33:38] TimStarling: http://en.wikipedia.org/wiki/User_talk:Werdna#Character-counting_parser_function -- Should we write a parser function for formatting quantities as the poster describes there? [07:34:59] just add a parser function to format a number in scientific notation [07:35:10] http://lists.wikimedia.org/pipermail/wikitech-l/2008-March/thread.html [07:35:59] TimStarling: they want to add spaces every three digits. [07:36:08] narrow spaces, I think [07:37:04] each to their own [07:38:07] anyway, what they're really after, I think, is StringFunctions. [07:38:13] But IIRC that needs some cleanup before it goes live. [07:39:18] Dantman: haven't we already talked about that? [07:39:40] werdnus: I don't think it should ever go live [07:39:54] TimStarling: oh, why? [07:40:07] werdnus: Yes, StringFunctions needs refactoring, it is using fairly inefficient methods of doing the simple things it does [07:40:09] it will make templates even more slow and complex [07:40:25] it will allow tasks to be done in templates that are currently impossible [07:40:47] thereby increasing the use of that unholy programming language [07:41:02] I liked your comparison of it to mIRC scripting [07:42:38] *werdnus points to Extension:Lua [07:43:30] TimStarling: on a related note... i have started to set up a template system of including bibliographic records. I don't think it's too terrible, but before deploying it on a large scale, would you mind looking it to it for a couple of minutes? the main entry point is http://de.wikipedia.org/wiki/Vorlage:ISBN [07:43:32] Lua is nice... but it was rejected for general use simply because not everyone can install it [07:43:39] werdnus: ex:Winter :) [07:43:58] Yuck, Winter was horid [07:44:02] :P [07:44:08] no, Brion didn't like it because implementing a programming language gave him the willies. [07:44:18] Winter went and reimplemented part of the parser in handling it's own {}'s [07:44:49] 03siebrand * r42983 10/branches/CentralNotice-SpecialPage-Integration/CentralNotice.i18n.php: Message updates [07:45:01] I doubt it's even properly compatible with the new parser [07:45:01] TimStarling: i'd also like to talk to you about the possibility of including this type of info from a central repos into multiple wikis... but I have to go afk for a bit soon, so now is probably not the best time for that. [07:46:18] Duesentrieb: I stopped reading at 'Template system' [07:46:52] is this an expected diff in the 'new preprocessor'? http://meta.wikimedia.org/w/index.php?title=Talk:Migration_to_the_new_preprocessor&curid=112097&action=history (verified at http://en.wikipedia.org/wiki/Special:ParserDiffTest?pdt_title=Main+Page&pdt_input=%5B%5BImage%3AWiki.png%7C19.px%5D%5D )... wow, 7 months late [07:46:55] 03siebrand * r42984 10/trunk/phase3/ (2 files in 2 dirs): Follow up on r42980. Register message and fix alignment. [07:47:18] werdnus: whatever you call it, it's less complex that the average infobox on enwp. [07:48:33] Splarka: I never saw 19.px. I would say it was a mistake [07:48:44] Duesentrieb: :) [07:49:14] it's not preprocessor related [07:49:18] it's just bitrot [07:50:36] something that happens to have changed behavior this year at some point, or did the old parser regress? [07:51:34] I don't know [07:51:45] does it matter? [07:51:52] not really [07:51:57] (and since the oldPP isn't apparently accessable anymore except via ParserDiffTest, is that special page even need anymore on wmf?) [07:53:35] probably not [07:54:10] should we open a +shell bug to remove it, like makesysop/makebot? since it will just continue to return more false positives at this point [07:54:30] it's gone [07:54:41] wow, that was easy [07:54:48] *Splarka didn't even feel himself submitting the bug [07:55:01] *Splarka gives Tim a bag of spooky candy [07:55:37] *werdnus completes a form, signs it in triplicate, sends it in, sends it back, queries it, loses it, finds it, subjects it to public inquiry, loses it, and buries it in soft peat for three months and recycles it as firelighters. [07:55:41] TimStarling: While you have it open, there are five extensions that can be disabled, if you want. [07:56:14] "Welcome to Wikipedia, where the reward for a job well done is 5 more jobs" [07:56:32] Isn't that true anywhere? [07:56:56] it is better when you get paid a commission [07:57:20] ^ good for used car salesman, ^ bad for salaried worker, ^ really bad for volunteer [07:57:56] heh [07:58:43] Poem, Makesysop, UniversalEditButton, Usernameblacklist, RenderHash. [07:58:57] 03(mod) query-continue incorrect for usercontribs - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=16195 (10roan.kattouw) [07:58:57] what did RenderHash do? [07:58:58] <_mary_kate_> why universaleditbutton? [07:59:10] it's in core. [07:59:26] I've never understood UniversalEditButton. [07:59:35] Poem, UEB, Makesysop, and RenderHash are in core. [07:59:42] Usernameblacklist is deprecated in favor of the Titleblacklist. [08:00:42] RenderHash has something to do with secure.wikimedia I think. [08:00:54] 03raymond * r42985 10/trunk/extensions/ (2 files in 2 dirs): ParserDiffTest disabled on WMF wikis. Obsolete now. [08:01:34] *werdnus watches the family guy star wars parody. [08:03:25] 03raymond * r42986 10/trunk/extensions/Translate/groups/MediaWikiExtensions.php: ParserDiffTest disabled on WMF wikis. Obsolete now. [08:27:44] <_wooz> lo [08:28:15] 'lo. [08:36:40] 03(mod) Provide preference-based autoformatting for unlinked dates - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=4582 (10s.r.e.turner) [08:59:20] 03(mod) query-continue incorrect for usercontribs - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=16195 (10thecentercannothold) [09:00:09] 03(NEW) Video metadata parsing borked for latest version of ffmpeg2theora - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=16208 normal; normal; Cortado: jheora; (sco_scam) [09:08:32] 03rotem * r42987 10/trunk/phase3/ (3 files in 2 dirs): Localization update for he, and reordering the right and action messages. [09:25:17] 03dantman * r42988 10/branches/titlerewrite/phase3/ (21 files in 4 dirs): [09:25:17] Committing todays work on the TitleRewrite branch. It's a fair start, [09:25:17] differs from the original branch and is probably a little more stable [09:25:17] than it. [09:39:54] 03tstarling * r42989 10/trunk/phase3/includes/HistoryBlob.php: [09:39:54] New DiffHistoryBlob storage format and compression method, which maintains a [09:39:54] good compression ratio in the case of vandalism by replacement with small text. [09:39:54] The storage format allows edit wars to be dealt with as well, but an efficient [09:39:55] algorithm for that may be more difficult. [09:40:30] <_mary_kate_> TimStarling: delta storage? [09:40:40] <_mary_kate_> or something more complex? [09:42:06] before that commit, it was delta storage [09:42:15] after the commit, it's delta storage with optional reordering [09:42:30] so you can arrange all the vandalism in one block, so it doesn't affect the main sequence of edits [09:43:54] TimStarling: ooh, spiffy. [09:44:19] If the text of two revisions is the same, does it store it twice in es, or does it nicely use the content already there? (that may help revert-only edit wars at least) [09:44:34] it stores it twice [09:44:45] <_mary_kate_> MattJohnston: that particular problem can be fixed without changing the storage method [09:44:58] <_mary_kate_> there just needs to be some logic to find an existing old_id with the same content [09:45:00] I was just curious whether it did [09:45:02] but if they're recompressed to the same HistoryBlob, then they can be stored effiicently [09:45:17] <_mary_kate_> (i think rollback already does this?) [09:45:23] I assume having some sort of checksum lookup would be the best way, like we do for images [09:45:37] <_mary_kate_> please, not like we do for images ;) images need a total rework [09:45:49] under certain circumstances, multiple revisions can share a text ID [09:45:57] I'm not sure if rollback is one of those [09:46:27] but if the edits are closely spaced and you recompress with DiffHistoryBlob, then it'll just be a null diff anyway [09:51:03] anyway, removing duplicate revisions alone would gain next to nothing [09:51:24] currently we're storing every item of text individually compressed [09:52:07] you get a compression ratio of maybe 2-4 [09:52:17] I don't think we're running out of disk space, anyway. [09:52:40] It's the size of an individual revision that matters, IIRC. [09:52:41] it's closer than you might think [09:53:11] (since IO is at a premium) -- correct me if I'm way off.. [09:53:14] with DiffHistoryBlob you can get a compression ratio of maybe 100 [09:53:20] whee [09:53:24] <_mary_kate_> some people on en are claiming that a proposed featured video isn't high enough quality [09:53:34] <_mary_kate_> but it's already 480Kbps... do they want people to actually be able to watch the video? [10:03:29] _mary_kate_: maybe they think we have video thumbnailing. [10:03:50] <_mary_kate_> that'd be pretty neat actually [10:03:58] yeah [10:04:34] very neat [10:04:45] but kinda expensive, CPU-wise. [10:04:48] <_mary_kate_> would probably need 2x cpu capacity [10:05:32] If somebody wanted to DDoS us, they'd just have to request hundreds of videos at 100px, 101px, 102px, 103px, ... [10:08:41] that would only DoS the scalers [10:08:46] maybe only allow powers of two [10:08:46] and maybe the image backends [10:09:08] 03dantman * r42990 10/branches/titlerewrite/phase3/includes/ (Article.php Title.php): [10:09:08] TitleRewrite: Making old revision viewing and page diffs respect the [10:09:08] title for the individual revision. [10:10:08] <_mary_kate_> TimStarling: depends, if there's something like the fast scaling method for jpeg (if you scale to a particular reatio) for mpeg/ogg, it might work [10:10:19] TimStarling: depends on whether DB connections are held open while images are scaled. [10:10:24] If so, it'd be fun. [10:11:50] As I remember, unless you are using GD the scaling is handled completely in a separate process from the web app [10:12:34] one of the bonuses of using ImageMagick instead of GD was that MW didn't need to wait for something to scale before it served out the page, so things ran smoother [10:13:38] Hmmm... guess it's time to ./manage offline for today [10:15:57] <_mary_kate_> we don't do image scaling on apaches, we use a 404 handler [10:16:28] does it open a DB connection, though? [10:17:06] <_mary_kate_> there's no reason it would need to, but i've no idea if it actually does or not [10:17:31] maybe to check the image table.. [10:18:51] <_mary_kate_> shouldn't need to do that [10:18:59] <_mary_kate_> except possibly for weird stuff like archived images [10:32:24] 03(NEW) Deleted images are not watched - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=16209 trivial; normal; MediaWiki: Images; (le.korrigan) [10:36:56] ^_^ Oh right, heh... I just remembered something... The title rewrite branch was why I got commit access in the first place [10:39:23] 03(mod) Resizing of edit box on screen using JavaScript - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=9370 +comment (10bawolff+wn) [10:58:58] 03(NEW) Date Format not always consistent - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=16210 minor; normal; MediaWiki: Page rendering; (paul) [11:20:52] 03maikmerten * r42991 10/trunk/cortado/src/com/fluendo/ (4 files in 3 dirs): Estimate playback position according to how many bytes have been read from source if no duration is given. [11:33:24] 03(mod) Parser function on Korean Wikipedia - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=13712 (10camway) [11:48:17] I wonder why MW shows the action bar of the main page in http://meta.wikimedia.org/?diff=1252937 and works in http://meta.wikimedia.org/?diff=prev&oldid=1252937 [11:57:21] because the first one is a broken URL and you should be happy that it doesn't just throw an exception [12:00:45] got to restart, brb [12:05:13] 03(mod) Install hanp arser function on Korean Wikipedia - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=13712 +need-review +shell; summary (10siebrand) [12:06:27] 03(mod) Install hanp extension parser function on Korean Wikipedia - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=13712 summary; +comment (10siebrand) [12:06:37] 03tstarling * r42992 10/trunk/phase3/ (5 files in 2 dirs): (log message trimmed) [12:06:37] * Added $wgDebugLogPrefix, to allow users (or in this case, multiprocess command-line scripts) to set a debug log line prefix [12:06:37] * Improved blob_tracking.sql docs [12:06:37] * Allow testCompression.php to run until happy instead of a specified revision count [12:06:39] * Added docs for some debugging globals in DefaultSettings.php [12:06:41] In experimental script recompressTracked.php: [12:06:43] * Fixed crippling bugs, seems to work now. Needs more testing. [12:43:11] *werdnus loses the game [12:50:04] 03(mod) Pseudo-Namespace on Korean Wikipedia - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=15880 (10camway) [12:53:56] 03minuteelectron * r42993 10/trunk/extensions/Babel/BabelAutoCreate.class.php: [12:53:56] * Use $user parameter to doEdit() rather than messing about with $wgUser. [12:53:56] http://www.mediawiki.org/wiki/Special:Code/MediaWiki/42666 [13:21:39] hi werdnus [13:22:11] can someone explain why you cant move a category? [13:24:17] 03tstarling * r42994 10/trunk/phase3/maintenance/storage/recompressTracked.php: Typo [13:26:39] TimStarling: Is there any particular reason you used $wgUser in the double redirect fixer instead of the $user parameter of Article::doEdit() - out of interest? [13:27:36] wasn't it added only after it? [13:29:01] ok then [13:29:54] just a feeling [13:32:07] the redirect fixer is switched off isn't it? [13:33:50] didn't you switch it off yourself :) [13:33:57] I don't know, but when I was developing the category auto-creation feature on my Babel extension I based it on some of your code but aaron said that my code should use the $user parameter - so I wondered if you had a particular reason for doing it your way? [13:35:28] new code should use the $user parameter [13:36:21] 03tstarling * r42995 10/trunk/phase3/includes/HistoryBlob.php: [13:36:21] Implemented xdiff_string_bpatch() in pure PHP, allowing DiffHistoryBlob objects [13:36:21] to be decompressed without libxdiff. Negligible performance difference. This [13:36:21] provides a stability guarantee -- if libxdiff disappears or changes, we can [13:36:21] continue to read old blobs. It also means we don't have to package libxdiff for [13:36:24] all app servers. [13:37:36] pretty cool eh? [13:37:41] MinuteElectron: speaking of Babel, that wasn't working for us a few weeks ago =( [13:38:16] TimStarling: wow, pure php as fast as C? [13:38:23] no [13:38:41] negligible in the context of the typical use? [13:38:54] it's probably 1000 times slower, but the difference is negligible compared to the time it takes to decompress a blob [13:39:14] it's probably dominated by gzip [13:40:03] that wasn't the cool bit, the cool bit is that there's no reason why we can't have diff-based history compression on wikimedia [13:41:02] which means we can defer some ES upgrades? [13:41:12] or is the load on apaches? [13:41:25] is wikitext allowed in comments of codereview revisions? [13:41:36] yes. [13:42:19] what does resolved mean in codereview? [13:42:36] it was broken, but now it's fixed. [13:44:36] chuck: oh, what's the issue? [13:44:59] MinuteElectron: I've forgotten now, sorry, but I can reinstall it, one sec [13:50:53] 03(NEW) Group 'emailconfirmed' not set after confirm link visited - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=16211 04CRIT; normal; MediaWiki: User login/settings; (huck.finn) [13:51:29] MinuteElectron: There used to be a php error, but now it's just not showing anything where I put the tag. Do I need to add templates for the languages? [13:55:43] what's the api to get the GlobalBlockList? [13:57:47] i dunno if there is one [13:59:11] I didn't write one [13:59:14] maybe somebody else did [14:13:22] chuck: no... it should just work. [14:13:32] chuck: for me and betawiki it just works out of the box [14:27:10] 03(mod) Provide preference-based autoformatting for unlinked dates - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=4582 (10ashtongj) [14:42:58] 03(NEW) Some messages should use PLURAL - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=16212 15enhancement; normal; MediaWiki extensions: NewestPages; (Eugene.Zelenko) [14:44:08] 03(mod) Group 'emailconfirmed' not set after confirm link visited - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=16211 +comment (10Platonides) [15:03:25] I am using enscript to highlight text, but everytime I highlight text its keep putting a footer saying "created by Enscript". How do I remove that???? [15:22:04] Hi. I'm trying to write a new special page. i adapted SpecialMIMEsearch to display 5 random file links. Atm it's not random yet and since it are 5 I don't want the navigation links. any hint how to remove them? here's what i have so far http://rafb.net/p/cQ30S892.html [15:26:35] Led_Zeppelin: first, check 'man enscript' for a command line option. if there isn't one, then find out where that line is being added in the source, and remove it [15:28:32] Led_Zeppelin: if you are generating HTML, the file to edit is lang_html.st [15:37:17] hello, I'm trying to change some text in Special:AllMessages, but on the screen it doesn't change. [15:37:43] you're editing the right mediawiki: pages? [15:38:28] carl-m: MediaWiki:Prefs-help-email/de [15:38:47] carl-m: this is the Mediawiki page, which i changed [15:39:32] Can it be problem about Cache? [15:39:48] i have closed both client and server side cache [15:40:15] then it's probably not the cache. double check that the change does appear on the mediawiki page if you view that directly [15:42:48] the change can be seen in Mediawiki:Prefs-help-email/de and Special:AllMessages [15:43:24] but not at the register page, where this message should be displayed [15:43:46] I see [15:44:41] what does that say? [15:46:47] carl-m, I don't see a lang_html.st [15:47:06] http://www.mindxuniversity.org/tingwang/allmessages.png here is a screenshot [15:47:07] also, I looked at the man page I see no where it says footer [15:47:28] Led_Zeppelin: the footer option doesn't seem to work. The file is located at /usr/share/enscript/hl/lang_html.st on my computer [15:47:40] might be elsewhere for you [16:00:37] 03river * r42996 10/trunk/tools/nsapi-dtrace/: new directory [16:01:50] 03river * r42997 10/trunk/tools/nsapi-dtrace/ (7 files): NSAPI module + sample scripts for DTracing sun web server [16:08:21] 03btongminh * r42998 10/trunk/extensions/CodeBrowse/ (CodeBrowse.i18n.php CodeBrowseView.php): [16:08:21] * Add a CodeBrowseBeforeView hook [16:08:21] * Add one missing i18n [16:23:10] 03aaron * r42999 10/trunk/extensions/CodeReview/CodeRevisionCommitter.php: Unbreak tag removing ;) [16:26:31] 03(NEW) hyphen-minus is unreadable - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=16213 15enhancement; normal; MediaWiki extensions: ConfirmEdit; (giecrilj) [16:43:35] Could someone go to my talk page on the English Wikipedia and post a test message? [16:43:47] I'm testing the new message notice [16:44:03] User Talk:Techman224 [16:44:35] 03aaron * r43000 10/trunk/phase3/includes/specials/SpecialBlockip.php: Pass user id to newFromDB() as well [16:45:52] posted [16:50:44] Subfader: Thanks [16:50:55] test was successful [16:52:55] techman224: do you like it? [16:53:14] you could more completely customize it with js, but I thought the CSS got it close enough [17:03:16] hi, is "MediaWiki:PDF Template Blacklist" for excluding templates from pdf generation? [17:04:15] I try to add a new special page to display 5 random image links. I adapted the code from SpecialMIMEsearch.php Since it are only 5 files to be displayed I don't need the navigation links. How can I remove them? Here's what I have so far http://rafb.net/p/aTk5LB61.html [17:10:01] 03(mod) Group 'emailconfirmed' not set after confirm link visited - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=16211 (10huck.finn) [17:30:13] :( [17:35:22] hello [17:35:25] 03ialex * r43001 10/trunk/extensions/CodeBrowse/ (CodeBrowseItemView.php SpecialCodeBrowse.php): fix some errors [17:35:30] hi Nikerabbit [17:36:55] *Nikerabbit is cleaning his hard disk from 0% free [17:39:03] O_o [17:39:54] ialex: yeah, isn't it nice [17:40:31] Solarka: Yes [17:40:49] Splarka: Yes I do [17:41:06] It's more user frendly [17:53:08] 03ialex * r43002 10/trunk/extensions/Configure/ (18 files): (log message trimmed) [17:53:08] Added database-based storage (inspired by ConfigureWMF extension): [17:53:09] * Users are able to choose beetween file-based and database-based storage [17:53:09] * a new maintenance script migrateToDB.php has been added to migrate configuration files to database storage [17:53:12] * database settings and memcached settings are not editable with it since they are required to get the configuration [17:53:15] * default storage is files for backward compatibility (don't forget that there's migrateToDB.php ;) [17:53:17] * dropped MediaWiki < 1.13 compatibility (requires r32578+) [17:55:29] 03brion * r43003 10/branches/CentralNotice-SpecialPage-Integration/ (SpecialCentralNotice.php SpecialNoticeTemplate.php): use makeLinkObj() instead of link() so it works on 1.13 [17:57:11] 03ialex * r43004 10/trunk/extensions/Configure/CHANGELOG: forgot to add an entry in that file... [17:58:22] ialex: back to 27% :) [18:03:35] 03(mod) Group 'emailconfirmed' not set after confirm link visited - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=16211 (10huck.finn) [18:05:17] 03brion * r43005 10/trunk/extensions/CentralNotice/ (13 files): [18:05:17] Merge work branch back to trunk... [18:05:17] svn merge -r41613:HEAD svn+ssh://svn.wikimedia.org/svnroot/mediawiki/branches/CentralNotice-SpecialPage-Integration . [18:06:48] brion: codereview is flooded with that stuff you know :) [18:07:45] that's kin dof the point dude [18:07:49] that way i can review the stuff :D [18:07:56] 03(mod) Group 'emailconfirmed' not set after confirm link visited - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=16211 (10huck.finn) [18:08:17] brion: yeah, just looks kind of accessive [18:08:21] soooo many commits [18:08:35] it makes the SMW people look like panzies [18:08:55] Yeah. Especially, I believe it was two days ago, like 30 CN-SP-Int commits in a few hours [18:09:23] :D [18:09:27] commit early, commit often [18:09:41] Heh [18:09:52] There's also something called spamming mediawiki-cvs [18:10:40] that's what she said! [18:10:59] What who said? [18:11:56] nudge nudge, wink wink [18:12:31] http://www.youtube.com/watch?v=DmPj5lqLrqQ [18:12:44] *AaronSchulz drools [18:14:10] ? [18:14:47] brion: then you just don't understand ;) [18:15:28] *AaronSchulz checks the Eagles' season record [18:15:59] 03mrzman * r43006 10/trunk/phase3/includes/specials/SpecialBlockip.php: pass user ID to Block::newFromDB in case of renames per comments on r42843. [18:16:24] 03(mod) query-continue incorrect for usercontribs - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=16195 (10roan.kattouw) [18:17:12] the eagles? like hotel california? [18:17:14] 03brion * r43007 10/trunk/extensions/ContributionReporting/ (7 files): Initial checkin of David Strauss's in-progress extension for donation reporting for the fundraiser [18:17:37] chuck: !rev 42974 works [18:18:51] brion: ok, now you are just pushing the buttons! [18:18:56] :D [18:19:52] bunch of sweaty men wearing tights running around throwing a ball, hey, whatever works for ya ;) [18:20:32] blasphemy! [18:20:37] yep, that does sound like The Eagles playing Hotel California in concert [18:20:45] :D [18:20:49] *Splarka seen twice [18:20:57] *brion marks some spam as spam [18:21:18] *AaronSchulz throws bug 11402 at brion [18:21:48] http://images1.wikia.nocookie.net/uncyclopedia/images/3/31/Spamburn.gif [18:22:29] *RoanKattouw throws bug 11430 at brion [18:24:39] 03ialex * r43008 10/trunk/extensions/ (2 files in 2 dirs): svn:eol-style native [18:24:56] 03siebrand * r43009 10/trunk/extensions/CentralNotice/ (12 files): [18:24:56] * put 'mk' translation back [18:24:56] * remove trailing whitespace and stylize.php [18:27:47] brion: any reason why ContributionReporting has 2 i18n files with 1 message each (and not one with the 2 combined -> mind if I merge them?) [18:29:16] siebrand: because david wrote em funny ;) [18:29:17] moment [18:29:36] FiloSottile: There's no API for anything CentralAuth-related yet. I'll probably write some of those modules soon [18:29:53] 03brion * r43010 10/trunk/extensions/ContributionReporting/ContributionReporting.php: cleanup a little [18:30:01] siebrand: if you can switch those to proper special page alias system that'd be awesome [18:30:53] brion: no problem. [18:30:57] thanks! [18:31:26] http://dev.donate.wikimedia.org/index.php/Special:Contribution_history <- ain't it purty though? [18:31:39] *brion goes to add a hack for yen to show without decimals [18:32:16] hmm, db error? :) [18:32:30] brion: are the i18n of centralnotice important to translate or for WMF staff only relevant? [18:33:02] Raymond__: there's L10n UI included. [18:33:10] Raymond__: not super relevant, but it wouldn't hurt. translators will (or can) use that interface to preview and update translations of the notice texts [18:33:32] ok thanks [18:34:56] 03(NEW) GlobalBlocking needs list=globalblocks API module - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=16214 15enhancement; normal; MediaWiki extensions: GlobalBlocking; (roan.kattouw) [18:35:38] 03(mod) GlobalBlocking needs API modules - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=16214 summary; +comment (10roan.kattouw) [18:36:28] 14(INVALID) Group 'emailconfirmed' not set after confirm link visited - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=16211 +comment (10Platonides) [18:36:55] 03brion * r43011 10/trunk/extensions/ContributionReporting/ContributionHistory_body.php: format yen without decimal point -- you can't break yen into smaller pieces :) [18:36:57] 03(FIXED) Protecting against creation protects expiring on the epoch - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=16207 (10roan.kattouw) [18:39:21] siebrand: are you making any other i18n changes to ContributionHistory? if so i'll hold off more refactoring for a little :) [18:39:52] brion: yeah, of possible stay out of there for another 10 mins or so. [18:39:57] if, even [18:41:00] :D [18:42:05] Hmm, this is fun. I'm writing list=globalblocks now and trying to figure out a module prefix, but there's this rule that prefixes can't start with a g ... [18:43:57] heh [18:43:57] name them sulblocks? or centralblocks? [18:44:13] No the name is OK [18:44:25] sure, but counterintuitive if the prefix doesn't match [18:44:33] But you can't call your parameters gbprop because that creates potential naming conflicts [18:44:38] wha? [18:44:38] So I settled on bgprop [18:44:46] "g is for generator, that's good enough for me" [18:44:51] urghl [18:46:28] *RoanKattouw is confused about the globalblocks table [18:46:53] How does this stuff work? Is it just IP blocking that's global? [18:52:00] brion: any change of getting to bug 11402 soon? Also you said you'd review bug 15649 after lunch? [18:52:06] brion: done [18:52:08] 03siebrand * r43012 10/trunk/extensions/ContributionReporting/ (7 files): [18:52:08] * delay message loading [18:52:08] * use $wgExtensionAliasesFiles for special page aliases [18:52:08] * merge 2 i18n files to one [18:52:08] * fix indentation, remove trailing whitespace and stylize.php [18:52:25] AaronSchulz: if it's not related to the fundraiser i'm unlikely to touch it for at least a few days [18:53:27] siebrand: woohoo [18:53:54] brion: it's just a trick I've done before :) I should automate it somehow :) [18:53:56] ialex: typo: handler comparaison [18:54:15] 03(mod) new history compression class - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=2310 +comment (10JSchulz_4587) [18:55:39] Nikerabbit: fixed, thanks! :) [18:57:42] 03(mod) new history compression class - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=2310 (10brion) [18:58:06] ialex: hmm [18:58:22] siebrand: why'd you remove the tripe-slash doc comments? [18:58:28] i don't think doxygen will pick them up anymore :( [18:58:33] *triple [18:58:54] brion: hmm, that's doxygen? I didn't know... [18:58:59] :D [18:59:05] or is it [18:59:08] fuck if i know [18:59:12] *brion shrugs [18:59:14] seemed familiar anyway [18:59:28] brion: stylize.php made "///" -> "// /", so I just "s!///!// /"-ed. [18:59:35] heh [18:59:38] no wories :D [18:59:47] brion: haven't seen it anywhere else if that helps... [19:00:16] http://dev.donate.wikimedia.org/index.php/Special:ContributionHistory [19:00:18] whee [19:00:37] sweet [19:00:39] no peeing in the chat room [19:01:15] ialex: is the default right needed to use configure bureaucrat? [19:01:59] Nikerabbit: I don't understand [19:02:02] brion: are you gonna have threshhold filters again, to block donations (by default) under $5 or such? and one to show only donations with comments? [19:04:27] ialex: which users can use it? [19:05:14] Nikerabbit: bureaucrat for Special:Configure and Special:Extensions, and sysop for Special:ViewConfig [19:06:31] Wow. Look at the guy who donated 3 yen... [19:06:49] 3 Japanese yen = 0.0238686647 Euros [19:06:51] LOL [19:07:03] I didn't know the yen was worth *that* little [19:07:10] 3? i saw lot of people donating 1 [19:07:13] ;) [19:07:26] Well a threshold of like 100 yen makes sense [19:08:27] Roan: previous years have seen ED/chan type vandals donate 1 yen many times over, with comments, for Lulz [19:09:16] a default threshhold keeps them somewhat suppressed, but if they were completely filtered they'd stop, which is a net loss ^_^ [19:15:41] 03brion * r43013 10/trunk/extensions/ContributionReporting/ (ContributionHistory_body.php ContributionReporting.i18n.php): i18n, add time to the date so we don't have an entire page of the same date all the time [19:15:57] there's also the issue of crackers testing their stolen CC numbers with small donations to non-profits such as WMF [19:16:10] RoanKattouw: yen doesn't use decimal portions; it's like counting all your dollar or euro amounts in cents :) [19:16:46] brion: I got that when I found out 1 yen is worth less than one eurocent [19:17:31] Which is why I said a 100 yen threshold would make sense [19:17:36] hehe [19:17:57] I should donate all my pennies [19:19:17] mmm, well, yen historically had sub-units, while cents _were_ subunits, which is why you can have fractional/decimal cents [19:20:46] mils! [19:21:18] http://en.wikipedia.org/wiki/Fils_(currency) [19:21:27] We're going to Sweden around Christmas, and it turns out Swedish crowns do have subunits: 100 öre to the crown. The pointlessness of the whole matter is that the smallest coin is 50 öre [19:21:39] Which is basically due to the fact that 1 crown is worth shit [19:21:50] roan: indeed, but not until somewhatg recently [19:22:00] everyone there has a coin jar with 10 and 1 ore coins [19:22:12] You're a Swede, then? [19:22:15] ^somewhat = this century (time moves slow there) [19:22:21] no, just been there several times [19:22:39] Yeah well over here people still have pre-euro currency in their jars [19:23:14] :o [19:23:27] 03catrope * r43014 10/trunk/phase3/includes/api/ApiQueryBlocks.php: API: Clean up ApiQueryBlocks, which was a stylistic mess. I can't believe I actually wrote that code (I was young and foolish, I guess :P ) [19:23:38] Well the euro was only introduced quite recently (2002) ... [19:24:01] I still have my 1 penni coin [19:24:15] > In 1961, the 1 and 2 �re and 2 kronor coins ceased production and the size of the 5 �re coin was reduced. [19:24:37] yeah, i never had any use for öre when i was there [19:24:39] what do you mean "whas"? heh [19:24:44] er "was" [19:24:46] just like i have no use for anything smaller than a quarter at home :) [19:25:04] *Splarka hands brion a 20 cent piece [19:29:39] 03siebrand * r43015 10/trunk/extensions/ContributionReporting/ (ContributionReporting.i18n.php ContributionReporting.php): Add extension credits and description message [19:30:21] 03siebrand * r43016 10/trunk/extensions/ContributionReporting/ContributionReporting.php: Oops. Fix syntax error [19:33:00] Splarka: wgNotice js var will be set when there's an active centralnotice [19:33:06] it'll contain the html of the notice :) [19:33:54] kk, but going live in ... 2 days now? [19:34:07] if they enable the JS on en.wp it'll take ~31 days (worst case) to be operative [19:34:37] you do realize we can flush things from the shared cache, right? [19:34:50] client side cache [19:35:58] unless the actual URI is changed, I don't think anything can be done about it even with a change to the headers [19:36:06] (at least for ~31 days) [19:36:06] well, life sucks eh? [19:36:10] http://web-sniffer.net/?url=http%3A%2F%2Fen.wikipedia.org%2Fw%2Findex.php%3Ftitle%3D-%26action%3Draw%26smaxage%3D0%26gen%3Djs%26useskin%3Dmonobook&submit=Submit&http=1.1&type=GET&uak=0 [19:36:13] yes, yes it does [19:36:27] we'll just have to survive [19:36:30] 03siebrand * r43017 10/trunk/extensions/Translate/ (aliases.txt groups/mediawiki-defines.txt): Add support to Translate for ContributionReporting [19:36:39] heh [19:37:17] right, so... [19:37:17] if(wgUserName == null && skin == 'monobook') addOnloadHook((function (){ [19:37:18] if(window.siteNoticeValue || window.wgNotice) return; [19:37:33] that should abort the script if either of those globals is set (local sitenotice, or global sitenotice) [19:37:49] it has to be done after onload because the centralnotice is in the (or last I saw anyways) [19:38:37] 03siebrand * r43018 10/trunk/extensions/Translate/groups/MediaWikiExtensions.php: Update Wikimedia meta message group [19:38:57] *Splarka gives brion 8 hours off, get some damn food and sleep [19:55:32] 03siebrand * r43019 10/trunk/extensions/uniwiki/Authors/ (Authors.i18n.php Authors.php): [19:55:32] Update for uniwiki/Authors: [19:55:32] * standard i18n [19:55:32] * complete extension credits [19:55:32] * update indentation, remove trailing whitespace, stylize.php [20:00:54] 03(mod) Include a radio button on the installation screen allowing the wiki' s contents to be licensed from the get-go as public domain - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=15125 (10nathanlarson3141) [20:10:39] looking doing stuff via the api and have found the perl module MediaWiki::API... In the docs of the module, it has 'cmlimit=max', but when I try to use that the error response says it must be a number... after looking into it, its defaulted to 500 for a user and 5000 for a bot. I'm writing a bot and was wondering now to configure mediawiki to have no limit [20:11:46] limits are good, they keep php from timing out [20:12:29] ok cool so how do I break up the requests [20:12:32] Or your browser from crashing. [20:12:38] say give me first 500 [20:12:40] query-continue [20:12:43] The API has a continue function. [20:12:47] cool [20:12:48] thanks [20:12:55] i'm doing this from the command line [20:13:15] for example: [20:13:17] http://en.wikipedia.org/w/api.php?action=query&list=categorymembers&cmtitle=Category:Physics [20:13:22] [20:13:22] [20:13:22] [20:13:23] going through all our code, stripping the docs, and putting them in the wiki [20:13:23] etc etc [20:13:29] [20:13:29] [20:13:35] so you pass on the next query... [20:13:46] http://en.wikipedia.org/w/api.php?action=query&list=categorymembers&cmtitle=Category:Physics&cmcontinue=Acoustic%20cloak| [20:14:05] 03(mod) Group 'emailconfirmed' not set after confirm link visited - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=16211 +comment (10huck.finn) [20:14:06] note it is sorted by sortkey rather than pagename [20:14:20] all so is that the last title name? .. so sortkey [20:14:26] beat me to it ;) [20:14:29] well, it is the _next_ [20:14:40] if you have 10, the continue is the 11th [20:14:46] cool [20:14:55] you can do it by timestamp too (which is less meaningful) [20:15:04] yeah i could see that [20:15:13] category members are fickle [20:15:30] i see there is a different limit for bots [20:15:47] is there just a flag in the database saying a user is a bot [20:16:00] *Splarka nods [20:16:04] sweet [20:16:12] Splarka++ # thanks for the help [20:16:19] MZMcBride++ # yo too :D [20:16:22] rar [20:16:31] you even [20:16:46] :-) [20:16:47] *Splarka can't see where the 500/5000 limit is set hmm... [20:16:55] might be hard coded like the 50/500 limits are [20:16:58] bah, quitters [20:17:17] $wgAPIMaxDBRows = 5000; ? [20:17:31] nah [20:17:33] const LIMIT_BIG1 = 500; // Fast query, std user limit [20:17:35] const LIMIT_BIG2 = 5000; // Fast query, bot/sysop limit [20:17:46] api/ApiBase.php [20:18:00] yah, that'd be it [20:18:13] http://google.com/codesearch?num=100&q=package%3Asvn.wikimedia.org%2Fsvnroot%2Fmediawiki%2Ftrunk+file%3Aphase3%2Fincludes%2Fapi%2F+500+5000 [20:18:16] *Splarka just saw it too [20:19:52] 03brion * r43020 10/trunk/extensions/CentralNotice/CentralNotice.php: [20:19:52] restore $wgCentralNoticeLoader option to disable client-end loading stuff [20:19:52] make $wgNoticeInfrastructure work properly w/ new classes [20:21:57] $logPage->addEntry( 'foo', $this->getTitle(), $logComment ); [20:22:06] is there any reason why that above code wouldn't work? [20:22:10] That's what she said. [20:22:21] >_> [20:22:39] 03brion * r43021 10/trunk/extensions/ContributionReporting/ContributionHistory_body.php: bump uggo hardcoded template names [20:22:49] the debug log doesn't report any errors [20:22:53] MZMcBride: zing! [20:22:59] but nothing shows up in the "foo" log [20:25:13] ? [20:26:50] http://p.defau.lt/?BUxud5jzl6nHF2_TT8lCbQ [20:27:04] that's the code I used to create the log thingy [20:27:11] and special:log/foo shows up [20:27:20] but nothing gets added to it [20:28:22] 03rotem * r43022 10/trunk/extensions/ (2 files in 2 dirs): Localization update for he, and several fixes in en. [20:31:41] 03brion * r43023 10/trunk/extensions/CentralNotice/CentralNotice.php: need to pull up perms for centralnotice :P [20:32:54] ok then [20:33:18] !wg LogTypes [20:33:18] --mwbot-- http://www.mediawiki.org/wiki/Manual:%24wgLogTypes [20:34:55] hmm [20:36:35] is there a straightforward way of converting a MediaWiki site over to a Google Sites site? [20:36:42] Basically you need to do $wgLogTypes[] = 'foo'; in the extension setup file [20:36:55] the log type that I added is exactly 10 bytes [20:37:09] so that shouldn't be the problem according to that page [20:37:34] Hi [20:37:39] Is it possible to save variables on one wikipage and use them on another? [20:37:58] RoanKattouw_away: I do :P [20:38:05] http://p.defau.lt/?BUxud5jzl6nHF2_TT8lCbQ [20:38:15] that has all the code in the extension's setup file [20:39:27] phobie: no. not beyond what templates do. semanti mediawiki can do stuff like that though [20:39:28] !smq [20:39:28] --mwbot-- I don't know anything about "smq". [20:39:31] err, [20:39:33] !smw [20:39:33] --mwbot-- SemanticMediaWiki is an extension that lets you conenct wiki pages with semantic relations. See and . Mailing lists are available at . [20:41:58] 03brion * r43024 10/trunk/extensions/CentralNotice/CentralNotice.php: drop the old purging hooks -- obsolete with the changes to loading infra [20:43:05] Duesentrieb: thanks, I found that page before, but hoped to do simple stuff without that... [20:44:04] phobie: there are other extensions that try to provide some of that functionality, but they are generally ugly hacks [20:44:13] what are you actually trying to achieve? [20:46:04] We want to save values for proposals on OpenStreetMap and use them on several places (i.e. [[Tag:key=value]], [[Map_features]] and [[Proposed_features/*]]) [20:46:07] 03siebrand * r43025 10/trunk/extensions/CentralNotice/CentralNotice.i18n.php: doesn't -> does not [20:46:35] 03(ASSIGNED) Include a radio button on the installation screen allowing the wiki' s contents to be licensed from the get-go as public domain - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=15125 (10nathanlarson3141) [20:46:47] Has anyone actually converted a MediaWiki site over to a Google Sites site? [20:47:26] Don't mess with dreagons [20:48:51] 03(NEW) Incubator mailinglist. - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=16215 15enhancement; normal; Wikimedia: Mailing lists; (Sterkebak) [20:49:09] 03(mod) Incubator mailinglist. - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=16215 (10Sterkebak) [20:51:01] 03(mod) Incubator mailinglist. - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=16215 (10soxred93) [20:52:20] Beyond cutting and pasting each individual page, is there a more automated and labour-saving way of taking content from a MediaWiki site over to a Google Sites site? [20:52:41] not that I am aware of [20:52:51] 03(mod) Incubator mailinglist. - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=16215 (10mohsensalek) [20:53:01] ah. Okay, thanks. [20:53:14] 03(mod) Incubator mailinglist. - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=16215 (10mohsensalek) [20:55:42] 03(mod) Incubator mailinglist. - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=16215 (10Sterkebak) [20:56:02] http://en.wikipedia.org/wiki/Image:Voterseyeview.jpg [20:56:08] brion: starting at that makes me angry [20:56:38] agh [20:56:53] 03(mod) Incubator mailinglist. - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=16215 (10mohsensalek) [20:58:40] "A later review of discarded ballots in Palm Beach County by the The Palm Beach Post showed that 5,330 votes were cast for the presumably rare cross-party combination of Gore and Buchanan, compared with only 1,631 for the equivalent cross-party combination of Bush and Buchanan. The number of votes that may have been mistakenly cast for Buchanan was well in excess of George W. Bush's... [20:58:42] ...certified margin of victory." [20:59:03] yay "butterfly ballot" [20:59:51] 03(mod) Incubator mailinglist. - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=16215 +comment (10cbrown1023) [21:01:44] 03(mod) Incubator mailinglist. - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=16215 +shell (10mattj) [21:01:56] so any idea why nothing is getting logged and it's not even mentioning any problems? [21:02:13] $log->addEntry actually returned true, even [21:03:33] 03(mod) Incubator mailinglist. - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=16215 (10mohsensalek) [21:04:11] 03siebrand * r43026 10/trunk/extensions/uniwiki/AutoCreateCategoryPages/ (AutoCreateCategoryPages.i18n.php AutoCreateCategoryPages.php): [21:04:11] Update for uniwiki/AutoCreateCategoryPages: [21:04:11] * standard i18n, change message IDs [21:04:11] * use wfMsgForContent in edit summaries and page creation [21:04:11] * complete extension credits [21:04:15] * update indentation, remove trailing whitespace, stylize.php [21:04:42] oh wait, I think I finally found the problem >_> [21:10:35] 03siebrand * r43027 10/trunk/extensions/uniwiki/CatBoxAtTop/ (CatBoxAtTop.i18n.php CatBoxAtTop.php style.css): [21:10:35] Update for uniwiki/CatBoxAtTop: [21:10:35] * complete extension credits [21:10:35] * update indentation, remove trailing whitespace, stylize.php [21:11:23] 03siebrand * r43028 10/trunk/extensions/uniwiki/CatBoxAtTop/CatBoxAtTop.php: Forgot stylize.php [21:18:07] Pinky: What was it then? [21:19:03] well this particular extension creates a new wiki [21:19:22] and so it was adding the log to the new wiki instead of the original wiki :P [21:19:38] still haven't quite fixed it to log it to the original wiki, but at least I know the problem [21:28:25] I think I'm gonna have to end up rewriting the saveContent function from LogPage.php in my extension so that I can make it use the right db [21:32:31] Just tweak it to take a $db parameter [21:32:47] *brion wonders how much is CZK 500.00 [21:33:10] RoanKattouw_away: would that get accepted into core though? [21:34:05] brion: http://www.google.com/search?&q=500+CZK+in+USD [21:34:08] Pinky, I wouldn't think so, it sounds more like a poor hack [21:34:18] There has got to be a better way to do it [21:34:22] Duesentrieb: thx ;) [21:34:26] brion: don't be lazy, czech! [21:34:35] it is "the czech is in the mail" [21:34:37] well rewriting the function works for me I guess [21:34:38] brion: you can say about google what you want, these guys know how to make handy tools :) [21:34:49] Would it be imperialistic to make all the donation figures have an extra column converting them to USD on Special:ContributionHistory? [21:34:49] :D [21:35:31] Pinky: Why don't you subclass LogPage? [21:35:36] MZMcBride: ��$ [21:35:57] More importantly, do we care? [21:36:49] Pinky: Override saveContent, in that function grab the db, use switchDB, call parent::saveContent(); then switchDB back. [21:37:19] 03midom * r43029 10/trunk/debs/wikimedia-mysql4/ (15 files in 3 dirs): Hello Mark ;) [21:38:59] 03tstarling * r43030 10/trunk/phase3/includes/HistoryBlob.php: Convert from old DiffHistoryBlob format. Untested, committing for test by Jelte. [21:40:57] Someone who can fix a bug in uniwiki/CatBoxAtTop/CatBoxAtTop.php? In lines 45-49 category links should be gathered. I only get "Array" as output... [21:41:01] I guess that works [21:42:02] 03siebrand * r43031 10/trunk/extensions/uniwiki/ (Authors/Authors.php UniwikiSettings.php): [21:42:02] * Fix a bug in Authors extension ($ missing) [21:42:02] * Comment out Authors extension. Method creditlink() appears to be missing. [21:42:02] Fatal error: Call to undefined function creditlink() in D:\mediawiki\phase3\extensions\uniwiki\Authors\Authors.php on line 83 [21:43:47] How do I make a skin compatible with the stuff in the default Sidebar message that inserts a SEARCH, TOOLBOX, LANGUAGES, etc? [21:44:15] 03(mod) Group 'emailconfirmed' not set after confirm link visited - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=16211 (10Platonides) [21:47:00] 03mrzman * r43032 10/trunk/phase3/includes/specials/SpecialBlockip.php: Fix rangeblock detection [21:47:07] Dantman: the $dbw in function saveContent is a local variable though [21:47:08] 03siebrand * r43033 10/trunk/extensions/uniwiki/CreatePage/ (4 files): (log message trimmed) [21:47:08] Update for uniwiki/CreatePage: [21:47:08] * standard special page addition [21:47:08] * split off special page body into a class file [21:47:08] * standard i18n for UI messages [21:47:11] * add i18n for special page aliases [21:47:12] not like $this->dbw [21:47:13] * complete extension credits [21:47:45] Pinky: Doesn't matter... Calling wfGetDB( DB_MASTER ); should always return the same object [21:50:56] so I'm doing $dbw = wfGetDB( DB_MASTER); $dbw->selectDB($therightdb); parent::saveContent(); ? [21:51:55] 03siebrand * r43034 10/trunk/extensions/uniwiki/CreatePage/CreatePage.php: Update documentation URL [21:52:26] Pinky: Yes, however remember to selectDB back to the original after [21:54:25] mmk [21:55:22] 03siebrand * r43035 10/trunk/extensions/uniwiki/ (4 files in 4 dirs): Extension credits URL updates [21:58:30] 03siebrand * r43036 10/trunk/extensions/uniwiki/CssHooks/ (CssHooks.i18n.php CssHooks.php): [21:58:30] Update for uniwiki/CssHooks: [21:58:30] * complete extension credits [21:58:30] * update indentation, remove trailing whitespace, stylize.php [22:07:05] 03brion * r43037 10/trunk/extensions/CentralNotice/ (SpecialCentralNotice.php SpecialNoticeTemplate.php): Allow non-priveledged visitors to view the notice campaigns and templates, just not edit them [22:13:51] cn you encrypt the mysql password in LocalsETTINGS.php? [22:14:30] can* [22:16:33] Is it possible at all [22:16:35] ? [22:16:40] gewt, I don't think so [22:16:44] gewt: No [22:16:46] :< [22:16:56] any encryption must be reversable to be used by mysql functions [22:17:10] just make that file unreadable by others [22:17:21] or include a different file containing the passwords [22:18:09] gewt: Why do you need to? [22:18:18] hm [22:18:26] i guess i wouldn't need to [22:19:05] 03tomasz * r43038 10/trunk/extensions/CentralNotice/ (3 files): adding spacing when no content is present [22:19:32] If you are a little security concious, the best thing you can do is chown the file to the php user, and chmod it to r-------- [22:19:44] might need a u+x [22:19:52] 03catrope * r43039 10/trunk/phase3/includes/api/ApiQueryBlocks.php: API: Followup on r43014: clean up some more [22:25:18] 03catrope * r43040 10/trunk/extensions/GlobalBlocking/ (ApiGlobalBlocks.php GlobalBlocking.php): GlobalBlocking: Add list=globalblocks API module, which is basically just a tweaked rip-off of list=blocks [22:31:08] 03brion * r43041 10/trunk/phase3/includes/GlobalFunctions.php: Update doc comment for wfArrayMerge() to note the behavior difference between array_merge() with string keys and the + operator on arrays. Now with less cursing. :) [22:32:07] more cursing [22:33:14] brion: now with 80% less cursing! [22:33:30] did it move to basic cable? [22:33:41] did you resize it to fit 4:3 and cut it for time too? [22:34:24] 03siebrand * r43042 10/trunk/extensions/Collection/Collection.i18n.php: Fix message [22:36:17] this comment has been reformatted to fit your screen [22:38:25] Thank God MySQL doesn't do that kind of stuff [22:38:45] *RoanKattouw imagines a world in which MySQL DBs are truncated to fit on one screen [22:39:01] 03(NEW) Wikitext parser inside tags partially functional - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=16216 normal; normal; MediaWiki extensions: Cite; (holek.n) [22:39:48] 03siebrand * r43043 10/trunk/extensions/AbuseFilter/AbuseFilter.i18n.php: Remove period from descriptionmsg [22:41:51] 14(DUP) Wikitext parser inside tags partially functional - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=16216 +comment (10roan.kattouw) [22:41:53] 03(mod) template substitution does not work inside tags - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=6157 +comment (10roan.kattouw) [22:43:50] 03siebrand * r43044 10/trunk/extensions/CentralNotice/CentralNotice.i18n.php: Fix message [22:47:03] 03(mod) Pre-save transform skips extensions using wikitext (gallery, references, pipe trick, subst, signatures) - 10http://bugzilla.wikimedia.org/show_bug.cgi?id=2700 (10holek.n) [22:47:23] 03(mod) Add Commons as an import source for Meta - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=16136 +shell (10mrzmanwiki) [22:48:02] 03brion * r43045 10/trunk/extensions/ContributionReporting/ContributionReporting.php: force utf8 connection [22:48:03] hi. is there a template parser with which i can make the template different depending on the day? [22:48:31] DeFender1031: Use {{CURRENTDAY}} and friends in combination with {{#if:}} [22:48:37] !parserfunctions [22:48:37] --mwbot-- "Parser functions" are a way to extend the wiki syntax. ParserFunctions is an extension that provides the basic set of parser functions (you have to install it separately!). For help using parser functions, please see . For details about the extension, see . [22:51:22] that help page doesn't seem to mention {{CURRENTDAY}} [22:51:30] !magicwords [22:51:30] --mwbot-- For more information about creating magic words and their inner workings, see . For a list of magic words, please see . [22:53:23] thanks [22:53:29] this'll be a great help [22:54:04] i have a page with lots of subpages which are categorised. Is it possible to "strip" the page name from the article list in the category so they show as "subpage" rather than "page/subpage" [22:54:06] ? [22:55:09] exobuzz: Not really. You can categorize them under the S(ubpage) rather than the P(age) by using [[Category:Foo|{{SUBPAGENAME}}]] (IIRC) [22:55:31] yeh. i use the pipe to fix the sort order [22:55:51] but it looks crappy still. [22:56:37] How to save variables on one wikipage and use them on another? [22:56:40] I know. But you can't really get around that [22:56:44] 03brion * r43046 10/trunk/extensions/ContributionReporting/ContributionHistory_body.php: [22:56:44] * proper num formatting for currency [22:56:44] * add USD conversion as a hover title for the lazy like me :D [22:56:44] !smw | phobie [22:56:44] --mwbot-- phobie: SemanticMediaWiki is an extension that lets you conenct wiki pages with semantic relations. See and . Mailing lists are available at . [22:58:57] RoanKattouw: I know, but how to fix my problem with that? [22:59:01] mwbot said semantic, get ... it [22:59:12] I searched on that side but did not find a resolution [22:59:17] phobie: I don't know. But SMW should be able to do such stuff [22:59:19] *Splarka lynches mwbot [22:59:55] phobie: There's an example about the population of some city (Munich?) IIRC, which looks a lot like what you want [23:01:00] http://semantic-mediawiki.org/wiki/Munich [23:01:16] The values set there are used on the same page and not on an other one [23:01:30] RoanKattouw: i guess the fix is in this case, not to use subpages and to put them in the main namespace. they probably should be really, but there will be a lot of pages to sort out conflicting names [23:01:33] They *should* be available on another page too [23:01:39] back in my day, if we wanted variables we made templates for them [23:01:43] exobuzz: Create another ns? [23:01:52] phobie: they can be used on any page though [23:01:59] RoanKattouw, what about hebrew calendar days? is there a way to get those? [23:01:59] {{Population of Munich}} would be [[Template:Population of Munich]] [23:02:13] phobie: smw is powerful. you calalso query it for all cities with a pupulation above x, etc [23:02:23] RoanKattouw: I guess that could work too.. although the subject matter fits into the main namespace, but when it was made neither me or the guy who made them had much wiki experience [23:03:04] phobie: See also http://semantic-mediawiki.org/wiki/Germany_(inverse_example) , the solution to your problem involves {{#ask:}} [23:03:31] Splarka: You'd be creating tens of thousands of templates that way. Since that's not what people want, they wrote SMW [23:03:31] ok [23:04:27] Roan: back in my day people didn't need that much info! damn kids, get off [[Lawn]] [23:04:51] phobie: http://semantic-mediawiki.org/wiki/Help:Inline_queries seems to provide more generic information [23:05:29] {{City population|Munich}} -> {{#switch:1|Munich=1,234,567|Berlin=12,345,678|Vastenhaufferschliedel=12|#default=Error, no city specified}} [23:05:31] *Splarka hides [23:05:48] *RoanKattouw finds Splarka and kills him [23:06:34] Mmm... "Damn kids, get off my [[Location::Lawn]]!" [23:06:43] Splarka: that could be a way to fix my problem [23:06:53] *Splarka bleeds [23:07:05] phobie: If you have a *very* limited number of cities and know that won't change, it could work [23:07:19] But it becomes *really* impractical with more than a few cities [23:07:27] *Splarka nods [23:07:27] hm [23:07:29] that it does [23:07:30] And the location where the city's pop. is stored is counter-intuitive [23:07:32] you have to scale it sensibly [23:07:41] we have several values... [23:07:41] heh [23:07:47] Is thgere any magic word to get hebrew date? [23:07:48] well, Roan... there is another way, even uglier [23:07:50] SMW lets you store the pop. of Berlin in [[Berlin]], not in [[Template:City population]] along with all other city pop.s [23:07:56] Splarka: Enlighten me [23:07:59] DeFender1031: what is the year offset, 3000 or so? [23:08:20] Roan: for each City article, have Population: 123456789 [23:08:21] DeFender1031: Maybe something involving {{#time:}} ? [23:08:31] then in the pop template: {{{{{1}}}}} [23:08:50] *RoanKattouw slams Splarka's head into a wall [23:09:01] *Splarka has a spare, but it stings, ow [23:09:09] !attack | RoanKattouw [23:09:09] --mwbot-- RoanKattouw: I don't know anything about "attack". You might try: !phpinfo [23:09:14] :( [23:09:36] lol phpinfo [23:09:38] rotfl [23:09:43] RoanKattouw, to do that i'd have to re-write the entire jewish date algorithm in parserfunction. [23:09:46] *RoanKattouw checks why that is [23:09:52] !phpinfo [23:09:52] --mwbot-- create a file called info.php in your wiki's installation directory. It should have the following content: ; Visit it with your browser to see information about your php installation. If it doesn't work, check file permissions. Remove the file when done, it makes life easier for attackers. [23:09:59] ah [23:10:00] lol "attackers" [23:10:01] Splarka, it's more than year offsets [23:10:02] heh [23:10:11] !fnord [23:10:11] --mwbot-- I don't know anything about "fnord". [23:10:17] aww... no suggestions? [23:10:20] DeFender1031: There has to be an easier way. You should ask some i18n guys like siebrand or Nikerabbit [23:10:38] il8n? whuzzat? [23:10:39] Hmmm... one more [23:10:43] !splarka [23:10:43] --mwbot-- I don't know anything about "splarka". You might try: !op !ops [23:10:54] omg dont do it [23:10:58] it mass pings [23:11:01] DeFender1031: internationalization. Stuff like translation, but also foreign calenders [23:11:16] btw, RoanKattouw, i know PHP has built in hebrew date stuff... doesn't seem to have been written into MW yet though [23:11:33] *Soxred trouts mwbot [23:11:38] Dantman: why would you do that?! [23:11:43] Well maybe it has. There's been some stuff to do over another foreign calender (Arabic?) a while ago [23:11:56] But you should really ask siebrand or Nikerabbit [23:11:59] *RoanKattouw hopes they show up [23:12:03] if you !oops (heh) d*m*s usually will kickban someone, so pray there is someone more kickbannable than you when typing it [23:12:16] afaik there's Hebrew calendar in MediaWiki. No idea how to use it, though [23:12:19] so, siebrand, Nikerabbit, DO you guys have a suggestion for me as to how to do a hebrew date without rewriting the algorithm in parserfunction? [23:12:28] Hey Splarka how's this for pinging people: !Splarka_I'm_poking_you [23:12:39] may want to ask Rotemliss. [23:12:40] siebrand, any idea where to look even? [23:12:44] buy me dinner first [23:12:45] Splarka: I think I will create one Template for every city like: {{cityvalues/{{{city}}}|population}} {{cityvalues/{{{city}}}|lat}} ... [23:12:53] RoanKattouw: Why not just !poke? [23:12:54] Is that a good idea? [23:13:00] Rotem Liss... sounds like an israeli name... [23:13:02] Yeah probably better [23:13:02] DeFender1031: nope. Well, work your way back from the messages file... [23:13:08] phobie: bit messy [23:13:19] siebrand, what messages file? [23:13:31] MessagesXX.php where XX is your language code [23:13:36] 'hebrew-calendar-m1' => 'Tishrei' [23:13:37] In the languages/messages/ directory [23:13:39] "And so was the world before the invention of Semantic MediaWiki." [23:14:54] "and so was the world after the great semantic cleansing" [23:15:09] ah. i don't have access to the source code on the particular wiki i'm actually trying to use this on. problematic a bit. i was hoping there'd be a way to do it just with parserfunctions [23:15:40] Splarka: what is better, if you have a lot of cities and several sites per city to use that values? [23:15:41] DeFender1031: The source code is public of course [23:15:59] http://svn.wikimedia.org/viewvc/mediawiki/trunk/phase3/ [23:16:12] true, but if it requires any modification, which is what i thought you were saying, then i'm screwed [23:16:51] DeFender1031: it's used in langauges/Language.php [23:17:56] okay, is there a way to translate it to parser functions, or is it only used internally by the php? [23:18:27] There's always a way to add a parser function wrapper for it [23:18:37] But maybe it can be used somewhere somehow [23:19:03] phobie: usually best just to duplicate the data whenever it is needed... but you must have a good reason to not want to.. how many times would a typical data bit be used? on how many pages? [23:19:28] RoanKattouw, a way without modifyingthe source? i somehow doubt it [23:19:57] *Splarka blinks [23:20:05] DeFender1031: did you even look at #time ? [23:20:14] http://meta.wikimedia.org/wiki/Help:ParserFunctions#.23time: [23:20:25] xjj The day of the month in Hebrew calendar [23:20:25] xjF The full month name in Hebrew calendar [23:20:25] xjx Genitive form of the month name in Hebrew calendar [23:20:32] Splarka: It is for http://wiki.openstreetmap.org/index.php/Proposed_features [23:20:39] 03tstarling * r43047 10/trunk/phase3/ (7 files in 3 dirs): (log message trimmed) [23:20:39] * Rewrote wfArrayMerge() in terms of array plus [23:20:39] * Two callers of wfArrayMerge() were bugs, both assuming strange and complex behaviour in wfArrayMerge() which has never been present or documented. [23:20:39] * Introduced wfMergeErrorArrays() to remove duplicates from merged error arrays, e.g. from getUserPermissionsErrors(). [23:20:42] * Rewrote the remaining callers of wfArrayMerge() to use array plus. It makes [23:20:44] the code clearer, assuming the reader knows more about basic PHP operators than [23:20:46] GlobalFunctions.php. Considering the two bugs discussed above, this seems like a [23:20:59] Splarka: so per proposal the values will be used on 3 sites [23:21:26] Splarka, time seems to be static [23:21:29] oooh [23:21:30] 3 pages or 3 different wikis ? [23:21:33] Splarka: and we have more than 50 proposals [23:21:40] DeFender1031: second parameter is the time, if not _now_ [23:21:40] unless i pass it {{CURRENTTIME}} [23:21:41] no wikipages [23:21:53] just omit the second parameter for now [23:22:12] won't that just continually show the time the page was saved? [23:22:25] depends [23:22:27] that's how i understood it anyway [23:22:28] DeFender1031: No, it'll update to the current time every hour or so [23:22:49] if you have a very good cache, it will be when it was last saved, or purged, or server-side cache expired [23:22:57] if you have no cache, every view will update it [23:22:59] ah [23:23:09] they do have a cache IIRC [23:23:12] but that is true of any parserfunction [23:23:15] Splarka: Pages containing time-dependant stuff are purged every hour [23:23:19] ah [23:23:22] Splarka, pages containing {{CURRENTX}} are set with a low expiry time, like an hour or something. [23:23:23] Splarka: i.e. use the values of Smoothness on [[Proposed_features]] and [[Proposed_features/Smoothness]] [23:23:25] roan: since when? [23:23:27] You need to use {{CURRENTTIME}} explicitly to enable that behavior though [23:23:31] so time, using the heb date format and the parser functions... [23:23:32] Simetrical: that's new innit? [23:23:35] Splarka, no. [23:23:44] yay TimStarling [23:23:50] It's *relatively* new. Like, a few months [23:23:51] i can do it [23:23:57] Simetrical: does #time do that with no second parameter? [23:24:00] brion: fair assumption? [23:24:04] Splarka: No. It [23:24:11] 's based on magic words, not parserfunctions [23:24:19] Simetrical: why did the main pages always need manual purges at midnight? [23:24:20] (IIRC) [23:24:39] Splarka: They shouldn't need that any more now [23:24:59] well, anyways, I was accurate, assuming he just used #time ^_^ [23:25:03] Unless you're pedantic about stuff and think it's a big deal if it takes till like 1 AM for the date to update [23:25:26] so how often will #time update if i'm not using a parameter? real-tim, or cache-time? [23:25:31] Yes, but using {{CURRENTTIME}} explicitly with #time will trigger the low expiry cache thingy [23:25:36] personally I just remember those two words from the PHP manual as a mnemonic: "NOT overwritten" [23:25:39] DeFender1031: Never, only on purge [23:25:50] Splarka: you ment it is good to duplicate the data. But not if the values changes and needs to be kept in sync [23:25:55] Unless you happen to have another time-dependent magic word on the page [23:25:58] and if i pass it {{CURRENTTIME}}? [23:25:59] 'never' is a long time [23:26:18] phobie: you might consider semantic then [23:26:19] DeFender1031: If you use {{CURRENTTIME}}. it'll update every hour at least [23:26:24] sweet [23:26:25] *Splarka sighs at saying that [23:26:36] maybe we need to invent rhymes for PHP irregularities [23:26:44] like the ones that exist for english spelling [23:26:56] I before E except after C and stuff like that [23:27:15] RoanKattouw, where's the length of the interval stored? Is it configurable? [23:27:23] It's hardcoded to 3600 IIRC [23:27:30] Ask brion, he implemented it [23:27:31] RoanKattouw, remember where? [23:27:37] why not add that for #time? [23:27:49] Simetrical: Remember what where? [23:27:52] (just with no second parameter, that is) [23:27:55] RoanKattouw, where it is in the code. [23:28:10] Splarka: Because it's not that advanced. It's just about the presence of a list of magic words on the page [23:28:15] wha? [23:28:16] Pfff [23:28:18] MagicWord.php ? [23:28:25] i don't remember that [23:28:29] brion: Short expiry time (1h) for {{CURRENTTIME}} and friends [23:28:34] I *seem* to remember you added that [23:28:38] Although I'm not sure [23:28:45] I think it was Tim [23:28:56] Array of caching hints for ParserCache */ [23:28:56] static public $mCacheTTLs = array ( [23:29:03] sounds like something i'd have done, or at leat advocated [23:29:04] Domas actually [23:29:06] no idea if it works ;) [23:29:08] !rev 29511 [23:29:08] --mwbot-- http://www.mediawiki.org/wiki/Special:Code/MediaWiki/29511 [23:29:09] ahh, each one gets a number, but none less than 3600 [23:30:08] so let me get this straight... basically i do {{#ifeq:{{#time:xjj|{{CURRENTTIME}}}}|somehebrewmonth|stuff i want special}} and that should do it? [23:30:24] If xjj does what you expect, yes [23:30:37] ... [23:30:39] DeFender1031, I'm pretty sure {{CURRENTTIME}} is superfluous there. [23:30:43] i'm having an issue with skins [23:30:48] no matter what skin i use [23:30:48] well i'll just have to play with xjj and see if that's true [23:30:50] Simetrical: not if he wants the de-cache hint [23:30:51] i don't get the skin [23:30:52] 03siebrand * r43048 10/trunk/phase3/ (30 files in 3 dirs): Localisation updates for core messages from Betawiki (2008-11-02 00:14 CET) [23:30:59] It's a wiki I am hosting for someone [23:31:09] Simetrical, we already went over that, without it it wouldn't update as often [23:31:35] *Simetrical mutters about inelegant, expensive extra calls to strtotime() [23:32:05] *Splarka wonders if parserfunction extensions could push to $mCacheTTLs externally [23:32:13] maybe we should have a quiz that people have to complete before they get commit access [23:32:17] Simetrical, you give me a better way to have it not cache out and i'll listen [23:32:24] DeFender1031, your way is find, never mind me. [23:32:25] "what are the 6 things that are false when converted to boolean?" [23:32:43] Hmm [23:32:45] 0 [23:32:47] '0' [23:32:48] NULL [23:32:50] '' [23:32:52] array() [23:32:52] which are true: '0' == 0, 0 == '', '0' == '' [23:32:53] TimStarling, "0", "0.0", "0.00", "0.000", "0.0000", "0.000000"? [23:32:58] and false [23:33:00] Oh and 0.0 of course [23:33:10] well done Roan [23:33:18] Simetrical: nope [23:33:19] *Simetrical fails [23:33:24] "0.0" is true, "00" is true [23:33:27] TimStarling: Isn't that seven? 0 0.0 '' '0' NULL array() false [23:33:30] *Simetrical just discovered that [23:33:34] any ideas why NO themes work? [23:33:43] gewt: You mean skins? [23:33:48] yeha [23:33:52] RoanKattouw: yes [23:33:54] none fo them work at all [23:33:59] possibly permissions error? [23:34:04] I was counting the number of bullet points in the manual, but one of them has two things [23:34:09] lol [23:34:13] '0' and '' ? [23:34:17] there's also SimpleXML but nobody cares about that [23:34:18] Splarka: Ok, if I can get the admin to install semantic... Can you give me a tip on how to use it for my problem? I only found php-code on the semantic-site but the variables must be usereditable... [23:34:38] no idea, I just hear people complain about it, there should be a channel here for it somewhere [23:34:52] *AaronSchulz digs through the manual [23:35:19] phobie: Assuming the population example: set [[population::1259677]] on one page [23:35:22] permissions are entirely messed up... [23:35:28] Then use {{#ask:}} on some other page [23:35:32] is it not easy to learn php [23:35:33] phobie: #semanticmediawiki [23:35:36] Not sure what the exact #ask syntax is though [23:35:44] phobie: But your best bet is the mailing list [23:35:46] Does that channel actually exist? :O [23:35:49] im just going to reinstall... [23:35:51] Yes [23:35:55] which of these has the haystack first? [23:36:08] I'm hosting it [23:36:14] strpos() has it last I believe [23:36:21] Or rather the group has me handling the IRC stuff [23:36:25] No in_array() has it last [23:36:32] I was going to make a little list [23:36:38] brion: 'array_merge() has inconsistent behavior when your string indexes happen to look like integers, which is undesireable for arbitrary input.' [23:36:40] *RoanKattouw apologizes [23:36:45] 03brion * r43049 10/trunk/extensions/ContributionReporting/ContributionTotal_body.php: [23:36:45] * accept all MW timestamp formats [23:36:45] * add fudge factor so we can easily include pledged donations not in the CRM database [23:36:45] brion: is that correct? [23:36:56] but yes, whatever the list is, the answer is strpos and friends [23:37:04] strrpos, stripos, etc. [23:37:08] well anyway guys, this put me on the right track. i'll ask for more help if i see i need it. [23:37:14] thanks so much! [23:37:40] AaronSchulz: the two bugs I mentioned in that last commit appear to have been both due to you [23:37:51] so you might want to look at the diff [23:37:56] AaronSchulz: array_merge() does two different things, depending on your array [23:38:12] if you have string keys, it behaves in one way, preserving keys [23:38:18] brion: where are decent docs on +? [23:38:24] if you have integer keys, it behaves in another way, where it throws away key values [23:38:28] brion: that's what I thought [23:38:33] 03siebrand * r43050 10/trunk/extensions/ (74 files in 67 dirs): Localisation updates for extension messages from Betawiki (2008-11-02 00:14 CET) [23:38:37] http://www.php.net/manual/en/language.operators.array.php [23:38:48] the + operator appears to have self-consistent behavior [23:38:56] it just overrides in the opposite order from what i'd usually expect :) [23:38:59] Hey I found a fun one [23:39:04] but if you know that, it's fine for most uses [23:39:05] brion: I was already there [23:39:13] well, that's as clear as they get :D [23:39:27] What's the output of: $parts[3] = 3; $parts[2] = 2; $parts[1] = 1; $parts[0] = 0; echo implode('.', $parts); [23:39:36] it's counterintuitive, but it's probably more efficient [23:39:36] hmm [23:39:50] the time parameter, what format gets passed to it? [23:40:08] updating an array value requires two steps, hashtable search and then update [23:40:17] if you don't overwrite, then you can skip the second one [23:40:50] De [23:40:51] RoanKattouw: 3.2.1.0 [23:40:59] Yep [23:41:50] Another one: $parts[] = 'foo'; $parts[] = 'bar'; unset($parts[1]); $parts[] = 'baz'; // What does the array look like now? [23:42:17] arrays made sense to me once I saw the way array elements are defined in C [23:42:31] arrays are a hybrid of linked list and hashtable [23:42:52] Most things make sense once you see how they work in C :P except maybe for higher level concepts such as hash table [23:42:54] s [23:42:57] so you just have to think "how would I implement this operation on a hybrid hashtable and list?" [23:43:15] unset() just unlinks the element and removes it from the bucket [23:43:19] TimStarling: hmm, title.php was using array_merge wrongly before that [23:43:43] AaronSchulz: arguably [23:43:49] *AaronSchulz hugs annotate [23:43:49] but your way wasn't better [23:44:13] Oh, a crazy one: What does require_once('Foo.php'); require_once('foo.php'); do on a) UNIX/Linux and b) Windows, assuming Foo.php exists and foo.php doesn't? [23:44:25] DeFender1031: it is pretty slutty, but your best bet is the output of {{CURRENTTIMESTAMP}} or ISO 8601 [23:44:49] RoanKattouw: I think I'd rather just use a consistent case :) [23:44:56] Splarka, thanks [23:45:02] timestamp appears to be working [23:45:04] Linux is case sensitive [23:45:09] Yeah me too. But this goes to show what kind of weird stuff happens when you don't [23:45:21] iy was the rewriet rule i was using that cuased issues! [23:45:22] infinisoft: Yep, and Windows isn't [23:45:23] anything else can be ambiguous, especailly dd/mm/yyyy vs mm/dd/yyyy, always avoid! [23:45:25] tissues -_- [23:45:37] [23:45:44] it is documented though [23:45:56] So what happens here is that on Linux, you'll get an error about foo.php not existing, whereas on Windows, foo.php is the same file as Foo.php and it'll be included *twice* [23:46:00] so a fair exam question [23:46:05] Yeah probably [23:46:08] WTF? Why does + take the left-hand value? [23:46:15] "This behaviour changed in PHP 5 - the path is normalized first so that C:\PROGRA~1\A.php is realized the same as C:\Program Files\a.php and the file is required just once. " [23:46:22] Oh [23:46:27] Good, actually [23:46:45] *Splarka looks at the exam "Who would use this stuff? better to ask how many pints of blood in the human body and how fast you lose it out of a stab wound... as well has how many reverts per day br*on can do" [23:46:48] '$c = $a + $b; // Union of $a and $b' [23:47:01] brion: heh, wrong use of the word UNION [23:47:19] AaronSchulz, why? [23:47:28] $a = array("a" => "apple", "b" => "banana"); [23:47:30] $b = array("a" => "pear", "b" => "strawberry", "c" => "cherry"); [23:47:32] Another nasty and well-known one is this: foreach($arr as &$a) { //do stuff } unset($a); [23:47:44] Well, except that it's really a union of the array keys, I guess, with the values coming along for the ride. [23:47:53] PHP IS EVIL. [23:47:57] Except if the keys are integers [23:47:58] Let's switch to Python. [23:48:08] Let's hear proposals for a migration path. [23:48:11] only if python introduce { } [23:48:13] lol [23:48:13] :) [23:48:20] Let's switch to a decent, non-interpreted language and make MW faster :D [23:48:29] maybe we just need alternative manuals [23:48:32] RoanKattouw: ansi c? :) [23:48:32] exobuzz, >>> from __future__ import braces [23:48:32] File "", line 1 [23:48:32] SyntaxError: not a chance [23:48:35] PHP isn't interpreted [23:48:42] RoanKattouw: Code it as a custom HTTP server in Assembly [23:48:42] it's a VM [23:48:49] hehe [23:48:57] Oh yeah right there's pre-compilation [23:49:02] is there any other project that has a chosen language so hated by every developer as mediawiki php? [23:49:02] RoanKattouw, there's nothing wrong with scripting languages. [23:49:06] 6502 asm! [23:49:17] Splarka, texvc? [23:49:30] *AaronSchulz should do his c++ lab homework [23:49:50] c++ always makes me hear the term "Logical Positivism" in my head [23:49:53] and then I giggle [23:49:57] *AaronSchulz is strange [23:49:58] What the hell [23:50:25] what about algol or cobol or something modern like that ? [23:51:09] maybe php will become non shit one day [23:51:16] exobuzz: FORTRAN? [23:51:17] i wouldnt mind if it required lots of "breaking" changes. [23:51:25] MattJohnston: why not ;-) [23:51:34] exobuzz, they do. They've explicitly rejected a lot of sane changes because of compatibility. [23:51:34] exobuzz: You'd be creating a new language, basically :P [23:51:38] MattJohnston: its got double precision at least! [23:51:42] Instead of taking the correct approach like, say, Python 3. [23:51:44] RoanKattouw: well .. true [23:51:46] exobuzz: Yeah :) [23:51:47] *AaronSchulz thinks array_merge was worse that using wfArrayMerge [23:51:56] in title that is [23:52:01] Why not just write our own language? Then we can even embed it inside of MW and allow it for scripting templates... heh [23:52:02] whatever [23:52:11] AaronSchulz: Does array_merge() still work for error arrays? [23:52:12] Dantman: NEVER! [23:52:21] heh [23:52:24] RoanKattouw: errors and perm stuff needs to be recursive [23:52:35] Dan: and write your own interpreters for it for every platform, IN it? [23:52:35] try: python has some weird habits as well you know [23:52:38] except: pass [23:52:42] *Dantman votes for JavaScript [23:52:46] as well as the compilers for those platforms, also written in it [23:52:58] RoanKattouw: so that won't work well :) [23:53:06] AaronSchulz: Is that a yes or a no? :P Should I merge error arrays with array_merge() or wfArrayMerge() [23:53:13] ? And what happens if I use the wrong one? [23:53:23] you should merge the with wfMergeErrorArrays() [23:53:26] Title.php used to use it array_merge() [23:53:28] which I just wrote [23:53:37] merging arrays is more an art than a science, it seems ;) [23:53:37] *Dantman casts a secondary acceptance of Lua [23:53:40] it was wrong because it didn't recursively merge the sub-arrays and such [23:53:44] Hmm the API uses array_merge() in some places IIRC, those would need to be fixed [23:53:48] array_merge() works just fine, as long as the caller expects duplicates [23:54:05] TimStarling: Well in the API the caller is just interested in the first element :P [23:54:09] there are too many different ways to do it that a function called 'array_merge' has a non-trivial chance of not doing what you wanted :) [23:54:20] TimStarling, what's that Python snippet supposed to illustrate? I just get a syntax error when I run it. [23:54:32] brion, mainly because of PHP's failure to distinguish lists from dictionaries . . . [23:54:35] brion: the problem is the word "array" [23:54:44] if it was really an array... [23:54:49] quite, that complicates things :) [23:54:58] still, sub-arrays are non-intuitive [23:55:12] T_T php's concept of arrays is even more troublesome with JSON [23:55:15] try: Simetrical: it demonstrates that python throws exceptions at all sorts of inopportune moments [23:55:17] there's an array_merge_recursive() but i'm pretty sure it never does what i want ;) [23:55:26] should we just care about the top keys and values, or try and merge sub-arrays on top key collision and so fourth? [23:55:31] try: for conditions that you don't really care about and would be happy ignoring [23:55:33] array() -> is it {} or []? [23:55:34] except: pass [23:55:36] except: pass [23:55:58] Well, okay, Python isn't perfect either. But it's better than PHP. [23:56:15] heh, Fallacy of the Excluded Middle [23:56:31] And if you try for( var in json ), expecting {}, but it gives you [], you end up iterating over all the array methods [23:56:32] python actually kind of drives me nuts since i frequently like to have ordered dictionaries. in python i end up using lists of tuples, which are harder to access randomly [23:56:54] but if you're just accessing them in order that's just fine [23:57:16] Yeah, ordered dictionaries are sometimes nice, I admit. [23:57:39] i loooooove list comprehensions though [23:57:43] also the python manual rarely documents its exceptions [23:57:58] i've never liked the python manual [23:58:13] the php manual is usually decent for procedural functions and insanely hard to search for anything else [23:58:37] the new categorisation in the PHP manual was a step backwards [23:58:42] I can never find modules anymore [23:58:53] you just have to search [23:59:34] what's a good place to buy hair dryers?