[00:00:07] Can someone help me? I'm trying to install mediawiki on a Times Sinclair XZ81. I have the 16k memory upgrade but it keeps complaing about space. [00:00:08] ^Timex [00:00:38] haha [00:00:40] lol [00:00:51] Splarka: add more swap [00:00:59] you can attach a cassette drive [00:01:17] k, sec [00:01:23] for added capacity, try replacing the cassette with an ipod w/ a recording attachment [00:01:24] which RCA plug goes where? [00:01:33] hehe [00:01:46] you can always find some way to pervert anything more appleish [00:01:48] actually, installing mediawiki on the ipod would work better ;) [00:02:01] brion: I already have lighty on my iPhone. [00:02:03] (srsly) [00:02:03] http://www.ipodlinux.org/ [00:03:05] I wonder if one day an Ubuntu sticker won't be a sticker with the Ubuntu logo on it, but would actually run Ubuntu [00:03:36] that'd be cool [00:04:06] I'd get a MediaWiki sticker, and stick it on top of my MySql sticker, stuck upon my Apache sticker, stuck upon my Ubuntu sticker [00:04:15] or whatever. [00:04:18] I'm confusing myself [00:04:20] 'Intel install' sticker, stick it on.. 'Intel installing....' 'Intel installing...........' 'Intel inside!' [00:04:26] lol [00:05:05] werdnum, you run iPhone Linux or it's in the iPhone store or whatever sissy equivalent of a package repo they have? [00:06:32] it's in cydia (the jailbreak package repo) [00:11:08] *Simetrical has no idea what that means, does not speak Apple [00:12:33] it's just a standard APT repo. [00:30:24] 03(mod) subcat count > 0 for cats with no subcats on dewiki - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=16614 +comment (10Simetrical+wikibugs) [00:34:51] 03vyznev * r44641 10/trunk/phase3/maintenance/cleanupTitles.php: Make cleanupTitles.php skip file pages (a bit late to fix that now, but oh well...) [00:36:55] I guess I kind of assumed that it was SOP to run cleanupImages first :/ [00:39:20] Restore from backup!! [00:40:31] I hate OpenOffice. [00:40:36] STOP TRYING TO BE INNOVATIVE WITH YOUR UI. [00:40:43] Why don't we require pages to re-render after 30 days? [00:40:45] use staroffice. it's closed source, so it must be better [00:40:47] OBEY THE UNIVERSAL STANDARDS THAT EVERY OTHER BLOODY DOCUMENT EDITOR DOES. [00:41:00] I just use actual Office. Works all the time. [00:41:00] Ctrl-I does not mean "make the last word I typed italic". It means "turn off italics mode". [00:41:21] wtf? [00:41:26] wow. [00:41:45] Anyway, page_touched can be as old as 2005 on en.wiki. So my question is, why is that a good thing? [00:42:13] because enwiki has lots of really obscure articles? [00:42:38] yksinaisyyteni, I'm amazed you don't use IpbWiki or something on the toolserver instead of that MediaWiki rubbish. [00:42:40] But what harm does the occasional purge from cache do? [00:42:43] Vyznev: ^ [00:42:54] MZMcBride, who says articles aren't occasionally purged from parser cache? [00:43:06] Simetrical: The page_touched column? [00:43:12] well, as long as it only happens when someone tries to view the page, nothing [00:43:23] but who says anyone has looked at those pages since 2005? [00:43:32] "Show us where the bad man page_touched you" [00:43:35] Does it matter if anyone has looked at them? [00:43:52] I mean, some things that don't re-render don't get put into appropriate tables... [00:44:03] Does page_touched actually reflect the last time the page's cache was in fact purged? [00:44:12] Or is it more sort of theoretical? [00:44:34] I think it's actual? [00:44:45] hmm, you do have a point... maybe it should be done as a background job [00:45:18] It doesn't necessarily need to be 30 days, but anything better than indefinite... [00:46:20] http://en.wikipedia.org/w/index.php?title=Dolores_Lake&diff=210457114&oldid=68679810 [00:46:31] Things like that... make me sad. [00:48:20] What, brion editing? [00:48:23] ugh [00:48:26] what, you think it should've DWIMed it? :) *g,d&r* [00:48:55] Reedy: Look at timestamps? [00:49:13] haha [00:49:38] two years of bleh? [00:49:39] wow. [00:49:51] How in the world do these things go unnoticed? [00:49:52] Presumably that page didn't make it into the appropriate table? [00:50:02] VartanSimonian: Because we don't force re-rendering. ;-) [00:50:08] Because nobody listens to me. :-( [00:50:19] To who? [00:50:25] I think I hear that darn wind again. [00:50:29] >_> [00:50:34] Someone close the window. [00:50:34] (sad face) [00:50:40] what, we have a table for horribly broken redirects now? [00:50:45] over and over and over again [00:51:03] Vyznev: We have a redirect table. [00:51:13] MZM: nobody listens to me either, or anybody else [00:51:24] Meanwhile, on a somewhat related rant, the categorylinks and protection tables have gone to shit. [00:51:47] But at least there are bugs filed about that, or something. [00:51:52] I want §ion=T-# links to behave like &redlink=1, I want diffs in the API to return the byte size of the diff returned, I want uploads to have SVG preview [00:52:05] "NOTE: as of August 2007, database dumps for Wikipedia and other Wikimedia projects as provided on http://download.wikimedia.org/ have incomplete data in this table: only redirect pages that have been created or edited recently are present. For older redirects, resort to using the pagelinks table." [00:52:07] well, brion listens to me, but only to my jokes [00:52:14] That's just sad. [00:52:28] I don't listen to myself. [00:52:35] Neither do I think through what I do say. [00:52:53] 00:44 < Simetrical> Does page_touched actually reflect the last time the page's cache was in fact purged? [00:52:57] 00:44 < Simetrical> Or is it more sort of theoretical? [00:53:01] Simetrical: it's used for If-Modified-Since et al. [00:53:03] More quoting. [00:53:25] [18:52] More quoting. <- QFT [00:53:40] :P [00:53:47] Splarka: WB. [00:53:59] MZMcBride: don't we have a maintenance script to fix that or something? [00:54:01] That whole 'nick collision' thang seems to be happening more often. [00:54:05] or if not, shouldn't we? [00:54:19] Vyznev: Sort of. [00:54:35] *Skizzerz wonders why Splarka has a wikia cloak [00:54:36] I mean, we can rebuild tables and things, but those scripts aren't run because they break things. [00:54:44] *Skizzerz also wonders why brion hasn't made mediawiki cloaks yet >:( [00:54:46] He's a Wikia import. [00:54:50] They don't 'break things'/ [00:54:55] Skizzerz: That's James_F's thing. [00:55:01] They just take forever to run. [00:55:10] werdnum: Tim says they break things. [00:55:14] but James_F isn't pokable at the moment :( [00:55:16] And I trust him more than you. :P [00:55:27] werdnum: Wikia breaks tons of stuff [00:55:55] plus if they enabled error_reporting, I'm sure they'd be getting tons of E_NOTICES and E_WARNINGS on each page [00:56:16] because they like to define their variables AFTER they check for them in if() statements [00:56:48] http://www.webvaluer.org/us/www.wikipedia.org [00:57:01] dang [00:57:01] Vyznev: I think something that ... runs all the time or something would be better. Is that even possible? [00:57:04] MZMcBride: The scripts or WIkia? [00:57:24] werdnum: Maintenance scripts break things. That's their purpose, according to Tim. [00:57:35] 03tomasz * r44642 10/trunk/extensions/CentralNotice/ (3 files): Renaming rights so they fit in with other extensions and for bug #16364 [00:57:43] *Splarka stabs the snow causing phone line failure [00:58:36] MZMcBride: sure, if we had an index on page_touched we could have, say, every edit also pull the oldest untouched page in the DB and reparse it, or something [00:58:56] 03(FIXED) Rights don't fit in - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=16364 +comment (10tfinc) [00:58:58] Hmm. Shall I file a bug? [00:59:08] why not? [00:59:17] when in doubt, file a bug :) [00:59:29] Heh. I usually get stabbed for my bug filings... [00:59:48] MZMcBride: my dialup just died twice in quick succession, for no reason, it doesn't do that more than maybe twice a year usually [01:00:16] Skizzerz: too lazy to switch it back, also lilo had to set it specifically on me due to a really weird bug that was making it 'wikicities' all the time, long after wikicities was deprecated [01:00:28] werkd [01:00:29] so I'd feel bad getting rid of it [01:00:31] *weird [01:02:44] Skizzerz: I'm about to fire off an e-mail about the cloaks. [01:02:53] many thanks :) [01:03:53] 03(NEW) Implement mechanism to force re-rendering of pages - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=16657 15enhancement; normal; MediaWiki: Page rendering; (public) [01:03:57] Vyznev: ^ :-) [01:04:48] 03(mod) Rights don't fit in - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=16364 (10matthew.britton) [01:05:06] 03brion * r44643 10/trunk/extensions/CodeReview/Subversion.php: [01:05:06] Run subversion proxy junk through PHP serialization instead of JSON... the json decoder seems a little unreliable in PHP for some inputs, like freaking out over nulls in a string. [01:05:06] Most confusing of all, sometimes it fails out by returning the raw JSON input as the result instead of failing sensibly with false or null return. :P [01:05:17] 03brion * r44644 10/trunk/tools/codereview-proxy/index.php: [01:05:17] Run subversion proxy junk through PHP serialization instead of JSON... the json decoder seems a little unreliable in PHP for some inputs, like freaking out over nulls in a string. [01:05:17] Most confusing of all, sometimes it fails out by returning the raw JSON input as the result instead of failing sensibly with false or null return. :P [01:05:34] >_> [01:05:42] A commit so nice, he did it twice. [01:06:59] 14(INVALID) Implement mechanism to force re-rendering of pages - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=16657 +comment (10brion) [01:09:08] well, that was quick :/ [01:10:39] make a bot to null edit old pages instead? [01:11:05] Well, I thought about that. [01:11:08] think how many bugs could be squished before hatching if br*on just paid attention here X_X [01:11:31] 01:10 < gurch> make a bot to null edit old pages instead? [01:11:34] ew ew ew ew ew [01:11:37] *werdnum stabs gurch repeatedly. [01:11:46] brion: Should I run a script to null edit all 11 million pages on en.wiki? :-) [01:11:54] no [01:11:58] only the old ones [01:11:59] no [01:12:09] touch.py just needs a minor modification or two... ;-) [01:12:25] gurch: I imagine I'd dump the list and work from oldest to newest. [01:12:42] MZMcBride: see my notes on the bug [01:12:58] MZMcBride: if you did that, you could quite possibly crash the site. [01:13:10] (if it was fast enough) [01:13:18] really? [01:13:30] Rendering pages *takes* *time*. That's why we cache them. [01:13:52] *gurch recommends the entire site not be crashable by a single user [01:14:04] not everyone is well-intentioned :) [01:14:12] just woke up... dreamed about ImageMagick [01:14:30] brion: I saw your notes. I want to update the links tables. Should I file a bug about running a maintenance script or two? And does it have any possibility of being resolved? :-) [01:14:37] <^demon> Mathieu1: At least you didn't dream about GD :D [01:15:27] ^demon, imagemagick is actually my nightmare these days [01:15:30] 03(FIXED) Some diffs are escaped thus appear in single line - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=16643 +comment (10brion) [01:15:35] Skizzerz: E-mail sent. We'll see what happens. [01:15:40] I don't even want to think about GD [01:15:40] thanks again [01:15:45] MZMcBride: refreshLinks, or maybe something else, depending on what you're actually talking about [01:17:25] Vyznev: Still about? Is refreshLinks.php what I want? [01:17:29] http://svn.wikimedia.org/viewvc/mediawiki/trunk/phase3/maintenance/refreshLinks.php?revision=35098&view=markup [01:17:48] MZMcBride: maybe you should describe the actual problem. [01:18:30] I did. I even filed a bug with links. [01:18:32] :X [01:18:45] 03brion * r44645 10/trunk/tools/planet/gmq/templates/index.html.tmpl: link icelandic from the english [01:18:46] Then maybe you should wait for that bug to be resolved. [01:18:54] *MZMcBride stabs werdnum in the throat. [01:18:55] 03(mod) Create Scandinavian Planet Wikimedia - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=16511 (10brion) [01:19:06] werdnum: It was resolved, you twit. [01:19:10] MZMcBride: can you point out which one? [01:19:20] https://bugzilla.wikimedia.org/show_bug.cgi?id=16657 [01:19:20] or do you mean the vague one? [01:19:23] *Splarka eats popcorn [01:19:36] The one about re-rendering pages before the next epoch... [01:19:41] this level of slapstick is usually only seen in the Three Stooges [01:19:47] do you actually want "fully populate redirect table"? [01:19:49] MZMcBride: No, you report the problem, not the solution. [01:20:00] Splarka: Please help. :-( [01:20:15] MZMcBride: What *problem* is being caused? [01:20:21] *^demon passes Splarka a soda. [01:20:24] old pages are old, redirects aren't always redirects [01:20:28] brion: Yes. But I was told by Tim that running maintenance scripts that rebuild tables isn't really feasible... [01:20:39] Not when the tables are so large and prone to breakage. [01:20:46] and revisions and pages don't all have sizes [01:20:56] --old-redirects-only : Only fix redirects with no redirect table entry [01:20:57] werdnum: Tables aren't populated correctly. If I knew which entries, I'd play the lottery. [01:20:58] sounds feasible to me [01:21:17] Splarka: well no revisions before april 2007 have sizes because the field didn't exist then. but that's not going to change [01:21:24] at least, I assume it isn't [01:21:26] *^demon wants a more useful redirect table. [01:21:27] brion: What about other pages? I mean, we change schema and all that regularly. Do we really want pages not re-rendering for years? [01:21:43] I was told it wasn't going to change when I asked back in april 2007 :) [01:22:06] I was told there'd be no math [01:22:17] I was told there'd be no Christmas. [01:22:20] MZMcBride: refreshLinks.php [01:22:27] it ain't got nothing to do with page_touched [01:22:40] it's been run many times over the years [01:22:56] So I should open a bug to have refreshLinks.php run. [01:23:03] if you like [01:23:05] *gurch runs refreshBrion.php [01:23:11] How can it refresh the links and not touch page_touched? [01:23:18] why would it touch page_touched? [01:23:28] there's no need to force recaching of the page [01:23:34] Because it presumably has to re-render the page to check the links, no? [01:23:44] it has to parse it. [01:23:58] that has no change on externally cached resources [01:24:06] Hmm, okay. [01:25:11] page_touched is "The last time that external caches had to be purged" [01:25:20] It's mostly for If-Modified-Since, IIRC. [01:25:25] (and Last-Modified) [01:26:33] 03(NEW) Run refreshLinks.php on WMF wikis - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=16658 minor; normal; Wikimedia: General/Unknown; (public) [01:27:26] 03(mod) Implement mechanism to force re-rendering of pages - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=16657 (10brion) [01:28:18] now bug 16658 gets wontfix'd :) [01:28:37] I hope not. :-( [01:29:38] 03(mod) Run refreshLinks.php on WMF wikis - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=16658 +shell (10brion) [01:31:33] werdnum, that says twcenter.net should be getting $285 a day in ad revenue. That's like what we get a month. I call BS. [01:32:01] Simetrical: *shrug* [01:32:08] The site tends to sell for like $2000, too. Not $300,000. [01:32:27] It says that site is worth $300k? [01:35:16] 03(NEW) Prettify permalinks - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=16659 15enhancement; normal; MediaWiki: General/Unknown; (brion) [01:36:20] brion: ooh, can we start using $wgActionPaths on Wikimedia? [01:36:26] *werdnum <3 $wgActionPaths [01:37:57] That bug sure looks tempting, but I think I want to be able to eat in SF. [01:40:18] 03(mod) Include permalink in printable version - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=16428 +comment (10brion) [01:40:44] morning [01:40:50] hi tim [01:42:47] 03(mod) Run refreshLinks.php on WMF wikis - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=16658 (10brion) [01:45:14] brion: I was in contact with David by IRC before the release [01:45:31] super [01:45:46] that's where the CVEs came from [01:51:26] 03(NEW) Database table cleanup tracking bug - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=16660 15enhancement; normal; Wikimedia: Site requests; (brion) [01:51:27] 03(mod) Phase I usernames with underscores break the system. - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=323 (10brion) [01:51:31] 03(NEW) 1.13.3 Patch is Missing IEContentAnalyzer.php - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=16661 04CRIT; normal; MediaWiki: Uploading; (joel.sherrill) [01:51:35] 03(mod) subcat count > 0 for cats with no subcats on dewiki - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=16614 (10brion) [01:51:36] 03(mod) Corrupt protections in page_restrictions table - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=16629 (10brion) [01:51:38] 03(mod) Run refreshLinks.php on WMF wikis - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=16658 (10brion) [01:52:32] 03(mod) UtfNormal replacement proposal - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=5303 +comment (10brion) [01:52:44] 03(mod) subcat count > 0 for cats with no subcats on dewiki - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=16614 (10Simetrical+wikibugs) [01:52:48] :O 16660! [01:53:02] >:D [01:53:12] Assigning 16661 to Tim. [01:53:39] 03(mod) 1.13.3 Patch is Missing IEContentAnalyzer.php - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=16661 (10Andrew) [01:55:24] <^demon> brion: What to with bugs involving the cur table? [01:56:24] *werdnum eats a Darrel Lea Christmas Nougat Pudding. [01:57:17] 03(mod) Database table cleanup tracking bug - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=16660 (10public) [01:59:17] 03(mod) Tracking bug (tracking) - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=2007 (10public) [01:59:24] 03(mod) Special:Wanted* - purge link tables - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=16112 (10public) [01:59:26] 03(mod) Database table cleanup tracking bug - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=16660 (10public) [01:59:55] <^demon> Splarka: I've read so many "it would be nice if..." bugs in the past few days..... [02:00:14] <^demon> Splarka: Yes, we talked about that earlier :) [02:00:36] ^demon: check if they need to be adapted to current schema or can be closed :D [02:00:36] 03(NEW) It would be nice if transcluded section edit links were hidden if user had no permission to edit , or treated like redlink=1 - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=16662 15enhancement; normal; MediaWiki: Page editing; (herd) [02:00:44] ^valid use [02:01:35] <^demon> brion: Referring to a non-existent query on a non-existent table in a non-existent function? [02:01:39] 03(mod) It would be nice if transcluded section edit links were hidden if user had no permission to edit , or treated like redlink=1 - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=16662 (10brion) [02:02:01] <^demon> *existing, rather. [02:02:29] 03(mod) subcat count > 0 for cats with no subcats on dewiki - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=16614 (10brion) [02:02:44] ok i'm heading home [02:03:00] darn, got his penultimate comment [02:03:16] What's wrong with "it would be nice"? [02:03:46] <^demon> MZMcBride: Nothing in particular. [02:03:49] *Splarka snickrs [02:04:12] >_> [02:04:57] 03(mod) Database table cleanup tracking bug - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=16660 (10innocentkiller) [02:04:58] 03(mod) Some categorized pages are (multi) listed as uncategorized - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=1444 (10innocentkiller) [02:06:03] Is there a way my extension can do a search without parsing Special:Search? [02:06:05] 03(mod) Database table cleanup (tracking) - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=16660 summary (10public) [02:06:07] *MZMcBride fixed naming convention. ^ [02:06:19] trust MZMcBride [02:06:22] MZNitPicke. [02:06:25] 14(WFM) deadlock error on edit save - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=1274 +comment (10innocentkiller) [02:06:59] I need to locate articles with a particaular template [02:07:05] and then parse bits out of that template [02:07:16] sprax_: [[Special:WhatLinksHere/Template:Foo]] [02:07:29] To parse bits, conditional ParserFunctions are best. [02:07:30] i have a problem... for some reason, this is malformed when used in pages: http://wiki.mobianlegends.com/wiki/Template:VG [02:08:00] {{#if:{{{param}}}|[[Category:Foo]]}} [02:08:45] MZMcBride I'm storing preferred NT ACLs in a custom template with a page for every path which is audited [02:08:59] so that ACLS/FILESERVER/E/SHARED/Accounting [02:09:06] *MZMcBride blinks. I only understand maybe half of that sentence. :-) [02:09:21] has an article with a template that has all the ACEs we went for that directory [02:09:27] yeah [02:09:29] its kinda nuts [02:09:35] my work is a special case... [02:10:05] so what I want to get a list of all the articles where the template exists [02:10:13] and parse out a specific user or group from it [02:10:30] 03(FIXED) Uncategorized images should be Uncategorized files - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=7793 +comment (10innocentkiller) [02:10:30] to generate a list of folders that a particular user or group has rights to [02:10:33] 03(mod) Prettify permalinks - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=16659 +comment (10Wiki.Melancholie) [02:10:36] like I said, it's pretty nuts [02:11:20] I already have a few parser hooks which use getContent(); [02:12:05] so if there is a function which will return an array of page names matching my query [02:12:08] it would be awesome [02:12:16] I'm rummaging about the manual now as well [02:12:35] but its not very long and my scheme is pretty hair brained [02:16:18] *Splarka ponders what else would be nice [02:18:04] Cigarettes and chocolate milk? [02:20:24] <^demon> MZMcBride: I'm there. [02:20:45] Excellent. [02:26:36] <^demon> brion: On the database cleanup...couldn't we drop cat_hidden since it's unused by everything? That info is in page_props, right? [02:26:55] He's gone. [02:27:05] <^demon> Missed that :p [02:27:45] ^demon: you don't notice that when tab-completing? [02:27:58] *MZMcBride wondered the same thing. [02:28:04] cat_hidden was added just in case. [02:28:11] werdnum, he could have started before brion left. [02:28:21] Also, brion is fast to type, maybe no need for tab completion. [02:28:22] Just in case what? [02:28:28] Just in case it was ever useful. [02:28:34] Because schema changes are like pulling teeth. [02:28:38] 3(NEW) It would be nice if MZMcBride had some cigarettes and chocolate milk - 10https://bugzilla.wikimedia.org/show_bug.cgi?id=16662 4BLOCKER; highest; MediaWiki: Excessive pedantic page deleting; (herd) [02:28:40] That's rather... [02:28:57] Splarka: :D [02:29:02] Splarka: those always freak me out. [02:29:07] Better to plan ahead and add useless fields rather than have to get sysadmins to change tables. [02:29:07] sorry:/ [02:29:11] *Splarka hides [02:29:14] Splarka, "BLOCKER" is supposed to be bold, IIRC. [02:29:22] <^demon> And red. [02:29:27] Simetrical: I think planning properly would be better... [02:29:30] is it, it is [02:29:36] Simetrical: I usually add useless fields when I create tables too. [02:29:38] MZMcBride, define "properly". [02:29:38] when was the last blocker opened? [02:29:57] Simetrical: Having a plan that will be implemented and fit the needs appropriately? [02:30:03] Fits. * [02:30:08] <^demon> Splarka: Better question. When was the last blocker opened that wasn't downgraded to at least critical within a day? [02:30:10] Splarka: when some clueless newbie thought "The second character on this page displays wrong". [02:30:17] Rather than adding random columns just to have them sit unused. [02:30:27] heh [02:30:28] MZMcBride, we're supposed to come up with plans for all possible uses of a table before we ever create the table at all? [02:30:33] MZMcBride: You should know that MediaWiki doesn't have a "plan". [02:30:49] werdnum: Yes it does. :P [02:30:56] If we had a functional process for getting schema changes approved in a timely and reasonable fashion, I wouldn't have bothered. [02:30:58] But nope. [02:31:07] Fair point. [02:31:28] But having duplicate columns / tables seems rather messy... [02:31:35] <^demon> Can I trade 1 useless field for 2 useful ones? [02:31:42] It wasn't a duplicate when I created it. [02:31:58] 138 blockers [02:32:02] Tim decided to use page_props instead of cat_hidden when he added the hidden category feature. [02:32:39] <^demon> Splarka: Only 1 is open for MediaWiki [02:32:55] latest was grawp [02:33:13] maybe because cat_hidden didn't exist at the time [02:33:17] ^demon, no, the exchange rate is one useful feature for ℵ₀ useless ones. [02:33:23] and I didn't think it was a good idea to add it [02:33:29] ^demon: https://bugzilla.wikimedia.org/buglist.cgi?query_format=advanced&bug_status=UNCONFIRMED&bug_status=NEW&bug_status=ASSIGNED&bug_status=REOPENED&bug_status=RESOLVED&bug_status=VERIFIED&bug_status=CLOSED&bug_severity=blocker [02:33:34] TimStarling, didn't exist at what time? [02:33:43] pretty red [02:33:49] It's existed since the category table has. [02:33:51] 03vyznev * r44646 10/trunk/phase3/maintenance/cleanupTitles.php: Followup to r44641: Only skip file pages that actually have a corresponding file. [02:33:55] Unless my memory is badly failing me. [02:33:58] when I added the hidden category feature [02:34:13] *Simetrical breaks out the svn blame-mobile [02:34:41] 32222 simetrical -- Reserved for future use [02:34:44] 32168 greg cat_hidden tinyint unsigned NOT NULL default 0, [02:34:56] huh, Simetrical right for once: 04CRIT->04BLOCKER [02:36:30] ah yes, indeed it has existed since the category table has [02:36:44] but of course the category table was added after the hidden category feature [02:36:56] I can't find the diff for the hidden category feature. [02:37:01] I could swear it was later, though. [02:37:32] One thing I'll say for git, blames are a headache when you don't have reasonable revision id's. [02:37:58] that clears some things up at least [02:38:31] the feature was added in r31113 [02:39:04] the page_props table was added in the main commit, the category table was added in 32084 [02:39:19] sorry 32085 [02:39:31] I thought when you added that category table you were just trying to spite me [02:39:43] Maybe I decided on the schema before you wrote the feature, and forgot about the extra field. [02:39:56] I can't remember. [02:39:59] adding a useless new field to make the point that I should have done it [02:40:00] *werdnum shrugs. [02:40:09] *werdnum spites TimStarling. [02:40:16] I get my HSC results tomorrow morning. [02:40:24] but I guess if you didn't even know the feature existed, that couldn't have been the case [02:40:36] It was a difference of three weeks. [02:40:59] It's reasonably possible I had some versions of the category table hanging around for that long. [02:41:45] hi all [02:41:57] does anyone know of APIs that can be used to parse wikitext? [02:42:29] action=parse [02:43:33] Splarka, ah thanks, Ill check it out. [02:43:56] I had another question - if I have the HTML for a page, is there a way to convert it to a format similar to Special:Export's? [02:44:01] takes raw text, page title, oldid [02:44:08] hmm [02:44:19] that involves reverse parsing, so no [02:44:47] Export exports raw wikicode, which you get with action=edit, action=raw, or prop=revisions&rvprop=content [02:44:48] why is http://wiki.mobianlegends.com/wiki/Template:VG screwing up? [02:45:18] <^demon> RingtailedFox: It's an empty page...? [02:45:28] Splarka, ah I see [02:45:53] vIkSiT: it is confusing to some that it is XML, but the XML is just a wrapper for the escaped wikitext and meta info about it [02:45:56] i mean, when i use it on other pages [02:46:09] example: http://wiki.mobianlegends.com/wiki/Sandbox/3 [02:46:09] <^demon> Why would anything show up? It's empty. [02:46:23] Splarka, yes, I've realized that. Im using lxml (py) to basicaly get everything inside [02:46:34] but now, I'd like to get sections out of the wikitext itself [02:46:56] and thus, if I can parse the markup straight-away - I can go ahead and do it [02:47:10] <^demon> RingtailedFox: Part of it is your thumbnailing is failed on the template. [02:47:15] i know [02:47:26] it is trying to load SVG files, but i don't have those enabled [02:47:58] vIkSiT: what you might do is use expandtemplates, which will give you an XML (third style distinct from raw and parsed) of the preprocessor tree, which should show you the section breaks more clearly [02:48:46] but you can get just the content of one section with &prop=revisions&rvsection=# [02:48:53] Splarka, Im having a bit of trouble with using the api though (action=parse). Could you point me to an example of using it? (I assume action=expandtemplates?) [02:49:21] http://test.wikipedia.org/w/api.php?action=parse&text=[[This%20is%20a%20link]] [02:49:24] <^demon> RingtailedFox: I'm not really sure. I'm not good with parserfunctions. [02:49:58] Splarka, ah, so it only takes text? I was trying to use title=New_York with no results.. [02:50:16] http://en.wikipedia.org/w/api.php?action=parse&title=New_York ? [02:50:36] vIkSiT: title= is the page title [02:50:39] page= is for specific pages [02:50:42] bit confusing too [02:50:52] http://test.wikipedia.org/w/api.php?action=parse&page=Foobar [02:51:15] ah gotcha, thanks [02:51:25] http://test.wikipedia.org/w/api.php?action=parse&text=This%20page%20is%20called%20{{FULLPAGENAME}}&title=File:Foo [02:51:32] you need the title for parser functions and magic words [02:52:26] I see [02:52:57] hi, All! [02:53:08] hmm, so for say the New York page - why does everything turn up in the form of "<ol" instead of

Splarka, ^ [02:53:42] escaped of course [02:53:48] is there a method to do the conversion through the parser? [02:53:54] grab inside and then de-escape the html entities [02:53:59] no, you should do it on the client side. [02:54:00] only <>&" are escaped IIRC [02:54:13] ah ok [02:54:13] that is because the API returns many formats, xml/php/json [02:54:17] right [02:54:19]