[00:19:01] Well 1-2 secs is a very long time for a read operation [00:19:29] A simple read query, with the DB conn already open, finishes well within 1ms [00:47:00] TrevorParscal: ALTER TABLE article_feedback MODIFY aa_user_anon_token varbinary(32) NOT NULL DEFAULT ''; [00:47:10] Also change the definition in the main schema [00:47:14] yes [00:50:29] ^demon: I need to add an SQL patch to an extension - I need some help [00:50:38] <^demon> ok [00:50:51] the contents of the patch are the code Roan just sent me [00:50:59] well, plus a /* */ before the table name of course [00:51:26] should I use addExtensionUpdate? and if so, what method should I use? I see things like addField and addTable [00:52:29] just applyPatch I think [00:52:30] right? [00:52:34] <^demon> Yeah, applyPatch is fine [00:52:54] thanks [00:53:24] <^demon> you're welcome [00:53:52] <^demon> array( 'applyPatch', "$dir/sql/file.sql", true ) [00:53:56] <^demon> The true is important :) [00:58:23] I wrote a bunch of define()'s last weekend in my toolserver's common.inc.php mostly booleans like DO_RETURN and DO_ECHO which I pass in print_r() and other self written functions that use booleans for aguments. [00:58:26] Helps a lot :) [00:58:30] keeps the code sane [00:58:59] rather than doStuff( 'foo', true, false, 20, null ); [01:02:43] so, in this case it would be IS_FULL_PATH = true or some such... [01:02:47] interesting [01:03:03] it could be doStuff( 'foo', KR_ECHO, KR_NO_DB, 20 /* number of seconds */ ) [01:03:32] function doStuff( $stuff, $echo, $use_db, $seconds ) [01:04:09] I've also seen some people use in-function call declaration of temporary variables [01:04:22] using the fact that php returns the assigned value [01:04:37] dofStuff( $stuff = 'foo', $echo = true ) [01:04:45] actually works as a function _call_ [01:05:24] although I dislike leaving messy tempory variables like that in whatever scope. [01:05:51] seems like it would be better to just pass in options as maps [01:06:02] php core uses it as well. STR_PAD_LEFT is a define()'ed integer for a case switch if I recall correctly. [01:06:21] doStuff( array( 'stuff' => 'foo', 'echo' => true ) ); [01:06:28] yeah, <3 wordpress [01:06:42] in JS its pretty much THE way to go [01:07:32] wordpress uses it it a lot. passing arrays or even queries. like doStuff( 'stuff=foo&echo=true' ); very weird, but it get's used to. [01:07:45] <^demon> Arrays are nice. [01:07:51] I like it for simple functions in wordpress. [01:07:56] wordpress does this thing too where if you pass a string it will get an array from it via parse_str, otherwise it will take an array [01:08:05] <^demon> There's a proposal to allow you to name your arguments in php. [01:08:28] <^demon> So you could define your function as function doSomething( $var1, $var2 = '', $var3 = null ); [01:08:42] And ofcourse in jQuery all the time $.extend(default_options, options_argument) [01:08:53] <^demon> And call it like doSomething( var1 => $something, var3 => $somethingelse ); [01:09:31] ^demon: that would be nice, could reduce the bloat of using maps as args [01:10:04] <^demon> There's benefits and drawbacks. [01:10:11] in MediaWiki I often see /* comments */ in the function call. Also interesting (not sure if it's common in MW, but I've seen if a few times) [01:10:19] <^demon> What would doSomething( var2 => 1, 'somethingelse' ); do [01:10:30] <^demon> Would somethingelse be var2? Or var3? [01:10:36] <^demon> Or since it's the first unnamed, var1? [01:10:53] ^demon: deja vu. wiki template calls [01:11:25] {{templ|1=foo|name=yay|lorem}} is "lorem" 2= or 3= ? [01:11:37] I often see that question. [13:16:01] Bryan: You have a minute ? [13:17:33] if it's really short [13:17:36] I'm at work [13:18:10] I'm documenting my code... [13:18:12] blergh [13:19:12] Okay, np. [13:19:51] I wanted to see if parser functions works properly in REL1_17 for you in a fresh svn checkout of both from branches/REL1_17 and using the web installer. [13:20:23] because for me, on a fresh checkout, althoguh the require_once() is in LocalSettings and it's visible on Special:Version, {{#if:}} stuff etc. does not get parsed. [13:26:01] PF needs to be installed separatelz [13:26:08] fucking QWERTZ [13:38:00] you gotta learn German keyboards ;) [13:38:10] s/learn/love [13:48:21] Bryan: ... I know that much ;-) [13:48:28] checkout 'both' from branches/REL1_17 [13:48:36] (phase3 and extensions/ParserFunctions) [13:56:51] Krinkle: purge? [13:57:14] Created a new page [13:57:22] guillom: I'm quite used to it by know, but I am having my laptop today standing next to the desktop [13:57:39] laptop is qwerty, but desktop is qwertz [13:57:47] checkout phase3, checkout PF, run installer, verify PF on Special:version, create new page, edited pages. Won't show up [13:57:54] Bryan: The same happened to JanPaul at the Hackathon [13:57:54] Krinkle: well... odd, is the only thing I can say :p [13:58:01] remember ? [13:58:07] no [13:58:11] Bryan, why don't you use a qwerty mapping on your qwertz? It's what I did with my German laptop (except I mapped it to azerty :) [13:58:27] He instead copied /w/ folder from an older checkout he had laying around. [13:58:40] But it would be great if someone could verify this breakage. [13:59:06] well, let me check [14:01:33] re keyboards, it's not very hard to get used to, but switching keyboards every other minute is :) [14:03:01] why is my IIS fucked up? [14:03:50] Krinkle: worksforme [14:04:13] <^demon> Bryan: Do you really want an answer to that? [14:04:14] if you're running a non-cgi php, try restarting your webserver [14:04:29] ^demon: rethorical questions [14:04:37] ? [14:04:46] Oh, I was typing something but replaced it with a ? [14:05:41] <^demon> ? ???? ?? ???? ???? ?????? ??????? ?????? [14:05:45] <^demon> Could get fun :) [14:07:28] Hm... seems trunk is broken. Notice: Undefined offset: 0 in /htdocs/w/includes/db/Database.php on line 1674
[14:07:44] from within function "OutputPage::addCategoryLinks". Database returned error "1054: Unknown column 'pp_value' in 'field list' (localhost)". [14:07:44] Backtrace: #0 /htdocs/w/includes/db/Database.php(741): DatabaseBase->reportQueryError('Unknown column ...', 1054, 'SELECT page_id...', 'OutputPage::add...', false) #1 /htdocs/w/includes/db/Database.php(1040): DatabaseBase->query('SELECT page_id...', 'OutputPage::add...') [14:07:52] SELECT page_id,page_namespace,page_title,page_len,page_is_redirect,page_latest,pp_value FROM `mw_page` `mw_page_props` WHERE (page_namespace = '14' AND page_title IN ('Testing','Hidden_cat') ) [14:08:26] somebody was poking at cats yesterday [14:08:34] shouldn't there be a comma after _page' ? [14:09:25] <^demon> r81067 [14:09:32] <^demon> Krinkle: ^ There's your error [14:09:36] r81087 [14:10:33] ^demon: Thx [16:10:49] hi folks....we're going to shift our test meeting over here [16:12:55] http://eiximenis.wikimedia.org/TestFrameworkDiscussion2011-01-27 [16:20:18] hi again [16:21:05] hi everyone [16:21:16] I left a note for nadeesha to join us here [16:21:17] Hi Rob [16:21:20] hello [16:21:27] jinesh: how's 1.17 testing going? [16:21:36] I informed Janesh to login to here [16:22:10] we have covered most of the areas in media wiki core and editor extension [16:22:39] great! which browsers have you tested with? [16:22:44] ogghandler was the problem since we came across either that server write issue or invalid MMIME type issue [16:22:54] jinesh: still? [16:22:59] IE6, IE7, FF 3.6, Chrome and Opera11 [16:23:26] yesterday i was able to upload ogg and ogv files, there's still an issue with duplicates [16:23:54] ah, that may be the case, we also using the same files I guess [16:24:15] yeah in that case it fails silently [16:24:20] we will double chack on that [16:24:41] i'll add a comment on your bug saying that [16:25:02] right now the search is not working as it suggest to use either Google or Yahoo search..... [16:25:28] bugzilla id 26922 [16:26:03] yeah lucene is not installed on prototype [16:26:10] i was chatting about that with folks yesterday [16:26:21] it's kinda involved, so i'm saving that for later [16:26:55] ah oki, its status wont fix, thats why we asked [16:27:24] for now at least :) [16:27:38] I'll let you know if that changes [16:28:12] also user rights link is not available now for the admin user given, according to bug comments its available for others [16:28:25] so we will leave that issue at that point [16:29:47] apart from that what we have to do is verify bug fixes, have detailed testing furtther and try Ogghandler again, I guess.... [16:30:35] that sounds about right to me [16:30:53] jinesh: which areas were you planning more detailed testing? [16:31:43] user preferences, and other user related settings [16:33:17] pdhanda (or anyone else in the channel that wants to weigh in), what areas have seen the most change in the codebase? [16:33:43] (I'm trying to figure out the highest risk areas to suggest to Calcey for focusing on) [16:33:58] hmm.. [16:34:28] fwiw, I think they've found some problems that apparently haven't been raised before [16:34:28] upload wizard maybe??? may want to check with neil first though [16:34:40] and are not the result of 1.17 changes [16:35:24] hexmode: so, some of the problems are also present in our current deployment? [16:35:45] a couple, yes [16:36:10] I think I made comments where I saw that [16:36:13] that's great finding those, by the way, but since they aren't regressions, that probably informs the priority for fixing them [16:36:31] sure, sure [16:37:00] Rob, is it useful to have a look with Mac/Safari, Mac/FF and Fedora/FF? [16:37:51] jinesh: I would say Mac/Safari yes, lemme think about the others [16:38:22] oki we will do with Mac/Safari too [16:39:35] the reason why I'm not sure it's worth testing the others is that it's going to be exceptionally rare to find a Firefox platform specific problem, but I'd like to know what others think [16:40:05] (that is, a Firefox platform specific problem that is actually a problem we need to fix) [16:41:12] It was recommended to focus FF on one platform [16:41:31] i have found that iceweasel sometimes behaves different from ff, although my colleagues say this is not possible :) [16:41:33] But we can distribute tests across platforms [16:42:52] Windows/FF already covered, so we can either be hapy with that or try one more [16:42:53] jinesh: maybe a very light pass with Firefox/Mac and Firefox/Linux [16:43:04] ok Rob [16:43:28] we will do that way, just a smoke test round [16:44:00] great! [16:44:50] anything else we need to discuss? [16:45:16] not for the moment [16:45:23] nope [16:45:42] not from our end too [16:45:48] ok, thanks everyone! [16:46:00] Thanks Rob [16:46:02] thanks everyone! [16:46:04] Thanks everyone [16:46:12] thanks everyone! [16:46:16] thanks, heading to the office [16:46:34] thanks [17:19:54] flipzagging: yh? [17:20:02] hexmode: hi [17:20:27] hey, so, on prototype I'm seeing JS errors in IE 7/8 [17:20:32] is that known? [17:20:32] yes [17:20:34] yes [17:20:35] ok [17:20:42] prototype isn't updated [17:20:48] just makin' sure [18:38:36] kaldari: http://en.wikipedia.org/wiki/Anthidium_manicatum [18:40:08] Here's my latest spider paper: http://peckhamia.com/peckhamia/PECKHAMIA_87.1.pdf [18:40:26] check out the crazy legs on those guys [18:45:01] jorm: The best thing she could do would be to put the paper online somewhere (not behind a paywall) so that other editors can refer to it. [18:50:44] jorm: and of course translating it to english wouldn't hurt her chances either :P [18:55:12] hexmode: Which tag is for Tim's queue? tstarling, right? [18:55:26] right [18:56:55] It's almost empty :O [18:57:56] <^demon> Give him some of mine :p [18:58:12] to be fair, I haven't been assigning to him. But if you think he needs some, offload! [18:58:55] <^demon> I gave him securepoll. that's definitely not me. [18:58:58] I just gave him one rev [18:59:07] "to be fair" ... but I could see how *you* wouldn't thinhk its fair. ;) [18:59:15] No worries [18:59:18] I don't have a wife and kid [19:00:07] He's had the wife, and he is out of classes. The kid might be a handful, though. [19:02:17] ^demon, RoanKattouw: any other comments on the assignments? SecurePoll I now know should be given to Tim, but any others? [19:03:29] <^demon> Non-trivial parser stuff. [19:04:17] Scary things, generally :) [19:05:42] <^demon> I'd like his input on the Message class tbh. [19:07:27] RoanKattouw: how can I recognise "scary things" if they don't have nasty parser-like fangs? [19:07:44] Not sure that's easy [19:08:03] Heh [19:08:12] ask everybody else and if they say "ew, I'm not touching that" then it's tim's :p [19:08:23] <^demon> haha [19:08:49] So now, b/c its too hard to identify what's scary, I'm just gonna be really, really frightened all the time [19:09:01] heh [19:09:03] and give all the revisions to Tim ;) [19:09:12] Mostly it'll be us moving stuff from our plates to his I think [19:09:16] <^demon> Hmm, things touching performance-sensitive code [19:09:21] <^demon> Message cache is one [19:09:55] RoanKattouw: sound like a fair division of labor to me :) [19:09:57] <^demon> hexmode: Did you see Tim's post to BZ about Postgres as well? [19:10:14] Not yet. Bug #? [19:10:32] <^demon> !b 26612 [19:10:32] --elephant-- https://bugzilla.wikimedia.org/show_bug.cgi?id=26612 [19:12:18] <^demon> I would love to drop ts2 support stuff. [19:13:19] is that the only reason to have a special schema? full-text support? [19:13:32] if so, then DROP IT, I SAY [19:13:56] *hexmode confuses his benvolent dictator hat for his bugmeister had [19:15:37] *brion adds extra security to his hat collection [19:17:21] <^demon> hexmode: It provided fulltext search to postgres < 8.3 [19:17:28] <^demon> Which as Tim points out, is nearing 3 years old [19:17:40] right [19:18:00] I repeat: drop it. [19:18:01] i think it's probably safe to discard ancient pg support [19:18:07] \o/ [19:18:41] see, now that the hat confusion is over, you have a proper BDFL decision [19:19:16] brion: per your suggestion the other day, would going through bugs marked "shell" and identifying those tasks that shouldn't require "shell" be good meta bugs? [19:19:36] hexmode, i'd say that's a good idea [19:19:47] some of those things will be easier than others though :D [19:20:04] sure [19:20:16] but things like logo updates we could knock out pretty easy [19:20:34] and i think there's a community pr win to getting more of those things through the queue quickly [19:20:57] Yes there totally is [19:20:59] yes, less places for people to gripe out the foundation is a win [19:21:01] About 2 yrs overdue [19:21:47] you know those domino's commercials where they're like "WERE SO SORRY OUR PIZZA WAS CRAP WERE DOING BETTER PLEASE LOVE US"? [19:21:53] well let's not look QUITE like that [19:22:50] <^demon> SORRY WE COULDN'T BE BOTHERED TO ADD A NAMESPACE PLZ LOVE US [19:23:06] *hexmode does not remember those commercials and resorts to teh google [19:23:39] <^demon> brion may have paraphrased ;-) [19:23:47] only slightly :D [19:24:02] best thing though was they started that ad campaign around the time when microsoft was ramping up windows 7 ads [19:24:19] which are all like "hey i want my computer to crash less" "oh great, microsoft FINALLY listened to their customers and made windows 7 crash less!" [19:25:08] http://www.businessinsider.com/dominos-pizza-stock-surges-thanks-to-awesome-new-pizza-recipe-2010-3 [19:25:19] haha [19:25:47] <^demon> I will admit, their pizza is way better. [19:26:00] "traffic have increased significantly since the launch of our new core pizza" [19:26:39] so, we should say "WP has become the #1 web property since we started listening to people" [19:27:06] hehe [19:28:20] now we just need to add more cheese [19:43:23] LiquidCheese [23:44:44] AryehGregor: Hey Aryeh, you had a wiki page about your censorship pet project somewhere, right? How far did you get with that? Just specs, or code also? [23:46:00] RoanKattouw, no code. http://www.mediawiki.org/wiki/User:Simetrical/Censorship [23:46:19] Thanks [23:46:43] Erik and Howie are dragging me into this meeting tomorrow about "Controversial Content User-Controlled Viewing Options" which I think is the very same thing [23:46:58] Yay. [23:47:07] So I figured I'd mention your thing at least, and see if you were interested in calling into the meeting [23:47:18] It's a voice meeting? [23:47:26] What time? [23:47:38] Not set yet, it's likely to be either 11-12:30 or 3:30-5 [23:47:44] Although I've said I'm not available for the latter [23:47:47] (both times PST) [23:48:03] (on Friday, Jan 28) [23:48:50] I certainly can't make the latter. The former would be 3-4:30 EST, right? I could go to that, but the question is whether it will accomplish anything. [23:48:58] I tend to prefer doing things in writing. [23:49:05] Yes, to the times [23:49:15] Right [23:49:32] Well dragging you in is just my idea, and there's no pressure of coures [23:49:58] I'll reconfirm the times with Dana (Erik's assistant) and let you know, and you can just tell me what you decide whenever [23:50:20] I guess I'll say no. [23:50:31] But if there's written discussion, I'd be mildly interested in looking at that. [23:54:51] OK [23:55:01] Well there's usually someone documenting stuff in an Etherpad live [23:55:23] Oh wait, the time is 2-3:30 EST, sorry [23:55:29] If we do go with 11-12:30 PST [23:55:45] I'll tell you the definitive time later and give you the Etherpad link, use it how you will