[00:00:13] I'm going to truncate fatal.log [00:00:17] So, on account creation these days, we're asking for "Real Name". Where is this being stored? [00:00:34] user table [00:00:45] it's been there for years [00:00:56] Evan Prodromou added it [00:01:51] lol [00:02:34] Cheat fix for the RC 808 is just to initialise it inline on it's declaration line [00:05:47] thanks! [00:17:56] so it was RCL rather than RC [00:18:35] That makes a bit more sense [00:20:12] fixed in r98955 [00:20:31] should we move on to the next wiki now? [00:20:42] neilk_ just gave me a heads-up about a fix for commons [00:21:05] syncing the 2 files then resources.php should be safe [00:21:16] Looks like the standard skin error is either rare, or has stopped.. [00:21:44] *robla pings AaronSchulz to check with him [00:21:55] the skin error? [00:22:08] AaronSchulz: should we keep investigating, or move on? [00:22:19] seems to have stopped...weird...OK [00:22:24] it's not like that many people use standard anyway ;) [00:22:29] next? [00:22:35] maybe they learnt the error of their ways [00:22:39] enwiki? [00:23:12] I see a few more Standard.php warnings [00:23:17] should we fix that too? [00:24:13] if you want [00:24:15] *AaronSchulz looks [00:25:51] nah....let's not block on that [00:26:18] enwiki? :) [00:26:51] Commons fixes are pushed [00:27:44] should we do commons now, then, while neilk_ is still here? [00:28:03] sure [00:28:24] hah, AaronSchulz just started to do enwiki [00:28:36] heh [00:30:09] [17:26] robla enwiki? [00:30:22] hmm, just categoryTree mostly [00:32:27] Oct 5 00:32:16 10.0.2.175 apache2[351]: PHP Warning: Missing argument 2 for efCategoryTreeAjaxWrapper() in /usr/local/apache/common-local/php-1.18/extensions/CategoryTree/CategoryTree.php on line 204 [00:35:55] fixed [00:36:00] same way as 1.17wmf ;) [00:36:08] Reedy: want to look at the Standard error? [00:36:45] Just gonna commit that category tree fix to trunk [00:37:38] Call to undefined method ApiQueryWatchlist::getSkin() in /usr/local/apache/common-local/php-1.18/includes/api/ApiQueryWatchlist.php on line 298 [00:38:09] heh [00:38:12] let's see [00:39:01] I bet that's from the context refactoring [00:39:09] Call to undefined method CologneBlueTemplate::getInternalLinkAttributes() in /usr/local/apache/common-local/php-1.18/includes/SkinLegacy.php on line 487 [00:39:59] $vals['parsedcomment'] = $this->getSkin()->formatComment( $row->rc_comment, $title ); [00:42:35] api fixed and pushed [00:43:27] standard error: http://paste.tstarling.com/p/NmJqjY.html [00:43:29] AaronSchulz, that looks potentially broken in trunk also [00:45:02] AaronSchulz, got a fix for that (linker changes) [00:45:03] should use Linker [00:45:05] yep [00:47:37] Fixed [00:47:47] So, just the standard skin error? [00:52:05] math nfs fun [01:05:29] Oct 5 01:04:56 10.0.11.30 apache2[7684]: PHP Fatal error: Allowed memory size of 125829120 bytes exhausted (tried to allocate 7864320 bytes) in /usr/local/apache/common-local/php-1.18/extensions/AbuseFilter/AbuseFilter.i18n.php on line 18732 [01:06:55] meh [01:11:14] now i want to create banners that go "Y U NO ENTER CODE CHALLENGE?" [01:15:01] <^demon> CUZ I HAZ REAL FRENDS [01:20:43] ^demon: http://www.mediawiki.org/wiki/Weekend_of_Code/Design [01:21:13] MY MOM KICKED ME OUT OF THE BASEMENT [01:23:25] <^demon> Nobody's told me anything about the contest :p [01:24:18] Who broke OpenSearchXML? [01:24:22] Oct 5 01:24:01 10.0.8.44 apache2[13592]: PHP Warning: array_map() [function.array-map]: An error occurred while invoking the map callback in /usr/local/apache/common-local/php-1.18/extensions/OpenSearchXml/ApiOpenSearchXml.php on line 88 [01:24:29] brion, ^ that looks familiar [01:24:35] ^demon: I didn't mention it before? [01:24:59] <^demon> I vaguely remember it. [01:25:07] grr [01:25:18] ^demon: i didn't know about it until the other day when i was told "hey, we need a design for this extension..." [01:25:24] brion, I seem to recall we had it as a transient issue and it went away [01:25:25] Yup.. [01:25:28] https://bugzilla.wikimedia.org/show_bug.cgi?id=27452 [01:25:36] it's just moved down 3 lines [01:25:47] Was a spammage of like 20 or so of them [01:26:06] Reedy, yeah probably MWSearch extension is failing and returning false or null instead of an array [01:26:54] $srchres = PrefixSearch::titleSearch( $search, $limit, $namespaces ); [01:27:02] if it's not an array, just set it to array() and debuglog? [01:29:22] lots of opensearch errors now [01:29:26] Arrrrrrgh [01:29:30] yeah [01:29:33] let me hack [01:32:23] Well that did naff all [01:52:38] Oct 5 01:52:08 10.0.2.185 apache2[12847]: PHP Warning: preg_match() [function.preg-match]: Compilation failed: missing ) at offset 11 in /usr/local/apache/common-local/php-1.17/extensions/Quiz/Quiz.class.php on line 768 [01:59:38] 1.17 ;) [01:59:48] oh yeah [01:59:49] heh [02:41:36] Uncaught TypeError: Cannot call method 'indexOf' of undefined [02:46:45] <metadata-expand> on image metadata [03:43:23] TimStarling: fyi, AaronSchulz is working on a workaround for 31179, since it seems to be implicated in 31367 [03:43:27] https://bugzilla.wikimedia.org/show_bug.cgi?id=31367 [03:43:34] https://bugzilla.wikimedia.org/show_bug.cgi?id=31179 [03:45:25] I reproduced it locally, the current situation is not pretty [03:46:15] SELECT page_namespace,page_title,page_touched FROM page WHERE page_latest = 0 AND page_len = 0; [03:46:21] fast since length is indexed [03:46:29] TimStarling: how did you repro? [03:47:03] just creating a page with the API [03:47:15] 109 rows [03:47:25] it's not really any better after r97291 [03:48:02] *AaronSchulz should work on fixes existing rows [03:49:54] ok [03:50:06] making a script [03:51:42] hi [03:51:51] hi guillom [03:52:26] hey robla -- sorry, I was having dinner and I completely forgot to join IRC. [03:52:34] How are we doing? [03:52:41] no worries....I'll give you an update in wikimedia-tech [03:52:56] ok [03:56:07] ha [03:56:31] the bug is caused by the logging hack for the same bug? [03:56:46] seems kind of self-referential but that's what the backtrace says [03:59:35] that's ...um....whut? [03:59:42] $wgRequest->getFullRequestURL() [03:59:51] the global is FauxRequest in that case (API) [04:00:39] I'm pushing out a revert, it seems to work locally [04:00:47] just pushed [04:01:55] page_latest=0 is confirmed also [04:02:51] before the revert I created such a page locally with the API [04:15:01] *AaronSchulz looks over script [04:15:14] AaronSchulz: this is the script to correct the db? [04:15:32] there's been no more page_latest=0 rows created since the revert was deployed [04:15:33] that or the one to make flap jacks [04:15:51] finish the flap jack script first, that sounds good [04:16:19] dammit, now I am hungry for pancakes [04:18:16] *molliug is vaguely trying to keep his eyes open. [04:18:45] molliug: go to bed :) [04:18:56] actually, seriously, go to bed [04:19:06] Not before 10, otherwise I'll still wake up at 4am [04:19:40] IRC + music is helping :) [04:22:43] *robla nudges AaronSchulz to get his script he's testing to TimStarling for review :) [04:23:32] mmmm....pancakes I can't wait [04:23:33] http://pastebin.com/4fWDAHnk [04:23:58] pancakes come to those who wait [04:24:17] big beautiful pancakes [04:24:30] TimStarling: see pastebin ^ ignore pancake banter :) [04:26:05] should work, have you tested it? [04:26:58] commented out the latest=0 cond and ran it on all pages on my testwiki [04:27:02] *AaronSchulz is looking around there [04:27:15] (and the len=0 of course) [04:27:46] maybe I should run it on the article I broke on my test wiki [04:28:04] good idea [04:29:21] looks ok on my wiki [04:30:02] not working for some reason [04:30:20] page_latest still bad? [04:30:49] $conditions['page_latest'] = $lastRevision; [04:30:52] that's why [04:30:52] MW_INSTALL_PATH strikes again [04:30:54] duh [04:31:30] that contention handling cond in updateRevisionOn() [04:32:10] works now [04:32:31] for new pages yes [04:33:11] TimStarling: were only genuine new pages affected? [04:33:47] some old pages have page_latest=0, maybe for a different reason [04:35:10] for example 20513897, it has nothing in the revision table for it [04:36:40] nothing in archive either, apparently [04:37:27] maybe it was from a previous version of the same bug, page_touched = september 29 [04:39:37] TimStarling: this is on enwiki? [04:39:45] yes [04:40:49] page_namespace=3 for all the older corrupted DB rows [04:41:40] TimStarling: have you poked around nlwiki much for these problems yet? [04:41:47] no [04:42:05] TimStarling: http://pastebin.com/dM9uukCb [04:42:21] should handle non-new pages [04:42:22] we only started noticing this on nlwiki. might be coincidence, or it might be that the problems just got more frequent with 1.18 [04:43:20] AaronSchulz: ok [04:44:29] just run that [04:45:11] Aaron just told me an interesting theory he has about what's going on here [04:45:48] not a lot of new info here, but I guess just piecing together the chronology [04:46:28] *AaronSchulz tacks the wiki ID onto the log dump [04:46:29] 1.17 (or earlier) possibly has a rare db corruption bug, which we're now fixing with this script [04:46:44] 1.18 additionally had 31179, which Aaron fixed earlier today [04:47:17] then he put in the logging, which introduced more page_latest=0 fun (oops) [04:47:48] so....(fingers crossed), once he runs this script, that should kill this problem dead [04:49:23] TimStarling: Done! Found 112 rows and fixed 95 rows. ;) [04:49:57] robla: yeah, it's more of a failure mode than a specific bug [04:50:12] there's lots of things that could go wrong that could lead to page_latest=0 [04:50:27] AaronSchulz: excellent [04:50:52] what of the other 17? [04:50:56] mysteries... [04:51:08] no revision rows, probably [04:51:44] mysql> select rev_id,rev_page from page,revision where page_len=0 and page_latest=0 and page_id=rev_page; [04:51:45] Empty set (0.56 sec) [04:51:47] for sure, but "why" is the question [04:51:54] deleted pages? [04:52:24] maybe the page row was inserted but then the process died before the revision row was inserted [04:58:57] so....I think we've done enough damage for one night [04:59:10] :) [04:59:44] TimStarling: AaronSchulz: I'm thinking we should schedule a window for same time tomorrow [05:00:51] 23:00-03:00 UTC [05:02:18] we *should* be able to get done well within the window. I'm thinking we hit dewiki, a big rtl wiki (e.g. arwiki) and then push to the rest in bulk all at once [05:02:37] TimStarling: AaronSchulz: sound like a reasonable plan? [05:04:41] yep [05:06:09] the dump files those made are in /home/aaron/logdump fyi [05:07:41] TimStarling: sound reasonable to you? [05:11:53] TimStarling: Aaron just left, and I'm about to go here as well [05:12:20] I'm going to post that schedule to the blog [05:16:13] was just getting some coffee [05:16:27] yes sounds reasonable [05:21:55] cool, thanks. I'd forgot to monitor this page, but it doesn't look like there was much reported there: http://meta.wikimedia.org/wiki/Talk:Wikimedia_maintenance_notice [05:22:47] alright...talk to you tomorrow! [06:38:58] guys, did we definitely updated wp to 1.18 [06:41:25] why this 1.18 works so different than trunk it seems that in this installation is more differences [06:45:17] most wikis are still on 1.17 [06:45:33] the rest will hopefully go tomorrow [06:46:10] http://blog.wikimedia.org/2011/10/04/mediawiki-1-18-deployment-today-to-all-wikimedia-sites/ see the update at the bottom of the post [06:46:24] ok, I wondering why huggle works fine with trunk but not what is deployed on wp now I tested it before with 1.19 to make it compatible [06:46:38] but it's totally broken [06:47:10] did something else changed in 1.18 what isn't yet in 1.19? [06:47:14] en pedia was converted; is that the project you are on? [06:47:30] I tried simple atm, en testing now [06:47:35] ok [06:47:58] en same... [06:48:23] care to report in wikimedia-tech? [06:48:35] there is problem with diff rendering it seems like if skin changed, or something with way how does it render pages [06:48:37] (channel_ [06:49:06] it's rather problem of huggle than wp I am trying to understand why it's ok with 1.19 which should have more changes [06:49:32] no idea [06:49:45] ok, thanks I will try to fix it [06:50:00] I hope this is final version of what is going to be there [06:50:08] of mediawiki [06:50:25] if more issues are found tomorrow then there will be some bug fixes that go out [06:50:30] it ought to be pretty close to final though [06:50:34] ok [06:50:56] is it possible to download sources of this [06:51:07] of what's deployed? sure [06:51:10] yes [06:51:55] svn.wikimedia.org/svnroot/mediawiki/branches/wmf/1.18wmf1/ [07:57:59] monobook is broken on en.wp [07:58:27] see -tech [08:00:02] sort of [09:09:24] guys http://it.wikipedia.org/w/index.php?title=Pagina_principale&action=render [09:09:48] appending it uncover pages that are supposed to be hidden [09:10:18] that's known I guess [09:10:25] right [09:10:39] your ordinary reader and most editors won't know to do that [09:10:49] they'll see the message, as intended [09:11:04] how long it's supposed to be there btw [09:11:17] https://www.facebook.com/pages/Rivogliamo-Wikipedia-No-alla-legge-bavaglio/185745561500946?sk=wall [09:11:32] the rumor mill says a day or so [09:11:46] right, then no fix is needed :) [09:12:02] yeah, it's a protest, not a permanent lockdown [09:12:20] there's a dicussion about it on meta and on foundation-l if you want to weigh in [09:12:34] where [09:13:01] got it [09:13:24] ok [11:39:21] petan: there's also http://it.m.wikipedia.org/ [13:19:50] Hey Reedy, feel like chasing some fatals? [13:19:59] 3 PHP Fatal error: Call to undefined method LocalFile::getFileKey() in /usr/local/apache/common-local/php-1.18/includes/api/ApiUpload.php on line 150 [13:20:04] 3 PHP Warning: Invalid argument supplied for foreach() in /usr/local/apache/common-local/php-1.18/extensions/LiquidThreads/classes/View.php on line 362 [13:20:12] 50 PHP Warning: array_map() [function.array-map]: An error occurred while invoking the map callback in /usr/local/apache/common-local/php-1.18/extensions [13:20:14] /OpenSearchXml/ApiOpenSearchXml.php on line 88 [13:20:23] (I'm ignoring errors in php-1.17 for now) [13:23:27] I will do in a few [13:23:38] RoanKattouw, that OpenSearchXML one happened around the 1.17 upgrade, then disappeared [13:23:52] I did try poking it last night, and made no difference, but it's seemingly intermittant [13:24:14] I recognize that warning [13:24:32] I filed a bug against PHP where I argue that that entire warning is stupid and should not exist [13:24:38] heh [13:24:40] Might be worth checking the exceptions log [13:24:54] Because that warning is thrown also if the map callback throws an exception [13:25:10] The exception propagates just fine, mind you, but there's no way to shut up the warning [13:25:17] https://bugzilla.wikimedia.org/show_bug.cgi?id=27452 is from last time, which i ended up resolving as WFM [13:25:21] (short of using @ which shuts up /all warnings and notices/ that occur in the callback) [13:25:37] Hm it's gone now [13:25:42] it'll come back [13:25:59] It's almost like there's an issue on an apache or 3 or something [13:26:02] I still see weird shit like [13:26:03] 1 PHP Warning: strrev() expects parameter 1 to be string, array given in /usr/local/apache/common-local/php-1.17/languages/classes/LanguageRu.php on line 105 [13:26:05] 1 PHP Warning: preg_match() expects parameter 2 to be string, array given in /usr/local/apache/common-local/php-1.17/languages/classes/LanguageRu.php on line 102 [13:26:06] But who cares that's 1.17 [13:26:22] hi [13:26:34] I had a quick look at some of the 1.17 ones, to see if they're likely still relevant [13:26:37] ohai hashar [13:28:31] Another weird one is 8 PHP Warning: preg_match() expects parameter 2 to be string, array given in /usr/local/apache/common-local/php-1.17/includes/Exif.php on line 1099 [13:30:28] Arrrgh, my first alumni email [13:30:56] hah [13:31:07] Though, it's somewhat relevant. IBM guest lecture [13:31:20] I got an e-mail from my student's association saying they'd automatically unsubscribed me [13:31:36] So I was like thank you, now I don't have to do that myself [13:31:50] 3 days later I get an invitation to their monthly meeting [13:32:01] (although I guess the latter was sent to the cs-all list) [13:34:09] 1 PHP Warning: rename(/tmp/53373aecc3eb11a8c5dc3e7fb92e4c15.png,/mnt/upload6/wikipedia/fr/math/5/3/3/53373aecc3eb11a8c5dc3e7fb92e4c15.png) [function.rename [13:34:11] ]: No such file or directory in /usr/local/apache/common-local/php-1.18/extensions/Math/Math.body.php on line 172 [13:34:24] 1 PHP Fatal error: Call to a member function wrapWikiMsg() on a non-object in /usr/local/apache/common-local/php-1.18/includes/specials/SpecialEmailuser.php on line 125 [13:40:54] Hmm [13:41:02] MWException wouldn't catch more generic exceptions.. [13:42:27] Rarrgh [13:42:34] *Reedy commits his working copy [13:45:45] hah, you got r99000 [13:45:50] Just thinking, if we weren't planning to move away from CR soon, being able to do TAG: 1.18, 1.18wmf1 in the commit summary would be nice [13:46:14] Lol [13:46:25] That's one done [13:46:32] Extrapolating that, I would expect r100000 to be reached during the NOLA hackathon [13:46:37] haha yeah [13:46:50] I've been wanting to autodetect "Revert r12345" forever as well [13:47:03] Would've been useful if I'd filed that when I thought of it, i.e. last year in DC [13:47:13] o_0 [13:47:24] Not worth the trouble now [13:47:27] The LQT view one is whinging about $_SESSION [13:47:39] Oh taht [13:47:48] Isn't $_SESSION being accessed in debugging code there? [13:47:59] *RoanKattouw also doesn't understand how the hell $_SESSION can be undefined [13:48:10] static function fixFauxRequestSession( $request ) { [13:48:10] foreach( $_SESSION as $k => $v ) { [13:48:10] $request->setSessionData( $k, $v ); [13:48:25] *Reedy moves on [13:49:48] 2011-10-05 13:48:24 srv189 enwiki: /w/api.php?action=userdailycontribs&user=Wiki Sysop&daysago=5 Exception from line 1342 of /usr/local/apache/common-local/php -1.18/includes/GlobalFunctio [13:49:48] ns.php: Internal error in ApiResult::setElement: Bad parameter [13:50:03] Ouch [13:50:07] userdailycontribs [13:50:11] I need moar screens [13:50:28] For a while there I was afraid we had a setElement error in core [13:51:57] Use WikiSysop :/ [13:53:40] Oct 5 13:53:28 10.0.8.5 apache2[1593]: PHP Fatal error: Call to a member function isAnon() on a non-object in /usr/local/apache/common-local/php-1.18/includes/api/ApiQueryUsers.php on line 262 [14:01:49] RoanKattouw, http://en.wikipedia.org/w/api.php?action=userdailycontribs&user=WikiSysop&daysago=5 [14:01:55] It's the bloody example url that fails :P [14:02:04] lol [14:02:15] Probably due to nonexistent user? [14:02:24] Cause https://en.wikipedia.org/w/api.php?action=userdailycontribs&user=Catrope&daysago=5 works [14:02:32] if ( !$user ) { [14:02:32] $this->dieUsage( 'Specified user does not exist', 'bad_user' ); [14:02:32] } [14:02:35] Hmm, no, https://en.wikipedia.org/w/api.php?action=userdailycontribs&user=klgdfjklghkhfjkhfjsfgjsf&daysago=5 [14:02:36] But it's hitting that [14:02:49] to eval! [14:02:53] ZOMG that user existed [14:02:57] https://en.wikipedia.org/w/api.php?action=userdailycontribs&user=klgdfjklghkfdafsdgdfghdhdfgghegerf4trt4erf4r2r233413423hfjkhfjsfgjsf&daysago=5 [14:03:02] That one does return a nosuchuser error [14:03:05] haha [14:03:13] the first had an id of 0 [14:04:09] $user = User::newFromName( 'WikiSysop' ); [14:04:11] That passes [14:05:34] So why do some usernames get a false positive [14:06:16] Oh, cause it can use it to save it [14:15:15] http://en.wikipedia.org/wiki/Special:CentralAuth/WikiSysop [14:26:33] Did you guys see http://commons.wikimedia.org/wiki/File:Unbenanntblafoosmiley.png ? nice bug :S [14:26:50] What's upw ith it? [14:27:01] Oh, it's been replaced [14:27:53] mm hmm [14:29:12] Few broken messages here and there [14:35:27] Well, that's the exception log a bit happier [14:52:40] apergos: Known bug? [14:52:50] not to me, no [14:53:01] I have been a bit out of the loop on image-related bugs for awhile though [14:53:14] had to reuse those brain cells for dump and/or revision related things [14:54:17] All those array_map errors happen accross the board at hte same time... [15:07:40] RoanKattouw, do narayam updates fall under the "Internationalization bug fixes/updates" deployments? [15:10:32] Apache error log has quietened down to the usual crap it seems [15:11:13] Software updates yes [15:11:21] At the discretion of the i18n people [15:11:30] They just give me stuff and I deploy it [15:11:35] <^demon> !log crap is within normal limits [15:11:35] --elephant-- Wrong channel! [15:11:46] That stuff usually includes a Narayam rev or two [15:11:48] <^demon> Oh shut up elephant2, that wasn't for you [15:56:40] Oct 5 15:56:27 10.0.8.51 apache2[6156]: PHP Catchable fatal error: Object of class AFPData could not be converted to string in /usr/local/apache/common-local/php-1.18/includes/GlobalFunctions.php on line 1197 [15:56:52] Oct 5 15:56:39 10.0.8.4 apache2[21331]: PHP Fatal error: Call to undefined method LocalFile::getFileKey() in /usr/local/apache/common-local/php-1.18/includes/api/ApiUpload.php on line 150 [15:57:03] Why does that look like things are out of sync? [15:57:26] <^demon> I tried tracking down the former last night, didn't get terribly far. [15:57:31] <^demon> Being out of sync is a possibility. [15:57:51] someone reported that AFPData one to BZ, apparently editing (foo) Bar did it [15:57:53] not for me.. [15:57:55] you saw the bug report for one of those right? [15:58:02] yeah, leading parens [15:58:10] I did a test edit on frwiki [15:58:11] I wonder if there has to be a close parens or not [15:58:14] worked fine [15:58:21] ugh [15:59:36] Yeah, 10.0.8.4 still has old code [15:59:38] *Reedy runs scap [16:02:58] FFS [16:03:06] WHY is 10.0.8.4 still out of date? [16:04:16] Because it's not responding to ssh? [16:04:21] Oh it is [16:04:36] yeah [16:04:45] I'm logged onto it looking at the apiupload file :P [16:05:00] I md5summed it [16:05:03] Looks up-to-date [16:05:16] http://www.mediawiki.org/wiki/Special:Code/MediaWiki/99003 [16:05:36] Whaat [16:06:10] Reedy: catrope@fenari:/home/wikipedia/common$ dsh -cM -g mediawiki-installation 'md5sum /apache/common/php-1.18/includes//api/ApiUpload.php' | grep -v 1be13ba4f124e0b760378b40ec6df865 returns nothing [16:06:17] Yeah [16:06:25] I svn upped and pushed the file earlier [16:06:41] Well you were alleging that srv254 was out of sync, right [16:06:45] yeah [16:06:49] i've just pushed it again [16:10:09] Oct 5 16:10:02 10.0.11.15 apache2[15670]: PHP Warning: rename(/tmp/a0c720e26fec1a9c6c050bce34717b18.png,/mnt/upload6/wikipedia/en/math/a/0/c/a0c720e26fec1a9c6c050bce34717b18.png) [function.rename]: No such file or directory in /usr/local/apache/common-local/php-1.18/extensions/Math/Math.body.php on line 172 [16:39:52] RoanKattouw: was the 1.18 upgrade everywhere successful? Is it safe for me to deploy those CentralNotice changes we discussed yesterday? [16:40:12] awjr, no [16:40:18] Most of the wikis haven't been upgraded still [16:40:42] is there anywhere i can easily follow along with progress so i dont have to keep bugging you? [16:40:55] Reedy ^ [16:41:05] Other than IRC/server admin log... [16:41:06] http://blog.wikimedia.org/2011/10/04/mediawiki-1-18-deployment-today-to-all-wikimedia-sites/ [16:41:11] "Update: October 5 05:14 UTC ? we didn?t get as far as we wanted, but we deployed to fr.wikipedia.org, pl.wikipedia.org, en.wikipedia.org, and commons.wikimedia.org. We?re planning to have one more deployment window in a little less than 18 hours (October 5, 23:00-October 6, 03:00 UTC) to deploy to the remaining wikis." [16:41:26] thanks Reedy! [16:47:38] awjr: did you hear from anyone on this triage? [16:54:00] Oct 5 16:53:44 10.0.11.31 apache2[22187]: PHP Fatal error: Call to a member function getText() on a non-object in /usr/local/apache/common-local/php-1.17/extensions/Collection/Collection.suggest.php on line 423 [16:54:06] oh, 1.17 [16:58:32] hexmode: no one's mentioned anything to me specifically - i forwarded on the announcement to fundraising-l, wikitech-l, internal-l, mentioned it to the civicrm community and posted in various WMF fundraising places [17:02:04] Anyone here for the fundraising bug triage? [17:02:36] *Jamesofur sulks [17:02:42] *pgehres is [17:02:56] cool! [17:03:02] lets start! [17:03:03] \0/ [17:03:22] http://etherpad.wikimedia.org/BugTriage-2011-10 [17:03:32] \0/ [17:03:59] (great way to spend my last 20 mins. in the WMF office) [17:04:07] :p [17:04:10] Special:FundraisingStatistics graph spills out of main content area [17:04:21] http://bugzilla.wikimedia.org/30820 [17:04:45] awjr: any comments about this? [17:04:53] that, i suspect, should be fairly straightforward to fix - and an important one to fix [17:05:14] particularly during the fundraiser, millions of people look at that page to check in on our progress [17:05:25] looks like it could be as simple as a CSS fix [17:05:37] that would be my guess [17:06:17] well, I'll make a comment on the bug [17:06:42] but otherwise, lets move on, unless someone wants to say "Me, me!" [17:06:53] wow. That page just took 33 seconds to be served. [17:07:06] siebrand: the etherpad? [17:07:12] siebrand perhaps file a new bug? [17:07:15] hexmode: special:fundraiserstatictics [17:07:19] it took a long time for me as well [17:07:24] :P [17:07:26] the results for that graph are supposed to get cached [17:07:30] special:fundraiserstatictics [17:07:31] Once upon a time, that thing had caching [17:07:33] damn, same typo twice. [17:07:35] yeah, another big thing we want to do for that is cache it. I don't think it's cached at all atm [17:07:38] Like, for an hour [17:07:48] Jamesofur: the results are cached, i believe [17:07:50] or at least are supposed to be. [17:08:00] I don't think they were last year :-/ I may be wrong though [17:08:21] awjr: if they were cached, then it shouldn't take me 30s, right? [17:08:27] The main reason it's wide is that there are many years in it. So is the solution to take years out of it, or make the bars less wide? [17:08:36] hexmode: more like <1s [17:08:54] hexmode: unless you're the first person to view the page after the cache expiry, you're right [17:09:01] im looking into the code now [17:09:16] and siebrand: you're right - it would be awesome if you could specify the years to display on that graph [17:09:22] in fact ithink that's part of another open bug... [17:09:32] awjr: force reload as logged in user took 1.9 seconds to be served. [17:09:40] k, it was pretty quick [17:09:51] yeah, the cache was probably stale [17:10:01] awjr: it may be that the query results were not cached and now they are? (not a popular page outside of fundraising season) [17:10:09] siebrand: likely [17:10:22] and yes - we have have another bug open about explicitly selecting years: https://bugzilla.wikimedia.org/show_bug.cgi?id=30818 [17:10:37] next bug? http://bugzilla.wikimedia.org/19895 -- Fold identical messages, or state that they should or could be different. [17:10:56] this one someone else added to the list - i am not familiar with it [17:11:10] I propose that a number of years is chosen by default that will fit a "normal" screen width, and that BA/PM considers if an option to add more years is wanted? :) [17:11:25] (that was on the previous issue that I'm done with now) [17:11:54] siebrand: adding it to the etherpad [17:12:12] awjr: I've no idea why I added 19895, but I did [17:12:14] :P [17:12:14] awjr: it came out of an i18n triage. [17:12:23] hexmode: ???? [17:12:31] siebrand: can you talk a bit about it? i dont know anything about it [17:12:51] Basically for ease of translation, it is preferred to reduce redundancy of messages, but *only* if the context is very similar or the same. [17:13:12] Apparently - I haven't verified - there allegedly is redundancy now, and the request is to remove that. [17:13:25] (so that translators can translate the extension(s) with less effort) [17:13:29] siebrand: when are you heading out? [17:13:38] AaronSchulz: in 8 mins or so. [17:13:40] hi [17:13:46] aude_: hey! [17:14:09] hmm I wonder if this is a donation Interface vs payflowpro gateway thing (the messages) [17:14:10] i dont know of anything actually using FundraiserPortal [17:14:22] i wonder if the fundraising extensions and civicrm stuff will be part of the wikimedia labs? [17:14:26] it looks like commits to it in SVN are pretty stale. [17:14:30] awjr: lemme check if we even support that. [17:14:45] have a test environment for developing? [17:14:46] aude_: what do you mean "part of"? [17:14:55] awjr: I know we removed a few previously developers and unused fundraising extensions from Translate [17:15:03] aude_: you could do it there, yes [17:15:06] ah ok [17:15:29] If we don't use the FR portal extension, this may be closed "Wontfix". [17:15:34] the barrier for volunteers helping w/ fundraising is higher because it's not just a mediawiki install w/ extensions and config used on wikipedia [17:16:05] siebrand: Hey so at some point I changed days-abbrev from 'd' to '$1 d'. Should I update all other languages in an automated fashion? Cause this is now causing a minor bug on at least plwiki [17:16:26] Or is this something TWN will handle? [17:16:40] (The change happened over a week ago, I was gonna poke you but forgot) [17:16:42] awjr: we still support FR portal extension. Is it still used? [17:16:58] RoanKattouw: in theory, that's something Raymond_ would do on change. [17:17:05] aude_: and we have a hack-a-thon in a 1.5 weeks with Ryan_Lane, so that would be a good time to talk about that [17:17:09] not that i am aware of - Jameofur, do you know anything about that ^ [17:17:16] RoanKattouw: have to go and travel. Best create a bug and assign it to me? [17:17:24] OK [17:17:24] aude_: you going to NOLA, by chance? [17:17:27] Will also CC Raymond [17:17:45] RoanKattouw: aye [17:17:51] hexmode: i won't be there :( [17:18:00] meh, i think we can safely mark this as wontfix. considering i've never heard of it, i'd be very surprised if it's still being used (At least for WMF fundraising) [17:18:03] ok, next issue. Jamesofur and I will look into this one. [17:18:05] *aude_ wishes for another hackathon in DC [17:18:30] Jamesofur: FYI: - ext-fundraiserportal # 2009-10-08: test.wikipedia.org [17:18:39] aude_: I'll try to get this brought up while I'm there. We want to make volunteer involvment easy [17:18:40] Jamesofur: probably never gone to production. [17:19:09] hexmode: thanks [17:19:10] 'wmgUseFundraiserPortal' => array( [17:19:10] 'default' => false, [17:19:12] 'testwiki' => true, [17:19:13] 'aawiki' => true, // ? [17:19:13] yeah looking into it. I think it was intended as a broader system that never really ended up getting used after we chose PayflowPro [17:19:13] ), [17:19:25] from InitialiseSettings [17:19:32] aawiki? alocoholics anonymous wiki? [17:19:34] aawiki? zomg [17:19:47] ok im gonna mark that one wontfix [17:20:20] siebrand: got to go? [17:20:42] glad you could help us with that one [17:20:59] next bug? http://bugzilla.wikimedia.org/23599 -- Add support for date range in ContributionTrackingStatistics [17:21:37] siebrand, I figured out what it is [17:21:46] it's that giant donate button on the side of test wiki [17:21:48] Jamesofur: oh! Be quick [17:21:51] we certainly don't need that [17:21:58] it's been sitting there for ever [17:22:00] *siebrand waves and heads for SFO [17:22:06] CYA! [17:22:08] *awjr waves back [17:22:12] by siebrand!!!! it was great seeing you in SF [17:22:15] travel safe [17:22:25] hexmode: https://bugzilla.wikimedia.org/show_bug.cgi?id=31382 is an awful bug, maybe you can find someone to fix it? :-) [17:22:40] multichill: looking [17:22:50] so bug 23599, this is the page it references: http://wikimediafoundation.org/wiki/Special:ContributionStatistics [17:23:14] it's intedned to give some high level contribution stats for the fundraiser [17:23:27] And we're workin on a survey for WLM. Who operates the WMF survey server? (we're supposed to have one) [17:23:30] multichill: after this triage I'll be chasing down people to fix 1.18 bugs, so yeah [17:23:33] it would be pretty slick if you could feed it a date range and get corresponding stats back [17:23:42] Thank you hexmode [17:23:52] How to tag these 1.18 related bugs? [17:23:53] multichill: check -operationss? [17:23:56] multichill: Right now I think that's no one because we found some pretty nasty security vulnerabilities in LimeSurvery [17:24:19] Right, so we'd better use an external survey tool? [17:25:23] multichill: 29876 is the bug 1.18 problems should block [17:25:47] multichill, We're going to get a new one but it is likely going to be a month or so (for the survey box) at least. If you want it earlier an external group is probably going to be better that's what the fundraiser is doing [17:26:20] Anybody have any other thoughts/interest in bug 23599? [17:26:24] *AaronSchulz digitally waves brion_ [17:26:48] Hi all, sorry I'm late - my clocks are playing up [17:27:12] Ok. Thanks guys [17:27:14] TheCavalry: I know the feeling [17:27:15] hexmode: perhaps we should move on [17:27:19] TheCavalry, You're here! That's what matters :) http://etherpad.wikimedia.org/BugTriage-2011-10 [17:27:21] http://bugzilla.wikimedia.org/23599 -- Add support for date range in ContributionTrackingStatistics [17:27:52] greetings, TheCavalry [17:28:40] so bug 23599 is essentially a pretty straightforward feature request for ContributionStatistics [17:28:47] *hexmode tries to get his doppelganger on the etherpad [17:28:51] http://wikimediafoundation.org/wiki/Special:ContributionStatistics [17:29:27] while i think it would be relatively trivial to build the feature in, the challenge will be in performance [17:30:31] this is definitely a good one for anyone who's interested in a nice intro to hacking MW for something to run on the cluster [17:30:47] you'll have to think about caching, dealing with concurrency issues, etc. [17:31:03] imho, fun and straightforward :) [17:31:08] heh [17:31:18] any takers? [17:31:33] but if we have no takers, lets move on to the next one... we're 30min in [17:31:38] http://bugzilla.wikimedia.org/24830 (Normal) -- International addresses not displaying and/or saving correctly [17:32:20] this is another fun one, although it's a lot nastier [17:32:51] this one relates to country/region data that comes with CiviCRM [17:33:07] awjr: CiviCRM, not Mediawiki, right? [17:33:10] correct [17:33:35] nearly all of the countries in the CiviCRM database relate to associated states/regions [17:33:46] aude_: you have any interest in CiviCRM? [17:34:18] 47 of those countries, however, have state abbreviations that are DIGITS rather than letters [17:34:22] http://etherpad.wikimedia.org/EngineeringRoadmap [17:34:27] ? [17:34:32] digit abbr? [17:34:37] ^ i guess tom's not in here nvmind :D [17:34:40] weird, right? [17:34:45] hella [17:35:09] awjr: is that just the way it is, or is that a CiviCRM bug? [17:35:35] i had looked into this last year and it seems like they had just sourced data from an iso 3166 specification that is not quite what was expected [17:35:58] it is a CiviCRM issue that has yet to be resolved [17:36:10] so civicrm bug, then, right? And no other ngo cares? amazing [17:36:17] most users of CiviCRM are nowhere near as global as we are [17:36:28] awjr: is there an upstream bug report? [17:36:41] and to be fair, it's a pretty minor bug as it only affects the display of regional abbreviations of 47 countries [17:37:01] i had posted in their forum last summer [17:37:07] http://forum.civicrm.org/index.php/topic,15431.new.html [17:37:10] how many countries are there in civicrm? 200? [17:37:23] *Jamesofur makes a note to kick people during the civi meetup [17:37:24] not sure of fth etop of my head, but probably around there [17:37:32] awjr: I'll make sure a bug is tracked in their db [17:37:47] fwiw, we are now using civicrm 3.4, not 3.1 [17:37:52] thanks hexmode [17:38:31] but that bit seems like a data sanitation task that some people might like [17:38:45] anyway, I've only got 22 more min [17:38:57] http://bugzilla.wikimedia.org/25430 (Normal) -- Fundraising-related Special pages need header and footer messages [17:38:57] [17:39:11] so that one ^^ [17:39:24] this one im not totally clear on - i was hoping MZ would show up and explain :) [17:39:39] Jamesofur, do you know anything about that one? ^ [17:39:49] *Jamesofur looks [17:39:53] awjr: think he is around more at night [17:40:55] yeah he is sadly. It's tbh less important I think since we really only use 1 of those pages now but it WOULD be really nice to put a simple mediawiki message on top and/or bottom of all of them [17:41:00] so that we can have headers like that [17:41:05] makes sense [17:41:05] I think that would be the simplest way [17:41:23] right now it's basically all hard coded as far as I've been able to tell [17:41:50] it also would allow us to put in links to things like Meta stats pages etd [17:41:51] *etc [17:42:26] Jamesofur: are you saying all tha tneeds to be done is plop in messages for header/footert/etc for those pages? like an empty message? [17:42:29] so, def easy, and should be done! [17:42:38] awjr, Yup [17:42:55] ok so this is probably the easiest of all of the fundraising-related bugs [17:43:01] heh [17:43:05] agreed [17:43:08] so far [17:43:33] http://bugzilla.wikimedia.org/25468 (Normal) -- Add a search box to Special:ContributionHistory [17:43:55] oops [17:43:58] wontfixed [17:44:19] oh yeah, that is no longer used [17:44:23] yeah [17:44:29] http://bugzilla.wikimedia.org/25608 (Normal) -- MediaWiki:Fundraiserstats-tab-ytd should not contain (USD) [17:45:10] damn i wish siebrand were still here [17:45:27] Nikerabbit: here? [17:45:47] or did he leave the same time? [17:46:04] i imagine he left at the same time [17:46:17] it looks like that message gets set in ContributionReporting/ContributionReporting.i18n.php [17:46:40] Sadly I think he did yeah :( [17:47:06] however, i'm not sure that particular message is getting used [17:47:10] awjr: code review link? [17:47:25] *aude_ scrolls up [17:47:51] awjr: that could easily be customized on-wiki, then, if needed. [17:48:23] hexmode: you're right [17:48:58] 12 more min. 5 more bugs [17:49:03] http://bugzilla.wikimedia.org/26387 (Normal) -- Add better key to Fundraiser stats page [17:49:21] for bug 24830, you can populate postal abbreviations with data from naturalearthdata.com [17:50:05] aude_: can you make that comment on the etherpad? [17:50:24] aude_: nm [17:50:55] it is possible that kaldari began work on bug 26387 once upon a time and never finished [17:51:08] there's potentially some magic for that key in trunk [17:51:15] heh [17:51:18] but if there'snot, this is likely another pretty straightforward fix [17:51:41] it's just adding a key to the stats graph: http://wikimediafoundation.org/wiki/Special:FundraiserStatistics [17:53:48] 7 more min [17:53:50] Bug 28903 - Make banner translations get deleted when banners are removed automatically [17:54:03] https://bugzilla.wikimedia.org/show_bug.cgi?id=28903 [17:54:08] jpostlethwaite: can you comment on that bug ^ [17:54:26] awjr: looking [17:55:45] that was one of the first bugs assigned to me. I have not done any work on it. [17:55:56] are you able to give any background? [17:56:01] jpostlethwaite ^ [17:56:16] jpostlethwaite: any idea on what needs to be done to fix it? [17:56:22] nothing more than is in the comments. [17:56:38] it needs to do a mass delete [17:56:56] jamesofur just went afk but i think he has more background [17:57:13] how are the translations stored? /lang? [17:57:21] *hexmode looks to see if it says [17:57:32] like: http://meta.wikimedia.org/wiki/MediaWiki:Centralnotice-2011MobileReaders-share/de [17:57:59] so you need to match the base url on the delete [17:58:00] right... so, iirc, Translate already does that [17:58:13] with a wildcard, would be my guess. [17:58:15] but we'd need to ask Nikerabbit [17:58:29] yeah, currently when we delete a banner it deletes the banner automatically along with the english messages but all the translations, if they exist, remian [17:58:36] and end up as just orphaned pages [17:58:40] and maybe use Translate for translation? [17:58:44] (that can be deleted manually but very tough) [17:59:02] i don't' know where/if it's reported, but ... i'm always annoyed that the banner text often overflows the banner on small screens [17:59:03] i think it was agreed, that it would not prompt for deleting all translations [17:59:18] *aude_ thinks i reported it last year [17:59:21] quite often [17:59:24] well Translate would require that those messages be in i18n files - but the banners are generated essentially as wiki pages [17:59:25] it should just delete all children since the parent has been removed [17:59:48] and the translation happens by hand [18:00:04] unless there's Translate magic i dunno about? [18:00:09] awjr: Translate is going to "maybe" be deployed on meta and used for the translation there [18:00:15] yes, magic [18:00:17] oo [18:00:24] do you know anything about timeline? [18:00:43] RoanKattouw needs to finish reviewing it [18:00:57] hexmode: *Reedy* needs to [18:01:02] We changed assignments :) [18:01:03] heh [18:01:06] heh [18:01:07] ok [18:01:11] font size is hard-coded i think for banners? [18:01:19] hexmode: RoanKattouw: Reedy, could you join for the 1.18 call? [18:01:44] robla: 1m and I'll be on [18:01:44] aude_: im not sure but i don't think so [18:01:51] awjr: i [18:01:51] Sure, sec [18:01:52] where are we? [18:02:00] awjr: i've got to go now [18:02:06] awjr: here http://meta.wikimedia.org/w/index.php?title=Special:NoticeTemplate/view&template=B_3009_Maryana_A_US [18:02:09] for example [18:02:18] ok hexmode, thanks for coordinating this :) [18:02:33] it always (almost?) gets hard coded and doesn't take into consideration screen size, which jquery could do [18:03:17] most of these banners: http://meta.wikimedia.org/wiki/Special:NoticeTemplate don't look so bad though [18:03:18] aude_: that would be good to talk about in #wikimedia-fundraising [18:03:39] ok [18:03:41] most of the folks who code those banners and no a lot more than i do about them hang out in there [18:03:52] robla: http://ganglia.wikimedia.org/graph.php?c=MySQL&h=db40.pmtpa.wmnet&v=406.493&m=disk_free&r=week&z=medium&jr=&js=&st=1317836378&vl=GB&ti=Disk%20Space%20Available [18:04:12] robla: http://ganglia.wikimedia.org/graph.php?c=MySQL&h=db40.pmtpa.wmnet&v=406.493&m=disk_free&r=month&z=medium&jr=&js=&st=1317836378&vl=GB&ti=Disk%20Space%20Available [18:09:14] http://etherpad.wikimedia.org/118deployment-scrum [18:09:31] https://bugzilla.wikimedia.org/show_bug.cgi?id=31382 [18:17:43] hexmode: http://ganglia.wikimedia.org/graph.php?c=MySQL&h=db40.pmtpa.wmnet&v=406.493&m=disk_free&r=day&z=medium&jr=&js=&st=1317836378&vl=GB&ti=Disk%20Space%20Available [18:20:11] http://commons.wikimedia.org/wiki/File:ReedyTestImage.png [18:29:09] RoanKattouw: just so I'm clear, we want that ganglia graph going up, not down, right? [18:29:17] Aye [18:29:21] It's a graph of free space [18:29:37] right [18:29:44] *RoanKattouw 's query is still running [18:29:52] *hexmode actually took time to read the title now [18:30:05] I'm running a query to see how much time is taken up by dead data that's left over from previous incarnations of the iOS hack [18:30:09] Which can be safely removed [18:30:20] k [18:30:32] and that may clear it all up? [18:30:42] *hexmode goes off to file a bug [18:30:44] Don't know how much of a dent it'll make yet [18:30:56] sure [18:39:50] Reedy: are you looking at any particular talk page atm? I'm gonna scan enwiki to be sure. [18:46:42] hexmode, not really. I was gonna go over the bugs in the last 24 hours, and put anything that's been "broken" by 1.18wmf1 onto the 1.18 post deployment cleanup thingy bug [18:47:11] Reedy: sure... I'll make sure nothing is missing from Talk pages [18:47:29] Reedy: 29876 is the cleanup bug, right? [18:47:38] !b 29876 [18:47:38] --elephant-- https://bugzilla.wikimedia.org/show_bug.cgi?id=29876 [18:47:52] yeah [18:53:53] Computers being sucky [18:53:54] brb, restart [19:06:31] So, hey. [19:06:57] Thoughts on the "Coding Challenge". What do you guys think about rebranding the entire thing from being a "combat metaphor" to a "super hero" metaphor? [19:12:31] Bugzilla is too busy [19:35:02] Reedy: now you know my pain [19:41:35] looking for me hexmode? [19:42:13] Nikerabbit: that was back during triage... let me see if I can see what we wanted to know [19:43:53] *hexmode pasted "nikerabbit ... irc bit" at http://paste2.org/p/1691366 [19:44:33] i can't open links with my phone [19:45:37] Nikerabbit: so, basically, it was about fundraising messages and translateion [19:46:13] Nikerabbit: Translate will delete translated messages when the main one is deleted, right? [19:46:22] it depends [19:46:52] you can delete a wholde page including all of its translations [19:47:22] Nikerabbit: translations are generally sub-pages, right? [19:47:31] guillom: Hey was there a CentralNotice about 1.18? enwiki people are kind of pissed off we didn't warn them other than posting to the tech blog [19:47:52] There was [19:47:52] (IIRC you said that, Nikerabbit) [19:48:04] *RoanKattouw wonders when and where [19:48:07] hexmode: always [19:48:23] RoanKattouw, it was in central notice last night during the deploy [19:48:31] Right, but not ahead of time? [19:48:36] Pass [19:48:46] Enwiki just like the drama [19:48:52] *Nikerabbit is surrounded by screaming children [19:49:01] RoanKattouw: i said what? [19:49:14] Nikerabbit: ?!? [19:49:19] I wasn't talking to you at all :) [19:49:45] oops, hexmode [19:49:51] heh [19:49:54] too small screen [19:49:55] hexmode Nikerabbit: translations are generally sub-pages, right? [19:50:20] yep [19:51:40] RoanKattouw, there was a CentralNotice ~15 min before the deployment. Also, enwiki people are pissed when we use the banner for anything else besides fundraising. So, enwiki people should make up their mind. [19:52:17] hah [19:52:21] Apparently it was posted to VP/T [19:52:29] By happy-melon [19:52:34] And then people complain they don't watch VPT [19:52:50] can't please everybody [19:52:54] Maybe we start notifying every user page on every wiki about upcominmg deployments [19:52:55] Of course these are the people that use upgrade in quotation marks [19:53:01] That sounds like a useful use of time [19:53:04] I was gonna say something about 800 village pumps [19:53:10] But yeah why not every user talk page on every wiki [19:53:19] Seriously, though, CN is the best vehicle for this [19:53:23] Reedy: how are you doing with the review? [19:53:30] And it sounds like it might be better to run it for a few days pre-deployment [19:53:32] on balance [19:53:42] I've not touched it due to 1.18 cleanup [19:54:14] how long will that take? [19:54:33] I was still up 4am this morning.. [19:56:39] is that synonym for loong? [20:01:30] jumpy Jamesofur [20:18:46] *Jamesofur grumbles. Stupid wifi [20:21:40] ugga [20:23:47] bugga [20:24:30] Nikerabbit, in SFO ? [20:47:08] Hi there [21:04:30] Hi there, Has anyone got time to discuss the in's and out's of a QWiki ?? [21:06:03] I don't think that's anything to do with us is it? [21:10:11] Well I was chatting to 'sumanah' last night and he said to downlad a book : A Beginner's Guide to Media Wiki : http://www.packtpub.com/mediawiki-1-1-beginners-guide/book , and I did and I am now at the stage where I am commenting on the PDF (using NitroPDF) and I have come to abit of an impass. ... Would you like to have a copy of what I have got so far (With sticky-notes ) ??? [21:19:57] heading bed have a good day! [22:07:04] Morning Tim [22:07:14] morning [22:19:10] "Wasn't there any testing done on this release? It's certainly introduced a load of problems that ought to have been caught by any proper test plan." [22:19:12] Right. [22:19:40] Heh [22:19:44] Isn't that always the case? [22:19:57] That and they had the oppurtunity to help test and didn't bother? [22:30:52] gar [22:30:56] ^ frustration [22:31:08] Wikipedians? [22:31:22] guillom: can you link? [22:32:01] sumanah, I don't have it handy, but it should be at [[:w:en:WP:VP/T#1.18 issues]] [22:32:51] The obvious answer is "no testing plan will ever replace production". [22:37:28] awjr: Still about? [22:37:39] yup [22:37:43] Myra: waht's up? [22:37:49] awjr: Re: https://bugzilla.wikimedia.org/show_bug.cgi?id=25430 [22:38:03] There should be MediaWiki messages at the top and bottom of the pages. [22:38:14] So that it's possible to write a short blurb about what the purpose of the page is, etc. [22:38:27] http://wikimediafoundation.org/wiki/Special:ContributionStatistics <-- This page has no intro. [22:38:34] The defaults can be blank, of course. [22:38:35] ah yes - Jamesofur helped clear that up [22:38:43] Ah, cool. [22:38:48] seems like a relatively quick fix [22:38:55] guillom: even more impressive, from the same page: "This is, by far, the worst corruption of the English Wikipedia since the dawn of time. It is a good time to talk about stopping future updates to Wikipedia: it can no longer function when updating to MediaWiki 1.18 or such." [22:39:09] Yeah, I can't imagine it'd take more than a few minutes. [22:39:09] we didn't really seem to get any takers from the bug triage though [22:39:42] i'm going to send out a report from the triage tonight or tomorrow and i'll make sure to make clear that it should be relatively trivial to implement [22:41:35] sumanah,instinctively, I would take this as a joke and smile. Then I'd realize the person is serious, and I'd LMAO. [22:41:46] I KNOW [22:41:49] IT'S LIKE, WHAT [22:42:15] sumanah: did you see my earlier message ? I'll check back in 5 mins. [22:42:43] mcbaine1: hi. where and when did you leave/send this message? In IRC, email, or other media? [22:44:32] TimStarling: Reedy: AaronSchulz__ : we have to make a go/no-go decision on 1.18 to the rest of the wikis here. I'd say "go", but I'm wondering if there's anything you see that should stop us [22:45:23] here's the list from BZ that might stop us: https://bugzilla.wikimedia.org/buglist.cgi?bug_id=27478%2C31388%2C30961%2C27720%2C28348%2C31374%2C31361%2C31383%2C31197%2C31379%2C30786%2C31362%2C31391&list_id=32459 [22:45:42] I was going to ask you about putting the banner up or not, robla :) [22:45:56] my vote is to go [22:46:22] *robla looks at https://bugzilla.wikimedia.org/show_bug.cgi?id=31383 [22:46:39] I can do a quick hack for bug 31379 now if we need to move fast [22:46:39] meh nm on 31383 [22:46:41] Half of that list isn't relevant [22:46:55] I was just going to reproduce and test it [22:47:57] *robla confirms AaronSchulz__ is ready to go [22:48:06] What's the plan for what to upgrade when? [22:48:16] obviously we don't want to switch 800 wikis at once [22:48:20] guillom: let's put the banner up [22:48:26] robla, ok [22:49:04] robla, I'm assuming we're going to do quite a bit of wikis at once? (checking, since the banner will show on all wikis) [22:49:28] Or, grouping them by family, or language? [22:49:47] Reedy: guillom: plan is this: let's do dewiki first, then arwiki (biggest RTL wiki) then see where we are [22:50:08] at some point, we'll have to do a big bulk migration, but I think we should maybe get the top 10 done before we do that [22:50:08] maplebed finished the dewiki DB stuff, right? [22:50:09] robla, ok, then I'll enable it in German and Arabic only [22:50:19] Reedy: yup [22:50:25] That way the false positive will be limited. [22:50:40] Or, actually, inexistent. [22:51:50] TimStarling: don't rush a hack out on 31379....I think you'll probably have a little time to fix it right while we're waiting for the dust to settle [22:51:52] Ok, maintenance banner should be up on de.wp and ar.wp shortly. [22:52:10] !log reenabling payments4 as mysqlhost in $wgDBservers on payments1 [22:52:10] --elephant-- Wrong channel! [22:52:21] :( [22:53:37] lol [22:55:34] elephant2 gets the smarty-pants-bot award. [22:58:11] sumanah:do you want me to copy and paste the earlier IRC message ?? check back in 5 mins. [22:58:56] sumanah: do you want me to copy and paste the earlier IRC message ?? check back in 5 mins. [22:59:04] mcbaine1: I have no idea what you're talking about, so, yes please. and let's move to #mediawiki please [23:00:16] Can we fix dewiki then? [23:04:43] Reedy: ? [23:04:56] I was going to say "Can we go and break dewiki?" [23:05:00] but that seemed a bit negative [23:05:49] yeah: AaronSchulz__, want to pull the trigger? [23:06:12] TimStarling: speak now or forever hold your peace [23:06:51] I guess [23:07:04] I wish the box peeped when I got pinged [23:08:11] Be careful what you wish for; or don't leave your laptop unattended :) [23:09:02] *hexmode likes how domas flatlined https://bugzilla.wikimedia.org/31388 [23:28:49] *robla has had lots of connectivity problems for the past 15 min or so [23:28:55] which is ossm