[08:03:53] <Nemo_bis>	 Wikinaut: solved? works for me now
[08:04:45] <Wikinaut>	 yes
[08:05:21] <Wikinaut>	 solutions: i) reboot of instances was needed ii) Ryan needed to reboot the proxy machine
[08:05:24] <Wikinaut>	 solved now
[08:05:26] <Wikinaut>	 over and out
[08:37:47] <Nemo_bis>	 sigh, all of this is still relevant http://thread.gmane.org/gmane.science.linguistics.wikipedia.technical/49535/
[08:49:57] <ori-l>	 Nemo_bis: well, let's see:
[08:50:33] <ori-l>	 "In community-based development, anyone who's willing to write good code can get it submitted and included into the product." Check. We don't do an amazing job, but better than a year ago.
[08:51:48] <ori-l>	 "Some employees apparently have shell access for the sole purpose of syncing their own code without going through the normal review process.  No volunteer has been given such access, to my knowledge -- indeed, AFAIK it's been years since any non-employee has been given shell access at all."
[08:52:42] <Nemo_bis>	 that was a definition, not a suggestion
[08:52:46] <ori-l>	 The former is quite firmly against the rules now and en route to becoming technically impossible as we move toward continuous deployment
[08:53:07] <apergos>	 review is done differently and volunteers are getting +2 so this is improvement
[08:53:36] <apergos>	 (just peeking in)
[08:53:43] <Nemo_bis>	 again this wasn't one of the recommendation
[08:53:45] <ori-l>	 and we added a bunch of volunteers to core
[08:54:15] <ori-l>	 hi apergos
[08:54:20] <apergos>	 hello
[08:54:24] <Nemo_bis>	 it's very dubious that git increased the ability of volunteers to get stuff merged
[08:54:28] <ori-l>	 There are a lot of paid developers who I've never seen in either #mediawiki or wikitech-l." Like who?
[08:54:42] <apergos>	 they might be in wikimedia-dv
[08:54:44] <apergos>	 *dev
[08:54:56] <apergos>	 but I would expect folks to be in one or all of those channels
[08:56:05] <ori-l>	 "There's a secret staff IRC channel" -- it's not secret and it's so terrifyingly boring that I /parted a while ago and haven't missed it much.
[08:56:34] <ori-l>	 I have never seen engineering decisions made during my time in there, either
[08:58:27] <apergos>	 no, it's not for anything technical
[08:58:54] <apergos>	 it is a digital substitute for remote workers who can't be in the office
[08:58:56] <ori-l>	 "Maybe #wikimedia-dev should be renamed to #mediawiki-dev" <-- good idea and not done, heh
[08:59:13] <Nemo_bis>	 ori-l: it's not secret?
[08:59:18] <apergos>	 well then #mediawiki and #mediawiki-dev really are one and the same
[08:59:37] <apergos>	 but I always thought we could have two channels instead of 20 :-P
[08:59:46] <apergos>	 (#mediawiki, #wikimedia-tech)
[08:59:57] <Nemo_bis>	 maybe it has no +s and yes everybody knwos it exists (since that email at least ;) ) but it's restricted to staff so it definitely matches most definitions of "secret"
[09:00:05] <ori-l>	 well, the distinction roan drew in his reply was that #mediawiki-dev is not for generic user support or bots
[09:00:14] <Nemo_bis>	 as a reminder, we have 769 open commits (615 unreviewed), 80 % of them are from non-WMF devs
[09:00:39] <apergos>	 I see
[09:00:59] <ori-l>	 Nemo_bis: it's not a great figure, no
[09:01:09] <Nemo_bis>	 nobody still produced any statistic showing that git allowed more volunteer devs or more patches from them get in
[09:01:46] <Nemo_bis>	 and the gigegat migration never was meant to increase volunteer partecipation, but rather less painful releases
[09:02:05] <Nemo_bis>	 so please let's forget git/gerrit in this discussion, it's totally unrelevant and useless
[09:02:24] <ori-l>	 you're the only one who brought it up
[09:02:52] <Nemo_bis>	 it what?
[09:03:00] <ori-l>	 git/gerrit
[09:03:05] <Nemo_bis>	 no, you did
[09:03:31] <Nemo_bis>	 ori-l> and we added a bunch of volunteers to core
[09:04:07] <apergos>	 if we said 'we added a bunch of people with commit access to trunk' it would be the same think
[09:04:09] <apergos>	 *thing
[09:04:14] <apergos>	 the tool doesn't matter
[09:04:17] <apergos>	 for this part.
[09:04:19] <Nemo_bis>	 except perhaps the status of this channel and (in a very small part) the "Consider what to do about code review" point, on all the other points we went in the opposite direction
[09:04:44] <Nemo_bis>	 apergos: it matters because it's omissive of the part where you *removed* access to volunteers :)
[09:04:58] <ori-l>	 it feels to me like you're trolling a little, tbh. but i agree, there are problems, and we could do better.
[09:05:24] <apergos>	 afaik we converted over eeryone's svn accounts as best we could, hunting down emails when there were none etc
[09:05:34] <Nemo_bis>	 then I'm not saying those recommendations were the good ones, just that they were not addressed and all applies to current situation even more than in 2010
[09:05:40] <apergos>	 where by we I mean demon I guess, not claiming that I did any of that work
[09:06:30] <Nemo_bis>	 ori-l: why trolling? :) I only said we have unsolved and unaddressed problems, I'm not aware of anyone seriously claiming we don't
[09:07:11] <ori-l>	 not conceding apergos's points; speaking in sweeping terms ('all of this is still relevant')
[09:07:54] <Nemo_bis>	 apergos: sure but that's useless if code doesn't get merged, it's hard to deny that it's a reduction of access... of course I have no stats on whether people previously having SVN access now have an average chance to get stuff merged
[09:08:10] <Nemo_bis>	 ori-l: why is that sweeping?
[09:08:19] <ori-l>	 some of it is still relevant. it's a good e-mail, at any rate, and i wasn't aware of it, so thanks. why don't you go over it, update it for the situation as of march 2013, and email it to the list again?
[09:08:49] <Nemo_bis>	 because I'm not competent enough to do so
[09:08:57] <Nemo_bis>	 because I don't see anything to update
[09:09:09] <Nemo_bis>	 and finally because when someone else did something like that it was a waste of time
[09:10:06] <apergos>	 it's not useless; because we have a blocker at a later stage doesn't mean it's useless to have made progress in earlier stages of a process
[09:12:59] <Nemo_bis>	 I don't see any such process
[09:13:47] <ori-l>	 Nemo_bis: you know, it's a very good thread. Susan has also pointed me to some good ones in the past, but I think it would be valuable to have a best-of-wikitech somewhere.
[09:14:23] <Nemo_bis>	 But again, I don't know what are the solutions, I'm not able and it's not my duty to find them. I can only see that the problems are the same, and move forward on doing something else.
[09:15:12] <Nemo_bis>	 ori-l: btw I bumped into that link in a 2010 it.wiki discussion where someone asked about AFT and someone else predicted it was going to be a failure, citing that post.
[09:15:34] <Nemo_bis>	 I was on the optimistic side at a time, so I'm not biased. ;-)
[09:15:39] <Nemo_bis>	 *at the time
[09:20:47] <ori-l>	 Nemo_bis: http://www.mediawiki.org/wiki/Wikitech-l/best_of
[09:21:26] * Nemo_bis  starts bikeshedding on the page location :p

[09:21:47] <ori-l>	 feel free to move it
[09:21:56] <ori-l>	 I don't have strong feelings about that
[09:22:14] <Nemo_bis>	 I was kidding
[09:22:27] <Nemo_bis>	 thanks for starting it, I'll think of adding something maybe :)
[17:29:06] <chrismcmahon>	 andre__: is there a way to have BZ show "Importance" in the search result not just as colors but as a sortable column?  I'm looking through preferences and not finding that.
[17:29:58] <andre__>	 chrismcmahon, bottom of buglist.cgi
[17:30:39] <chrismcmahon>	 andre__: got it, thanks!
[17:56:41] <chrismcmahon>	 andre__: do you make decisions about changing to WONTFIX status?  I'm thinking of https://bugzilla.wikimedia.org/show_bug.cgi?id=34255.  I'd like to resolve some Search bugs if I can.
[17:57:35] <andre__>	 chrismcmahon, if it is "we should not fix it" or "this does not fit into the plans / design" it's WONTFIX. else it's lowest priority.
[17:58:07] <andre__>	 chrismcmahon, also, if just one person has seen the problem, I'd reset to UNCONFIRMED status, and add a "testme" keyword
[17:58:23] <Nemo_bis>	 "I'd like to resolve some Search bugs if I can." <--- oooh, we love you!
[17:58:38] <chrismcmahon>	 andre__: OK, 34255 seems to be clearly in the "this does not fit" category, I'll WONTFIX it.
[17:59:34] <chrismcmahon>	 Nemo_bis: it
[17:59:40] <chrismcmahon>	 Nemo_bis: it's time :-)
[17:59:41] <fabriceflorin>	 Hello mlitn : Look forward to our AFT5 deployment on English Wkipedia today, when we will disable ArticleFeedbackv4, as well as switch to an opt-in version of ArticleFeedbackv5, as outlined in this Bug 45538: https://bugzilla.wikimedia.org/show_bug.cgi?id=45538
[17:59:59] <Nemo_bis>	 chrismcmahon: too bad though, it's a legitimate wish
[18:00:47] <andre__>	 chrismcmahon, probably for everybody (me included to some extend), the existing Bugzilla components "MediaWiki > Search", "MediaWiki Extensions > MWSearch",  "MediaWiki Extensions > Lucene Search", and "Wikimedia > lucene-search-2" create some confusion. I am aware of https://wikitech.wikimedia.org/wiki/Search but still have to read it
[18:01:00] <chrismcmahon>	 Nemo_bis:  true, but not WMF's area afaict
[18:01:12] <Nemo_bis>	 chrismcmahon: but that's not the WMF search component
[18:01:16] <andre__>	 so if anybody can explain what is what in one sentence, I could add clarification to each component.
[18:01:24] <chrismcmahon>	 hrm, thus I ask
[18:02:52] <chrismcmahon>	 andre__: I simply became discouraged by 117 open bugs in MediaWiki->Search, many of which seem like noise.
[18:03:18] <andre__>	 and I highly appreciate help :-/
[18:04:08] <andre__>	 I am wondering if turning the meaning of UNCONFIRMED into "not reproducible" (even when something clearly has happened) would help
[18:09:27] <chrismcmahon>	 andre__: I'm looking at those other 3 components.  Do you know the expression "can of worms"?  :-)
[18:10:28] <andre__>	 chrismcmahon, yes.
[18:10:33] <andre__>	 it is.
[18:14:55] <kaldari>	 parent5446: https://gerrit.wikimedia.org/r/#/c/48995/
[18:19:41] <hashar>	 kaldari: hey :)   I am coming to 6th in a few.  Have to arrange a meeting time with someone else.
[18:21:07] <gerrit-wm>	 Change merged: Krinkle; [integration/jenkins-job-builder-config] (master) - https://gerrit.wikimedia.org/r/52195
[18:37:02] <parent5446>	 kaldari: Mhm, will get to it in a sec.
[18:49:56] <mwalker>	 kaldari, jorm: when creating forms in MediaWiki I was sorta under the impression that HTMLForms did a lot of the styling for me -- and it's not :p Is there a different Form class I should be building forms with? or is it really that it's totally up to me to decide how it looks and add all the jQuery/CSS stuff?
[18:52:02] <mwalker>	 I should amend that to say; it's not automatically styling it like the way the form style guide informs me that I should be building forms
[18:52:33] <parent5446>	 mwalker: Generally a lot of styling should be done for you (with the mw-input and whatnot classes).
[18:53:38] <mwalker>	 that's what I was assuming -- but all it's giving me is very basic styling -- not at all what the style guide suggests
[19:00:53] <mlitn>	 fabriceflorin: AFTv4 has been disabled completely, AFTv5 is now only opt-in through the use of whitelist categories
[19:03:50] <fabriceflorin>	 Thanks, Mlitn! I will test your new AFT4 and AFT5 revisions right away on Enwiki. chrismcmahon, could you give us a hand as well?
[19:03:50] <chrismcmahon>	 sure
[19:03:52] <Reedy>	 You can't test AFT4 ;)
[19:03:54] * chrismcmahon  investigates negative space

[19:04:59] <mlitn>	 fabriceflorin: I'll refer to https://bugzilla.wikimedia.org/show_bug.cgi?id=37616#c18 for some example articles that fell within the lottery bounds (e.g. http://en.wikipedia.org/wiki/1974_in_art *was* in 10% lottery range, but is now no longer displaying)
[19:05:44] <fabriceflorin>	 Reedy: I was just able to check some articles that used to have AFT4 and they no longer do, so that's a good sign :)
[19:05:49] <mlitn>	 https://en.wikipedia.org/wiki/Golden-crowned_Sparrow is an example where it's still enabled because it's whitelisted
[19:05:57] <chrismcmahon>	 mlitn: I just found that
[19:06:13] <mlitn>	 and clicking around on a number of other pages should quickly reveal AFT4 to be gone
[19:06:33] <chrismcmahon>	 mlitn: so the whitelist is known to work?
[19:07:02] <fabriceflorin>	 We can also check articles in this Article_Feedback_5 category, which editors are now encouraged to use on an 'opt-in' basis if they want to show the AFT5 feedback form on articles they are working on: http://en.wikipedia.org/wiki/Category:Article_Feedback_5
[19:07:26] <mlitn>	 it has not been disabled, and the few pages that I've checked still work
[19:09:12] <chrismcmahon>	 http://en.wikipedia.org/wiki/Special:ArticleFeedbackv5 and links off it look OK to me
[19:09:59] <fabriceflorin>	 Yes, I can confirm that as well. Also, pages with the 'Article_Feedback_5_Additional_Articles' category are working as intended.
[19:11:14] <fabriceflorin>	 Cool. Is there anything else that we should check at this time? Would any back-end process be impacted by this?
[19:13:22] <fabriceflorin>	 I can confirm that this 'Harlem shake' page which used to have AFT4 no longer has a feedback page link in either its talk page nor in the Toolbox on the left sidebar of the article page, as intended: http://en.wikipedia.org/wiki/Harlem_shake_(dance)
[19:14:03] <mlitn>	 fabriceflorin: looks good to me :)
[19:14:22] <fabriceflorin>	 Sorry, that  'Harlem shake' page used to have AFT5 via lottery, not AFT4.
[19:15:28] <fabriceflorin>	 Cool. For anyone who's interested, next week we plan to deploy the new features which are now being tested here: http://www.mediawiki.org/wiki/Article_feedback/Version_5/Testing
[19:15:54] <fabriceflorin>	 Next steps for Article Feedback are outlined in this updated 2013 release plan:
[19:15:54] <fabriceflorin>	 http://www.mediawiki.org/wiki/Article_feedback/Version_5/Release_Plan_2013
[19:18:01] <fabriceflorin>	 mlitn: Anything else you think we can test at this time? Or should we call this done? If we're done, would you like to close Bug 45538?
[19:18:38] <mlitn>	 just closed it - looks like we're done!
[19:19:10] <fabriceflorin>	 mlitn: Nicely done. Thank you very much for the smoothest
[19:19:40] <fabriceflorin>	 mlitn: Thank you very much for the smoothest AFT deployment yet !  :)
[19:24:45] <kaldari>	 mwalker: You might want to use the Agora extension
[19:24:59] <mwalker>	 is that deployed on the cluster?
[19:26:57] <kaldari>	 parent5446: Thanks!
[19:28:28] <fabriceflorin>	 Reedy and Kaldari: To follow up on last week's IRC chat, we are not planning to remove AFT4 on other wikis besides Enwiki at this time, because that's a community decision. We are contacting these Wikis and should have answers in coming weeks. Thanks for your understanding.
[19:29:31] <Wikinaut>	 parent5446: hi, can we chat ?
[19:37:25] <AaronSchulz>	 anomie: maybe an mw.title.getContentModel would be useful too
[19:38:35] <anomie>	 AaronSchulz- The title object already has a ".contentModel" property
[19:38:51] <AaronSchulz>	 yep, I see it
[19:38:54] <AaronSchulz>	 nice
[19:57:22] <parent5446>	 Wikinaut: Send you a PM
[20:15:54] <yurik>	 hi, any hope to get all "xxwiki"-like links into interwiki table? https://bugzilla.wikimedia.org/show_bug.cgi?id=45116
[20:17:01] * Reedy  shrugs

[20:27:21] <yurik>	 is there a dump of all interwiki tables anywhere?
[20:31:49] <Reedy>	 yurik: We don't use the interwiki table
[20:32:17] <Reedy>	 So I truncated them all
[20:32:23] <chrismcmahon>	 andre__: how much longer are you around?  I'd like to send you a draft of something about Search bugs, but it can wait if it's too late for you.
[20:32:49] <Reedy>	 We use a cdb with the interwiki map from meta o it, and do magic
[20:32:52] <yurik>	 Reedy: so we are now finally on to the sites table?
[20:33:14] <Reedy>	 I think only wikidata uses it
[20:33:38] <yurik>	 so how does titles object resolve itself again?
[20:34:00] <Reedy>	 if ( function_exists( 'dba_open' ) && file_exists( "$IP/cache/interwiki.cdb" ) ) {
[20:34:00] <Reedy>	 	$wgInterwikiCache = "$IP/cache/interwiki.cdb";
[20:34:00] <Reedy>	 }
[20:34:22] <yurik>	 no , i mean - it where is the master list
[20:34:42] <Reedy>	 [20:32:49] <Reedy> We use a cdb with the interwiki map from meta o it, and do magic
[20:34:47] <Reedy>	 https://meta.wikimedia.org/wiki/Interwiki_map
[20:36:56] <yurik>	 do you think we should add all possible "..wiki, ..quotes, ..voyage" to match with the http://en.wikipedia.org/w/api.php?action=sitematrix  -- dbname ?
[20:37:04] <yurik>	 to that list?
[20:37:59] <yurik>	 this way all links can uniformly refer to the site+language without ambiguity, regardless of where the link resides
[20:40:42] <yurik>	 oh, and Reedy, that list is only for site links, not for langlinks. Where do you store those?
[20:41:32] <Reedy>	 err
[20:43:53] <yurik>	 you clearly don't create that cdb file from scraing that page... i hope...
[21:03:08] <apergos>	 no, there's a script that does it
[21:09:17] <Reedy>	 hence "do magic"
[21:11:53] <yurik>	 apergos: the script copies interwikies. Where do we store language list?
[21:12:24] * yurik  will look for a shaman tambourine before visiting SF again

[21:13:14] <apergos>	 the language list comes from looking at the list of known language codes to MW
[21:13:24] <apergos>	 and I forget where that is, I read all that code a little while ago
[21:13:37] <apergos>	 language class somewnere?
[21:13:48] <Reedy>	 we have a langlist file
[21:13:51] <yurik>	 ok, so should we introduce the global namespaces?
[21:13:55] <Reedy>	 Which does.. stuff..
[21:14:29] <Reedy>	 And then there's languages\Names.php
[21:15:06] <Reedy>	 /apache/common/langlist
[21:15:22] <Reedy>	 Which is used by other things, including some dns config I blieve
[21:17:32] <apergos>	 I don't remember the langlist file being inplicated in this bit
[21:17:39] <apergos>	 languageNames I think it was
[21:18:07] <Reedy>	 for IW we do.
[21:18:09] <Reedy>	 		$this->langlist = array_map( "trim", file( $this->getOption( 'langlist', "/home/wikipedia/common/langlist" ) ) );
[21:20:39] <apergos>	 bah
[21:20:43] <Reedy>	 lol
[21:20:50] <apergos>	 shows what I remember cause I read all that code >_<
[21:20:53] <apergos>	 oh wellz
[21:21:02] <Reedy>	 heh
[21:21:20] <Wikinaut>	 parent5446: can you pls. code review (easy) https://gerrit.wikimedia.org/r/#/c/52279/
[21:21:39] <apergos>	 http://www.mediawiki.org/wiki/Interwiki_cache/Setup_for_your_own_wiki   hmm no code notes, only notes on the cdb file, oo bad
[21:21:39] <apergos>	 *too
[21:23:04] <apergos>	 (and no I am not going to document the code path, forget it)
[21:23:28] <Reedy>	 It's why I called it magic further up
[21:33:36] <parent5446>	 Wikinaut: Sorry, got distracted. Will do now.
[21:33:57] <Wikinaut>	 no, there's a poroblem with the patch
[21:34:04] <Wikinaut>	 			»	»	$preferences['openid_trust'] =	274
[21:34:05] <Wikinaut>	 			»	»	»	array(	275
[21:34:07] <Wikinaut>	 			»	»	»	»	'type' => 'hidden',	276
[21:34:08] <Wikinaut>	 			»	»	»	»	'default' => $user->getOption( 'openid_trust' ),	277
[21:34:10] <Wikinaut>	 			»	»	»	);
[21:34:14] <marktraceur>	 !pastebin
[21:34:14] <wm-bot>	 Please do not paste more than 2-3 lines of text into the channel as it disrupts the flow of conversation. When sharing multiple lines of code, please use a pastebin such as <http://dpaste.com/> or <http://p.defau.lt/> and post a link to your paste in the channel.
[21:34:16] <Wikinaut>	 when saving, it changes
[21:34:39] <Wikinaut>	 the value
[21:34:49] <Wikinaut>	  $user->getOption( 'openid_trust' ),
[21:35:18] <Wikinaut>	 is "type" => "api" working differently
[21:35:46] <Wikinaut>	 i.e. preserving the binary separator between the elements in 'openid_trust' ?
[21:35:58] <parent5446>	 Left a minor comment.
[21:36:16] <parent5446>	 Also, stick with hidden because type=api was only added recently.
[21:36:57] <Wikinaut>	 parent5446: the value is changed ( a little bit )
[21:37:02] <Wikinaut>	 can this be avoided ?
[21:37:21] <Wikinaut>	 or not, becuase the content is html coded and then re-read ?
[21:37:45] <parent5446>	 Hmm, I'm not sure.
[21:37:48] <Wikinaut>	 then I must think about introducing a different method, or separator
[21:38:33] <Wikinaut>	 see here https://gerrit.wikimedia.org/r/gitweb?p=mediawiki/extensions/OpenID.git;a=blob;f=SpecialOpenIDServer.body.php;h=a7bf87c61e500a67bd24bcf12db8e04306cffe9d;hb=HEAD#l503
[21:38:39] <Wikinaut>	 how it is _generated_
[21:38:46] <gerrit-wm>	 New patchset: Hashar; "jobs for mw/ext/JsonData" [integration/jenkins-job-builder-config] (master) - https://gerrit.wikimedia.org/r/52282
[21:39:01] <gerrit-wm>	 Change merged: Hashar; [integration/jenkins-job-builder-config] (master) - https://gerrit.wikimedia.org/r/52282
[21:39:04] <Wikinaut>	 hex 1E and 1F values are used
[21:39:11] <Wikinaut>	 (not my idea...)
[21:39:29] <parent5446>	 ...I mean, you can just use serialize(), or maybe encode it to JSON.
[21:39:49] <Wikinaut>	 yes, this is what I meant above
[21:39:57] <Wikinaut>	 must introduce a breaking change
[21:40:01] <Wikinaut>	 (what is possible)
[21:40:31] <Wikinaut>	 do we have JSON encode/decode in the core, i think i have seen it
[21:40:43] <Wikinaut>	 (i use this often outside mw)
[21:40:44] <parent5446>	 Oh yeah, that would break stuff. Almost forgot. And yeah, it's the FormatJson class.
[21:41:10] <Wikinaut>	 breaking is ok for this option
[21:41:47] <Wikinaut>	 but the patch isn't that easy as expected :-(
[21:45:02] <gerrit-wm>	 Change merged: Hashar; [integration/zuul-config] (master) - https://gerrit.wikimedia.org/r/52206
[21:47:19] <gerrit-wm>	 New patchset: Hashar; "triggers for mw/ext/JsonData" [integration/zuul-config] (master) - https://gerrit.wikimedia.org/r/52285
[21:47:53] <gerrit-wm>	 Change merged: Hashar; [integration/zuul-config] (master) - https://gerrit.wikimedia.org/r/52285
[23:00:56] <andre__>	 chrismcmahon, you can always send stuff by email, that makes it timezone-independent :)
[23:01:24] <chrismcmahon>	 andre__: yeah, the difference is how much effort I put into the draft, you'll benefit from further research :)
[23:01:45] <andre__>	 heh
[23:01:54] <andre__>	 I might go to bed very soon, so up to you
[23:02:13] <chrismcmahon>	 see you tomorrow
[23:02:19] <mwalker>	 RoanKattouw: as you're the nearest contributor to HTMLForms -- AAAHHHHHH -- why did y'all design it so that 'sections' were always wrapped in 'fieldsets'!? WITHOUT IDS!!!!!
[23:02:39] * mwalker  feels better now

[23:03:07] <Susan>	 HTMLForm.php was written by Andrew G., as I recall.
[23:03:30] <mwalker>	 there's an impressive number of contributors actually
[23:03:44] <mwalker>	 but mr garret does seem to be the oldest
[23:05:18] <RoanKattouw>	 Oh the fieldset thing is my /only/ contribution actually
[23:05:18] <RoanKattouw>	 I forgot about it as quickly as I could
[23:05:18] <RoanKattouw>	 Also why do you need IDs on the fieldsets?
[23:06:04] <mwalker>	 because I want to move form elements around
[23:06:09] <Susan>	 Two Ts.
[23:07:25] <mwalker>	 and it's really hard to style something that has no id or class
[23:24:06] <mwalker>	 RoanKattouw: could you be really awesome and review this 1 line patch for me please! :D https://gerrit.wikimedia.org/r/#/c/52302/1
[23:26:00] <AaronSchulz>	 anomie: https://gerrit.wikimedia.org/r/#/c/52301/1
[23:26:49] * AaronSchulz  skins the cat

[23:27:51] <RoanKattouw>	 mwalker: Oooh I see, it doesn't apply the ID down in the recursion
[23:27:55] <RoanKattouw>	 Well that's just stupid
[23:27:57] <RoanKattouw>	 +2ing
[23:28:00] <mwalker>	 correct! hang on!
[23:28:07] <mwalker>	 I put the damn id in the wrong place
[23:28:17] <mwalker>	 ok -- revision pushed (and this time I actually tested it)
[23:28:28] <RoanKattouw>	 PS2 is the right one then?
[23:28:38] <mwalker>	 yeppers
[23:28:45] <RoanKattouw>	 (I see it moves a dash around relative to PS1)
[23:28:47] <RoanKattouw>	 OK
[23:29:07] <RoanKattouw>	 Should be merged in 2027 when Jenkins finishes CIing it
[23:29:21] <RoanKattouw>	 You
[23:29:25] <mwalker>	 awesomely possumly! thanks
[23:29:35] <RoanKattouw>	 You'll be able to download the new code using your 7G device on the SF-LA high speed train
[23:30:02] <mwalker>	 it didn't take it too long to do the +1
[23:30:12] <mwalker>	 what extra shtuff does it have to do for +2?
[23:30:18] <mwalker>	 compile X?
[23:31:48] <RoanKattouw>	 Yes, but it's cross-compiled to Brainfuck too
[23:32:06] <RoanKattouw>	 No, it runs the unit tests, which aren't usually run on submission unless you are a "trusted user"
[23:32:25] <mwalker>	 ahh -- good choice -- don't trust the fundraising team :)
[23:32:30] <mwalker>	 we do crazy things
[23:32:37] <AaronSchulz>	 TimStarling: https://gerrit.wikimedia.org/r/#/c/52301/
[23:33:00] <RoanKattouw>	 (The concern is that anyone can get an account and running the unit tests is kind of an arbitrary code execution vulnerability as it is)
[23:33:36] <mwalker>	 yep; I recall the discussion
[23:35:02] <mwalker>	 looks like I should poke anomie again about getting the ssh keys installed on the jenkins server so that I can run jenkins jobs in labs! (because that opens the possibility of using labs for extra executors to take care of this annoying queue thing we've got going on)
[23:35:47] <anomie>	 mwalker- Huh? I know nothing about that.
[23:37:28] <mwalker>	 anomie: uh; this might be a name confusion thing on my part -- are you, in real life, antoine musso?
[23:37:45] <anomie>	 mwalker- No. Antoine is hashar
[23:37:53] <mwalker>	 aaahhhh
[23:37:57] <mwalker>	 sorry!
[23:38:00] <anomie>	 no problem
[23:39:18] <hashar>	 yeah I am antoine :-
[23:40:10] <mwalker>	 :) I was attempting to give you a friendly poke to install the fundraising-ci ssh key onto the jenkins server
[23:40:34] <mwalker>	 and thinking we could use the same concept for queue overflow