[11:48:44] ugh, wtf [11:48:50] *werdna slaps jQuery around a bi [16:00:30] werdna: How did jQuery wrong you this time? [16:12:58] .data() was being inconsistent [16:13:18] Oh yes, saw that on identi.ca [17:12:07] nkomura: online now [17:12:09] morning :) [17:15:21] hey Ryan_Lane [17:15:44] RobH and i were trying to chat with Mark, Systemadmin in netherlands [17:15:56] nkomura: ok. is he still online? [17:16:21] he seemed to be away from his desk [17:20:08] hey [17:20:52] mark, RobH, and Ryan_Lane, can we chat about the server set-up for QA automation [17:21:20] we might have lost RobH [17:21:35] he has an appointment with sun technician [17:21:40] ok [17:22:01] mark: so i have email in progress to send you about QA automation [17:22:24] so this has been planned since october, to systematically test new UI initially for usability project [17:22:42] but eventually opening up to broader projects of WMF [17:23:10] what kind of broader projects? [17:23:10] initial users are the usability dev team and tech consultants who work remotely on QA [17:23:45] for example, flaggedrevs [17:24:01] I'd rather like to keep this separate from the development virtual servers we were talking about in the tech meeting [17:24:01] new features of MediaWiki which will be deployed to production [17:24:24] especially if this server is going to be managed externally [17:24:36] mark: separate from a network perspective? [17:24:44] separate in every way [17:24:53] like, let's keep this server restricted to the usability team [17:25:08] I am here, sorry was making food. [17:25:51] mark: can we include external folks as long as they are within the usability team? [17:25:57] absolutely :) [17:26:39] ooooh qa automation makes me happy [17:26:47] so i do not know what it means to separate these QA sever from dev cluster network wise [17:26:50] i'm confused about the level of separation. I was under the assumption that this would be more of a development system [17:27:02] but one dependency i can think of is [17:27:12] I was under the impression that you wanted to run browser tests and such on it? [17:27:13] that all test cases will be submitted for code review [17:27:24] perhaps that is unclear :) [17:27:36] i get why it would need to be on a separate network, but confused why only usability team can use it [17:28:03] alright... what do you want to do on this box exactly? [17:28:35] this link has the topology [17:28:36] http://usability.wikimedia.org/wiki/Resources#Interaction_testing_automation [17:28:40] mark: initially selenium rc testing. which would be in a number of VMs [17:29:06] what OSes will be running on it? [17:29:15] mark: those VMs would launch browsers to run tests. the tests would be submitted by code review, or possibly remote testers [17:29:32] mark: linux and windows. the xserve would be running OS X (tiger and snow tiger server) [17:29:49] mark: the host OS is slated to be vmware esx [17:29:56] on the dell anyway [17:30:04] we're talking about multiple servers now? [17:30:13] mark: two servers [17:30:34] one dell server for esxi and windows/linux testing and one apple xserve for safari and osx testing. [17:30:56] for general WMF use we want to set up a cluster for spawning arbitrary vservers for development work [17:31:13] i don't see why this couldn't be used for that [17:31:43] Ok, but why does the future of these servers involve them being used for other dev testing, it really isnt a discussion for right now... [17:31:47] I'd rather keep that separate if this server is not going to be managed by the ops team [17:31:56] right now the discussion is where the two usability dev servers will live [17:32:19] they should not be included as part of the wmf ops team domain, other than to get them up in the rack and LOM ready. [17:32:32] indeed [17:33:09] the discussion should be why these servers should be in the DC, or if they belong in the office. I know we had this between Ryan, Naoko and I, but now we need to rehash that so Mark knows why we want it in the DC [17:33:14] i want to make sure the set-up is expandable to the core dev team as well though [17:33:22] nkomura: but we don't want that [17:33:24] not that i want to complex the topic [17:33:30] it's much easier and less complex if we don't [17:33:31] why not? [17:33:39] basically, I'd rather give you one or two boxes that you can manage and do whatever with [17:33:46] so that it doesn't make our own needs more complex [17:33:47] nkomura: If you have folks with direct access to that server, they have to be vetted for that kind of access on the cluster, if you are including it in that domain. [17:34:01] makes it easier from a networking and management perspective [17:34:06] (oversimplified answer, but works) [17:34:24] plus, I'd rather not use VMWare at all for our own operations :P [17:34:29] AFAIK, i'm likely to be the only one with shell access to the VMs [17:34:53] let's say a new volunteer developer come a long, it will be matter of giving that person an write access? [17:34:58] Yes, there is a long running debate on what software to use in cluster for potential VM software, so best keep this outside that bailiwick. [17:35:03] mark, RobH: you should be looking at RHEV/RHEV-M btw [17:35:13] RHEV? [17:35:30] nkomura: well, I will have access to the server for you, to do setup of the initial server. Then Ryan will have full access [17:35:33] mark, RobH: that's a discussion for later, but red hat plans to have an upstream version soon. [17:35:37] anyone else you want to have access, you can give access to [17:35:45] Ryan_Lane: we don't use red hat [17:35:46] as long as it isnt in the ops domain that the rest of the tech team works in [17:35:52] indeed [17:36:14] if its under ops team management, and not in its own sandbox, then adding more folks would be more difficult. [17:36:21] as they would need vetting by tech team [17:36:36] if its in the sandbox, on its own, and you get a volunteer dev you want to give access to it [17:36:37] let's put it this way [17:36:42] ryan can merely grant it when you ask. [17:37:18] RobH: are we worried about VM access to the cluster, or host access? [17:37:18] (what Rob says) [17:38:05] I'm assuming host access will be extremely restricted. VMs can be put into different VLANs [17:38:09] both. [17:38:16] well, actually [17:38:19] Ryan_Lane: Canonical/Ubuntu has a similar system... since we use Ubuntu exclusively, we're more likely to use that :) [17:38:21] if its in its own sandbox, neither [17:38:29] but lets move on to why it needs to be in the dc, or doesnt need to be in the dc [17:38:47] mark: let's discuss that in a bit. :) [17:38:55] the other matter is mostly settled in that if you want the server on line in a timely fashion, it has to be in a sandbox networkwise and merely reside in wmf racks [17:39:58] that's fine. [17:40:07] since the immediate need is for usability team [17:40:42] so my understanding of what this server is doing is a large number of automated testing on multiple software platforms and versions that will spin up and may require many hours to complete. During which having a power system dedicated to providing uninterrupted power is ideal. Also it will be pushing and pulling large amounts of data at semi-regular intervals as tests are run and data is collected. That data may be a bit [17:40:42] for the office internet connection, which is only a 4/4 line. [17:41:36] Ryan_Lane: if I did not sum that up correctly, please do not hesitate to say I am wrong =] [17:41:57] it can be in our dc just fine of course [17:42:01] RobH: No. I think that is correct [17:42:11] I think the UPS argument is bollocks, but I appreciate the office connection argument ;) [17:42:18] RobH: We may also look at moving prototype to the box as well [17:42:32] in which case even more traffic. [17:42:35] yeah [17:42:38] alright [17:42:39] but makes sense [17:42:46] the stats on prototype spike high occationally [17:42:52] so let's set up these servers in the data center, in a special sandboxed network environment [17:42:58] Ryan_Lane: yea, i get the emails when they do [17:43:24] excellent [17:43:31] nkomura: that means I can order your dell today. [17:43:35] Ryan_Lane: good morning...did we ever do anything with idea torrent? [17:43:45] and since these servers need to be managed by the usability team (Ryan), I'd like to keep their use restricted for usability [17:43:53] nimish_g: no. I'd like to discuss that today though. [17:43:58] for other development/test work within WMF we'll set up a different cluster [17:44:10] which will be managed by us, of course [17:44:18] ok. we'll wait till other members of the team mozy in [17:44:20] that keeps things simple, and responsibilities clear [17:44:47] yay! [17:44:51] mark: at some point in time my contract is going to be ending. I'm not sure who will be managing it then [17:45:08] actually, i think my contract is up in about 20 days [17:45:14] hehe [17:45:30] Ok, I need to head down to the DC to meet the engineer [17:45:43] thanks RobH [17:45:44] Ryan_Lane: we would love to renew it of course [17:46:00] nkomura: i'm free if you chose to do so [17:46:01] mark: i'm a little worried the cost for setting up a different qa cluster though [17:46:11] but that is a separate discussion [17:46:21] these servers are not cheap... [17:46:36] relatively cheap compared to the scale of other things we're doing ;) [17:46:46] we wanted to set up a QA cluster on a bit larger scale anyway [17:46:59] and integrating the two would only complicate it a lot [17:47:08] understood [17:47:17] and if our QA automation attempts work [17:47:31] we can apply the practice in the larger scale [17:47:37] hehe well [17:47:45] thank you for your support on this mark [17:47:49] mark: i'm not sure how it would complicate things much. network separation can be done with vlans at the host level [17:47:52] the linux bits I guess :) [17:48:09] mark: i don't necessarily need host access [17:48:33] mark: separation of duties can be done via virtual center so that I could have access to some vms, and not others [17:50:03] mark: rob seemed to want vmware. i was recommending xen on opensolaris or ubuntu [17:50:44] I want to use ubuntu enterprise cloud [17:50:56] mark: that is missing tons of functionality [17:51:15] I'd rather not use vmware at all, as we like to use free software whenever reasonable [17:51:20] but that won't work with windows of course [17:51:29] mark: RHEV/RHEV-M is going to have an opensourced upstream version that is comparable to vmware [17:51:58] feel free to use that for usability then :) [17:52:14] mark: as a bonus, it uses SELinux to ensure VMs can't interfere with each other [17:52:47] *Ryan_Lane shrugs [17:53:29] you guys can use whatever you want. i'm just recommending things based off of the research i've been doing, and the clusters i've been managing [17:56:03] nkomura: nimish_g was looking to use ideatorrent, but we still have to work out the authentication issues. brion mentioned setting up an openid provider. [17:56:36] brion: is it possible to test central auth and openid on prototype, or does it have to be tested on the cluster? [17:56:36] i'd like to understand what that means [17:56:48] since we use SUL with other WMF projects [17:57:00] spiffy [17:57:05] i don't think introducing OpenID is that simple [17:57:07] nkomura: right. essentially, the provider would work in conjunction with SUL. [17:57:11] Ryan_Lane: you can certainly test openid on prototype [17:57:12] Ryan_Lane: Prototypes don't have SUL set up, and I don't wanna start doing that either [17:57:14] although i am supportive of the idea though [17:57:25] could set up a local centralauth just for kicks but i don't envision any difficulties [17:57:28] (as provider anyway) [17:57:30] right [17:57:36] nkomura: it would allow users to use their user page as their openid [17:57:39] Yeah a local CA between prototypes maybe [17:57:43] mw openid provider is working as of my last tweak to it a few weeks ago at least [17:57:47] But setting up CA is hell, and an undocumented one at that [17:57:55] heh [17:57:57] oh wayt [17:58:00] heh [17:58:00] wait [17:58:10] our usability wiki is in prod [17:58:18] RoanKattouw: I was thinking of using one of the spare sandboxes for it, but I trust your judgement on this one :) [17:58:29] CA across all prototypes would be fine [17:58:37] CA between prototypes and WMF wikis not so much' [17:59:13] RoanKattouw: I think that would likely be sufficient for testing [17:59:14] what is CA? [17:59:22] nkomura: cental auth [17:59:33] central too [17:59:34] The technology behind SUL [17:59:40] ah [17:59:49] i see [18:00:13] let's test it on prototype first before opening up ideatorrent to wider audience though [18:00:23] RoanKattouw: really we just need to ensure that the openid extension doesn't conflict with CA [18:00:29] Yeajh [18:00:33] i want ideatorrent to be part of usability wiki [18:00:35] nkomura: sounds good [18:00:44] But CA across prototypes is not a bad idea [18:00:54] nkomura: A *part* of usability wiki? How? [18:01:28] i am not seeing ideatorrent as a separate instance from user perspective [18:01:44] it should be seamless experience using usability wiki or idea torrent [18:02:00] nkomura: so we should be integrating the software somehow? [18:02:16] ideatorrent should be a part of usability wiki in a long term [18:02:35] eventually yes [18:02:45] authentication can be the same, but merging the software is likely to be difficult [18:02:56] I think drupal supports portlets though... [18:03:07] Ryan_Lane: do you mean merging ideatorrent with mediawiki? [18:03:35] i think merging authentication is a good start [18:03:50] nkomura: yeah, that's why we are looking at openid [18:04:07] openid is likely to replace SUL at some point in time correct? [18:04:21] Not sure [18:04:24] *RoanKattouw looks at brion [18:04:27] heh [18:04:42] i remember reading his slides at some point in time, and seeing that. I could be remembering wrong though [18:04:56] I'm not familar enough with OpenID to see how it could replace SUL; I think as of now we just want to be an OpenID provider, but not accept OpenID logins [18:05:08] *Ryan_Lane nods at RoanKattouw [18:05:36] RoanKattouw: it does essentially the same thing as SUL (or can) [18:05:43] RoanKattouw: do you use facebook? [18:05:50] Noep [18:05:52] ah [18:06:07] Ryan_Lane: are you on FB? [18:06:17] well, they accept external OpenIDs. I use google's OpenID provider to log into facebook [18:06:22] nkomura: yep. joined recently [18:06:32] ok, i'll find you [18:06:41] or find me, naoko.komura [18:06:50] RoanKattouw: when I go to facebook's login page, it automatically logs me in if I'm logged into google [18:06:56] Ah yes [18:07:49] RoanKattouw: most sites don't do automatic login, but it can be done [18:09:36] It would probably be good to investigate migrating from SUL to openid, but that isn't really in this project's scope :) [18:10:49] nkomura: oh. I was going to add content to the sandboxes today. I was thinking of importing the english dump, and another language's dump. Will this cause problems? Is there another way you'd like me to import content? [18:11:35] i'd of course ensure the main page goes back to sandbox's main page [18:11:39] supporting multiple languages in sadbox is great [18:11:48] but how would that work? [18:12:14] if it is in one database, perhaps it is better to keep one language? [18:12:33] the language pages shouldn't collide [18:12:51] ok [18:13:08] lots of wikis have mixed language contant [18:13:10] content too [18:13:15] that's right [18:13:27] Ryan_Lane: Have you considered two wikis might both have a page called [[Michael Jackson]]? [18:13:57] RoanKattouw: would that cause a serious issue in the history? [18:14:31] Well one of the language's versions would come out on top (the one with the most recent edit) and the history would be very confusing [18:14:37] ah [18:14:39] true. [18:14:52] RoanKattouw: so I should just stick with english content? [18:14:56] You could do it if you searched for dupes beforehand and took measures by renaming stuff [18:16:01] Ryan_Lane: so back to your question about data import to sandbox, please proceed as you planned [18:16:56] in addition to english, german and japanese are good language to imports [18:17:08] Ok. I'll check for dupes, and rename pages when needed. [18:17:09] from language characteristics point of view [18:18:32] RoanKattouw: how's it going? [18:18:38] I'm good [18:18:50] Coming up on my 10th shell request now :) and I've deployed your year-to-date feature [18:18:55] RoanKattouw: for renaming pages in the dumps, I just have to worry about ? [18:19:03] <RoanKattouw> Just now; after checking with Rand of course [18:19:18] <RoanKattouw> Ryan_Lane: I think so. I'm not terribly familiar with the XML dump format [18:19:46] <Ryan_Lane> RoanKattouw: me either :). I'll check for more references to the title. [18:28:02] <RoanKattouw> TrevorParscal: Is the FundraiserStatistics data cached by any chance? [18:29:20] <TrevorParscal> no [18:29:28] <RoanKattouw> Hm [18:29:46] <RoanKattouw> I wonder why syncing two .php files and requesting new stats caused such a load spike then [18:29:51] <RoanKattouw> Oooh, l10ncache, of course [18:29:58] <RoanKattouw> You added a message, so it got rebuilt [18:30:28] <TrevorParscal> we could add in some memcache stuff I'm sure.. [18:30:53] <TrevorParscal> it doesn't need to be rebuilt more than once per minute or so... maybe even every hour would be OK [18:31:04] <RoanKattouw> Don't worry about it [18:31:11] <RoanKattouw> I was wondering where the load spike came from, is all [18:31:43] <RoanKattouw> Hm Tomasz changed the message to "Cumulative total (USD)" [18:34:35] <TrevorParscal> RoanKattouw: are you sure that the lack of PST is to blame for the changes new line bug? [18:35:04] <TrevorParscal> I think it's quite possible that the new line is being artificially injected by the browser [18:35:11] <TrevorParscal> and could be removed [18:41:34] <RoanKattouw> Could be [18:41:41] <RoanKattouw> But that leaves the signature bug [18:51:06] <Ryan_Lane> ok. using a dom parser on the dumps on prototype was a bad call :) [18:53:10] <RoanKattouw> :) [18:53:17] <RoanKattouw> regexes? [18:53:47] <Ryan_Lane> i'm going to scp them to my local system, fix them, then upload them [18:53:54] <Ryan_Lane> worst case I'll use a SAX parser instead [18:54:19] <Ryan_Lane> i ate 500MB of ram, which was essentially all that was left on the system [18:54:22] <RoanKattouw> Ah [18:55:00] <TrevorParscal> who's got what jquery.wikiEditor.* code out right now? [18:56:54] <RoanKattouw> TrevorParscal: I've got massive changes to highlight, one (5 lines in 1 cluster) in templateEditor [18:57:22] <RoanKattouw> Oh and some comment changes to wikiEditor that I should probably commit now [18:58:23] <TrevorParscal> go ahead [18:58:40] <RoanKattouw> Done; didn't recombine or raise style versions because I just changed comments [18:58:47] <TrevorParscal> I'm going to audit some of the iframe initialization stuff, I suspect something strange is happening with the references [18:58:52] <RoanKattouw> (lest I get another "Listen to yourself" thrown in my face :) ) [19:07:54] <TrevorParscal> RoanKattouw: line 208 of jquery.wikiEditor.js - in what case is it possible for $.wikiEditor.modules to not exist? [19:08:13] <TrevorParscal> it's initialized as an empty object on line 16 [19:08:15] <RoanKattouw> Dude that's not my code [19:08:27] <TrevorParscal> who's is it? mine? [19:08:28] <TrevorParscal> hmmm [19:08:29] <RoanKattouw> It's Nimish's, I removed it but he put it back when he reverted half my changes [19:08:44] <RoanKattouw> He committed some code that looked like unrelated debugging code to that file [19:08:44] <TrevorParscal> in jquery.wikiEditor.js ? [19:08:48] <RoanKattouw> Yes [19:09:07] <TrevorParscal> nimish_g: did you get an error or something that led you to add that, or were you just adding extra validation? [19:09:15] <RoanKattouw> http://www.mediawiki.org/wiki/Special:Code/MediaWiki/59912#c4744 [19:09:24] <TrevorParscal> nice blame [19:09:26] <TrevorParscal> :) [19:09:34] <nimish_g> let me check [19:10:10] <RoanKattouw> heh [19:11:03] <RoanKattouw> Surely if that code were mine or yours, I'd l o o k l i k e t h i s, right Nimish? ;) [19:11:22] <RoanKattouw> s/yours/Trevor's/ [19:11:32] <TrevorParscal> seriously! [19:11:38] <nimish_g> c l e a r l y =) [19:12:06] <Ryan_Lane> oh wtf. this xml isn't welformed? [19:12:23] <RoanKattouw> Ryan_Lane: Cut-of UTF-8 maybe? [19:12:26] <RoanKattouw> *off [19:13:03] <TrevorParscal> The opposite would be when my wife was chatting with me from her computer where my daughter got some junk into her space bar and it stopped working... she was like "imtryingtogetsomethingdonetodaybutthekidsaregoingbonkersandmyspacebardoesn'tworkcausecadencewaseatingcookiesonit" [19:13:04] <Ryan_Lane> maybe. it's missing end tags for sure [19:13:15] <Ryan_Lane> regex's it is :( [19:14:49] <TrevorParscal> beware, i might move some things around, but I'm not making any functional changes [19:15:05] <TrevorParscal> this file is a bit disorganized (jquery.wikiEditor.js) [19:15:38] <RoanKattouw> Cool, I'm not touching it [19:38:03] <howief> here's a verbatim from a user about the tabs [19:38:25] <howief> "removing the tops off the tabs is akin to Seinfeld's concept of removing tops off muffins. Nice try but try harder. Thanks." [19:38:55] <RoanKattouw> Is this metaphor intended to capture the fact that the top is the good part of a muffin? [19:39:13] <howief> haha maybe [19:40:08] <mdale> .. appears the "setTimeout" function does not work in "WINE safari" while all other javascript and browser display works fine ... ~gasp~ some productivity lost to linux :( [19:42:31] <RoanKattouw> Wait what setTimeout() doesn't work? [19:42:32] <nkomura> howief: it'd be helpful if you can pull up the comments about users expressing difficulty clicking tabs as they are not sure where tabs end [19:42:38] <RoanKattouw> That's like a fundamental JavaScript function [19:42:52] <RoanKattouw> "WINE Safari"? Screw the 5 people on this globe using that? ;) [19:42:53] <mdale> yea I think Wine Safari is fundamentally broken [19:43:00] <howief> yes, i'm going through the comments now [19:43:21] <mdale> I just thought I would try native instead of the VM... but turned out to be a waste of time [21:33:43] <TrevorParscal> someone needs to write a plugin for eclipse that lets you write commit comments while editing the file, which get used when you actually commit the file.. [21:34:04] <TrevorParscal> a little textarea at the bottom of the screen would be pretty easy to do... [21:34:32] <mdale> synchronize with repository is ~sort-of~ like that [21:35:48] <TrevorParscal> yeah, but it's a bit on the heavy side [22:36:45] <YaronK> Here's the demo page for the latest version of TemplateInfo, 0.3: http://discoursedb.org/wiki/TemplateInfo_test [22:37:53] <YaronK> Main changes in this version: the display switched to CSS, and "unhandleable" XML can now be defined, via the "type=" attribute. [22:44:48] <Ryan_Lane> testing content import on my test wiki, then importing to sandboxes. I'll send an email when done. afk for like 1-2hrs while testing [23:00:12] <nkomura> i am seeing french text in table dialogue on prototype [23:16:58] <RoanKattouw> nkomura: Link? [23:25:20] <RoanKattouw> YaronK: You're using Html::rawElement() in TemplateInfo::tableRowHTML() and friends, so where are you escaping the input then? [23:27:09] <YaronK> The Html::rawElement() is being passed in the output of Html::element(), so it should be fine, I think. [23:27:56] <YaronK> Oh, oops, not in tableMessageRowHTML(), though... [23:28:41] <YaronK> I assume that's the right way to do it, though (calling rawElement() on element()) - no? Otherwise it escapes the HTML tags themselves. [23:28:55] <RoanKattouw> Yes [23:29:07] <RoanKattouw> But any input should be escaped somewhere [23:29:16] <RoanKattouw> But you're right that double-escaping messes stuff up [23:29:37] <YaronK> Right, okay. It looks like I had it right in tableRowHTML(), but not in tableMessageRowHTML(). [23:32:01] <RoanKattouw> Also (I just submitted a CR comment about this), you had "$css_class" in those functions; those quotes are unnecessary [23:32:16] <RoanKattouw> I think they even make PHP run the code slower, although that difference is probably very small [23:33:31] <YaronK> Are you sure it's unnecessary? Don't you need class="className"? [23:33:41] <YaronK> As opposed to class=className? [23:33:56] <RoanKattouw> Not if you're doing array( 'class' => $class_name ) [23:34:05] <RoanKattouw> The Html:: functions are smart enough to add the quotes themselves [23:34:14] <YaronK> Ah, okay. [23:34:18] <RoanKattouw> And in PHP, $class_name and "$class_name" produce the same result [23:34:32] <YaronK> Oh yeah... oops. [23:34:33] <RoanKattouw> If you'd wanted to add those quotes, you'd need "\"$class_name\"" or '"' . $class_name . '"' [23:34:43] <YaronK> Right. [23:35:29] <RoanKattouw> + $data_type = htmlspecialchars($data_type); [23:35:51] <RoanKattouw> While this works, I think it's cleaner to just change the rawElement() call to element() and not escape $data_type yourself [23:36:22] <RoanKattouw> Oh wait [23:36:25] <YaronK> The problem is, there's no HTML tag for just that string. [23:36:37] <RoanKattouw> Never mind, you're appending a <span>, only just notice [23:36:38] <RoanKattouw> d [23:36:54] <YaronK> I could make a <span> tag around that string too, though, if that's preferable in some way. [23:37:01] <RoanKattouw> $content = is_null($value) ? $data_type : "$data_type: " . HTML::element('span', array('class' => 'rowValue'), $value); [23:37:43] <RoanKattouw> You may want to parenthesize this, not everyone (probably more like hardly everyone) knows that . takes precedence over ? : (which I /think/ is what's happening here) [23:37:53] <RoanKattouw> *hardly anyone [23:39:09] <YaronK> Ah - I'll just switch it to an 'if' statement; it's a little unwieldy now anyway. [23:40:07] <RoanKattouw> TrevorParscal: Could you review http://www.mediawiki.org/wiki/Special:Code/MediaWiki/59971 ? It's all CSS and I added some comments, but I'm not really qualified to review large portions of CSS so I didn't mark it as OK; please do so if it passes your review [23:41:24] <TrevorParscal> yes, nearly done turning jquery.wikiEditor.js on it's fragile little head, shaking out all the loose bits [23:42:42] <YaronK> RoanKattouw - switched to 'if' now. [23:42:49] <RoanKattouw> k [23:51:08] <TrevorParscal> RoanKattouw: take a look at http://www.mediawiki.org/wiki/Special:Code/MediaWiki/r59987 [23:51:18] <TrevorParscal> I will look at http://www.mediawiki.org/wiki/Special:Code/MediaWiki/59971 [23:54:30] <RoanKattouw> TrevorParscal: Combined&minified files: pass. Style versions: fail ;) [23:56:04] <TrevorParscal> ha ha [23:56:06] <TrevorParscal> DANG IT! [23:56:11] <TrevorParscal> doing it now [23:57:57] <TrevorParscal> done [23:58:07] *TrevorParscal needs a script to do this for him