[10:42:15] @trusted [10:42:15] --elephant-- [c-67-170-206-100.hsd1.ca.comcast.net, wikimedia/Catrope, adsl-70-137-172-57.dsl.snfc21.sbcglobal.net] [14:59:41] good morning RoanKattouw [14:59:45] Morning [15:00:06] are you back home? [15:05:51] RoanKattouw: how late are you available today? [15:06:14] Yeah I'm at home, today's lecture was canceled so I never left [15:06:34] I'll be out at the usual Thursday morning time, except an hour later due to your DST [15:06:34] lucky you [15:06:49] So that's 10:45-12:45ish [15:07:07] (PDT) [15:07:59] I'll probably stick around till 4 or 5 [15:08:01] most of us will be in meetings from 9am to noon PDT today [15:08:10] Right so I'm not missing much then [15:08:33] Why all those meetings? :) [15:08:55] planning meetings for the coming fiscal year [15:08:56] Is it me or has the meeting density increased since the last time I visited? [15:09:01] Ah [15:09:06] The 5-year strategic plan thing? [15:09:10] it is dense right now [15:09:15] yeah [15:10:09] so RoanKattouw, we do need to enable dialogs to existing beta users who do not have turned on in their preferences [15:10:35] will you start the preparation [15:10:38] Sure [15:10:41] i have a blog post ready [15:10:45] I'll start by counting how many users we're talking about [15:11:52] from your stats, i estimate about 450K [15:14:25] RoanKattouw: shall we tentatively say we do config change around 12:30pm PDT / 8:30 CET? [15:15:37] I'll be out at that time [15:15:40] 10:45-12:45 [15:20:07] ah, right [15:21:12] how about 3:30pm PDT / 11:30pm CET? [15:21:36] Should work [15:22:52] i'm seeing a bug report re S&R [15:28:04] Hm I assigned them to Adam but I guess he's not working today? [15:28:51] oh it's thursday [15:29:14] yeah he has his kid today [15:44:42] OK I'm now building per-wiki lists of users with the toolbar but without the dialogs [15:46:09] nkomura: Have you seen http://meta.wikimedia.org/wiki/Wikimedia_Conference_2010/Accommodation ? [15:46:23] can we weed out users who turned off dialogs after feb 4 release? [15:46:49] no [15:46:59] nice photos [15:47:04] It's a hotel-boat :) [15:47:14] oh [15:47:16] cool [15:47:18] I will, thanks for reminding me [15:49:36] RoanKattouw, ohhh, neat! (the hostel-boat) [15:50:40] I don't know if the chapter people "get" to sleep there, I think they're being put into a proper hotel as opposed to a hostel [15:51:16] They got a pretty good rate too, E 19 per person per night plus E 4 for breakfast [16:54:26] Hi, is there already a target date when the new skin shall be deployed? [16:59:06] You mean as default? [16:59:26] I don't recall offhand, nkomura probably knows [17:03:25] hi hejko [17:03:34] hi [17:03:36] i'm in the meeting [17:03:39] i've got your email [17:03:54] ok :) [17:03:59] we are shooting for the week of April 5th [17:04:15] i had a question about pdf server readiness [17:05:08] pdf2 is up and running [17:05:13] before the exposure to anons [17:05:16] k [17:05:18] hejko: There seems to be something odd going on with load balancing between pdf1 and pdf2: pdf2 is pulling much more load than pdf1 according to Ganglia [17:05:26] let's chat more about that later [17:06:50] hejko: Although, strangely, pdf2 is supposed to be a beefier box and seems to actually be cranking out /less/ in terms of MB/s [17:08:05] we think that this issue will be solved if you switch to nfs version 4 [17:08:05] the job queue is shared via NFS. NFS4 should be available as a configuration option. [17:08:35] Ah yeah that's what I was thinking [17:08:43] That one of the boxes is pulling NFS load on its own [17:10:13] I'll forward these questions to Ralf. But I think if NFS is configured to version 4 everything should run smoothly. [17:11:44] OK [18:21:54] does anyone know if paru is in the office? [18:25:31] ... [18:27:13] hannes_^^, she's right next to me, we're in a strategy meeting [18:27:41] she's going to her desk now [18:55:49] Rah rah ramama, roma ramama gah gah ooh la la. [19:03:46] RoanKattouw_away: hi [19:58:03] Good afternoon [20:00:01] hi RoanKattouw [20:06:59] RoanKattouw: TrevorParscal and i are still determining whether table insertion bug is a show-stopper or not [20:07:30] I need to poke at it [20:22:19] i just committed a publish dialog fix that needs to get pushed to the cluster at some point [20:22:22] todayish [20:22:29] looking at some other bugs [20:22:49] OK [20:22:52] RoanKattouw: https://bugzilla.wikimedia.org/show_bug.cgi?id=22877 [20:22:55] any thoughts? [20:23:04] Why don't you poke those bugs some more and I'll just update the cluster to trunk state [20:23:11] *TrevorParscal fires up VMWare instance [20:23:28] ") [20:23:30] :) [20:25:38] clearly that bug has to do with the textarea loosing focus and thus loosing cursor position... [20:25:46] looking into save/restore of that [20:30:25] Yeah [20:30:29] We've had it before [20:40:58] RoanKattouw: i need help with this [20:41:04] got a min? [20:41:15] Sure [20:41:38] I'm doing math homework so I can jump in and help you any time [20:42:31] What specifically do you need help with? [20:44:37] well, I'm trying to save the selection when a button is clicked, and restore it just before things like encapsulateSelection get called [20:45:14] Right [20:45:14] the issue is that the heading tool and dialogs cause the textarea to loose focus since they require 2+ clicks [20:45:37] we don't have a getSelection function that returns the equiv of the input of setSelection [20:45:37] Yeah [20:45:45] getCaretPosition() [20:45:46] that's my feeling, that I need to write that [20:45:47] ? [20:46:15] getCaretPosition would not keep tes[ting testing 1]23 (where [ and ] are the selection) [20:46:24] and depending on the tool, that may or may not be important [20:46:52] it would just resore the cursor to tes | ting testing 123 [20:46:59] getCaretPosition(true) should return a tuple [20:47:10] ah! [20:49:01] *TrevorParscal reuses context.savedSelection which has been dormant since iframe switch.. i think [20:49:13] Probably [20:50:09] hmm [20:51:15] $().textSelection( 'getCaretPosition', { 'startAndEnd': true } ) [20:52:20] context.$textarea.textSelection( ... ) [20:53:20] Does anyone know a dark version of vector skin? [20:53:45] I'd say just tweak some CSS and there you go [20:59:00] hmm [21:03:03] Where is the "add media wizard" gadget source on svn? [21:04:06] mdale: ---^^ [21:04:32] js2-work/phase3/js/mwEmbed/modules/AddMedia [21:05:40] JeroenDeDauw1: http://svn.wikimedia.org/svnroot/mediawiki/branches/js2-work/phase3/js/mwEmbed/modules/AddMedia/ [21:11:05] Found it @branches/js2-work/phase3/js/mwEmbed/modules/AddMedia/ [21:11:28] JeroenDeDauw2 sorry must have been sent while you disconnected? [21:12:06] No, sorry about that, the wireless here failed for a moment [21:16:34] TrevorParscal: Are you having any luck with that bug? [21:16:51] a bit [21:17:13] one of the issues was that we need to use e.preventDefault together with return false for IE to behave as FF does [21:17:17] always [21:18:44] Interesting [21:18:48] You really need both? [21:19:12] yes [21:19:50] the other thing is that getCaretPosition is returning completly insane values [21:19:51] ^^ [21:19:58] Oh? [21:20:51] yeah, I have it setup so on blur/focus it console.log()s the getCaretPosition [21:21:09] on focus, it's sane [21:21:19] on blur, it's seemingly random [21:21:53] RoanKattouw: what is your availability for a meeting early next week? [21:22:09] i am trying to schedule a conference call with mark, trevor, tomasz and you [21:22:53] well, it's repeatedly but inconsistently 3191, 5099, 5097 - the article is only 1300 characters long [21:24:06] TrevorParscal: Blur happens after the element loses focus, so the textarea will no longer be focused, that's why you get bogus [21:24:30] right [21:24:36] nkomura: Assuming it's gonna be during office hours PT, anything works [21:25:20] nkomura: What's the call gonna be about? [21:25:37] mostly discussing the go default plan and system impacts [21:25:48] OK [21:32:43] nimish_g and parutron: how are the template collapsing and expansion working for test users? [21:32:57] as it is the day for dry run for tomorrow [21:33:39] it'd be good to have calcey test it [21:33:40] will test it now [21:35:19] nkomura and parutron I've been playing with pasting into an expanded template and there doesn't seem to be a quick, easy fix for it right now, so we know that's broken, as is pasting a template capsule [21:42:19] nimish_g: don't you want calcey test templates today? [21:42:25] i mean tonight [21:42:46] yes, but just know that we know that pasting is broken so they don't need to test that [21:43:12] why don't you write test instruction [21:43:14] ? [21:44:46] ok [21:50:04] nimish_g: you may want to set up a separate account for calcey [21:50:17] on enwiki? [21:50:21] so that we do not need to share the usability study accounts [21:50:38] prototype would be ideal [22:20:14] TrevorParscal: Any luck? [22:21:23] RoanKattouw: I'm trying to create a calcey test account on prototype [22:21:37] OK [22:22:01] I went into the DB and set highlight, preview, publish and template editor to 1 for that user ID [22:22:11] and it didn't show up [22:22:16] is there more that I need to do? [22:22:33] Well [22:22:38] Those features aren't enabled on prototype [22:23:10] You might as well create those accounts for them on enwiki [22:24:29] Or tell them to use the deployment prototypes, which are closer to production code than the regular prototypes [22:26:05] RoanKattouw: TrevorParscal went into a meeting [22:26:06] he won't be out for another hour [22:26:06] OK [22:26:06] nkomura: is it ok for us to test on the deployment prototypes? [22:26:06] i don't think we can enable dialogs today [22:26:07] nimish_g: sure [22:26:07] ok [22:26:08] actually, RoanKattouw those aren't enabled on the deployment prototypes either...I'll enable them on deployment-en [22:26:24] Oh [22:26:26] Hm [22:26:41] unless there's some step I'm missing [22:26:52] You're probably right [22:27:09] All the deployment prototypes run off the same LocalSettings.php file so you only have to enable it once [22:30:47] nimish_g and RoanKattouw, if it is complicated to test account on prototype [22:30:52] Add $wgWikiEditorModules['templateEditor']['user'] = true; after the line that includes PrototypeSettings.php [22:31:08] we should consider either using one of the accounts created in production or create another account dedicated for testing [22:31:13] It's not that complicated [22:31:17] and 'highlight' [22:31:29] Yes [22:32:32] k [22:38:54] RoanKattouw: we are likely to have a conference call with mark b monday around 11am PDT [22:39:16] OK [22:39:40] When it's final, could you invite me using Google Calednar? [22:41:27] RoanKattouw: will do [22:41:33] Also, assuming Mark also has an extension in the phone system, we wouldn't need the call-in bridge [22:48:24] RoanKattouw: we don't have any other settings like "reallyEnable" or anything right? [22:48:28] No [22:48:34] hm [22:48:39] That's just for NTOC [22:49:46] ooh, right there's no iframe...how do I re-enable that? [22:49:47] For templateEditor and friends you just gotta turn on the prefs [22:50:01] You don't need to? [22:50:25] You should just force those prefs in the DB and it should work [22:51:37] well I tried that at first, then just enabled them in localsettings, and went into the user acct, turned on highlight, toc and templates and there's still no iframe [22:51:54] Wait which wiki is this? [22:51:58] http://prototype.wikimedia.org/d-en/index.php?title=Grammy_Award&action=edit [22:53:08] OK [22:53:14] Lemme peek at the LocalSettings file [22:54:31] Hm [22:55:01] you getting the same result? [22:55:10] I checked and the code's up to date [22:55:21] and exists [22:59:04] Ah [22:59:15] It's a very subtle brokenness in the code that sets up the iframe when required [22:59:41] .wikiEditr('addModule', {'foo': options}) triggers it, but .wikiEditor('addModule', 'foo') doesn't [22:59:46] Lemme fix that [23:04:16] k [23:04:39] are you going to push that fix to deployment? [23:05:04] (the branch)...if not, can you send it to me and I can just merge it locally on prototype? [23:12:35] I'm committing, hold on [23:13:24] Committed to trunk. You can svn up on prototype, even wikis/d/extensions/UsabilityInitiative runs trunk [23:13:26] No need for merging [23:19:45] « So far, every single thing I've seen that the foundation has paid for has been completely unusable. Thank goodness I perma-blocked the sitenotice after the fundraising banner fiasco. Glad to see the money's being well spent on these web-designers from 1994. » [23:19:48] yay [23:20:04] :) [23:21:49] Well I haven't been impressed with our outsourced design work at all [23:22:23] Admittedly I'm not a designer [23:26:15] guillom: What's the fundraising banner "fiasco" BTW? [23:26:39] RoanKattouw, probably the WIKIPEDIA FOREVER thing [23:26:47] Right [23:27:14] The cheesy slogans used in the first stage [23:28:51] RoanKattouw: sorry, netsplit [23:29:16] Committed to trunk. You can svn up on prototype, even wikis/d/extensions/UsabilityInitiative runs trunk [23:29:18] No need for merging [23:29:32] excellent, thanks [23:33:51] RoanKattouw: you should be in bed... yes? [23:35:17] Maybe [23:35:34] Friday's my day off though [23:35:54] RoanKattouw: I know you need to sleep soon...but you sure your fix works? it's still not working right in prototype [23:36:50] Lemme take a look [23:41:41] Oh wait crap [23:41:47] UI is not from trunk [23:41:49] WTF [23:42:00] *RoanKattouw cleans up mess [23:43:35] nimish_g: I was wrong, wikis/d/extensions/UsabilityInitiative does not run from trunk. Merging stuff in now [23:44:00] nimish_g: Try now [23:44:01] k, I figured it was *all* from the deployment branch (which would make sense) [23:44:13] Yeah but svn status seems to indicate otherwise [23:44:20] So at some point I probably did switch the UI dir to trunk [23:44:30] Don't wanna do that now because wmf-deployment contains live hacks [23:44:48] ok yay it works now [23:45:00] Good [23:45:07] now, get to bed =P