[00:02:49] Can either of you use a different version? [00:02:55] werdna: Does disabling IE7 mode change anything? [00:03:33] let me see [00:04:55] RoanKattouw: still broken [00:05:04] Grr [00:05:09] TrevorParscal: Does using IE8 change anything for you? [00:05:29] checking [00:08:50] IE 8.0.6001.18783 on Windows Vista Ultimate with Service Pack 1 works [00:10:30] in IE7 mode and IE8 mode [00:10:55] OK screw that, I'm gonna commit this other fix and get some sleep [00:11:02] ha ha [00:11:12] your fix seems to work for me [15:46:05] RoanKattouw_away: please ping me when you are back [15:47:35] nkomura: can you tell parul that if there were some things she took out of the studies about the Wikipedia documentation, to write them down in blogpost or something ? [15:47:56] we are currently merging 4 pages that deal with explaining search for instance [15:48:21] and we added a "contact us" link at the bottom of the page because that is where apparently many are looking for it. [15:49:36] thedj: parul should be online in two hours [15:49:48] yeah, but then i'm not :D [15:50:59] so as far as the second study is concerned, we did not probe help documentation [15:51:42] our main focus was evaluating acai and babaco (current beta) [15:52:16] yeah, i figured something like that from the blogpost [15:52:36] the editing tasks we asked them to do were carried out with the new navigation and the toolbar [15:52:55] i loved the external internal link conflict :D [15:53:00] you may want to contact marlita, PM for bookshelf [15:53:26] good point [15:53:41] yeah, we are addressing the link dialogue right now [16:25:43] hey RoanKattouw [16:25:50] Morning nkomura [16:25:50] there are two of you [16:25:56] Hmm [16:26:00] Oh right [16:26:11] i've been talking to your shadow ;) [16:26:22] *RoanKattouw reads logs [16:27:07] i was asking your shadow to ping me, that's all [16:27:21] Oh OK well here I am [16:28:07] how did the code review go? [16:28:21] i still see lots of "new" for acaifix [16:30:02] Yeah apparently Tim didn't get to those or something [16:30:36] Oh there's only 5 "new" [16:30:59] Well most of Tim's comments were minor, except of course where they concerned ClickTracking [16:31:50] There were no actual issues in what he reviewed, or he would've raised them [16:31:59] we have to revisit click tracking [16:32:12] is this the right path? [16:32:14] http://www.mediawiki.org/w/index.php?title=Special:Code/MediaWiki/path&path=%2Fbranches%2Fusability%2Facaifix [16:32:43] There is this one thing where we made toolbar sections expand on mousedown instead of click, which improves perceived performance (looks faster) but causes a weird issue where Firefox interprets a middle-click as a left-click as well. I'll have to talk to Trevor about that [16:33:01] http://www.mediawiki.org/wiki/Special:Code/MediaWiki/tag/acaifix [16:33:06] What you're looking at are just merges [16:33:19] The merges themselves aren't reviewed, but all they do is merge code that has been reviewed [16:34:46] I'm gonna grab some dinner now, after that I'm gonna make the first preps for today's deployment [16:34:46] I'll have to step out for a bit at 9:30, but I should make it back before 10 [16:35:38] one of the revs which are not reviewed is critical fix for IE right? [16:36:42] werdna? [16:36:55] yeees? [16:37:23] nkomura: Yes [16:37:46] i also see adam's RTL tab fix is deferred [16:37:47] r59185 and r59191 are IE fixes [16:37:51] hi werdna! [16:37:55] No idea why that is [16:38:19] do you have time for code review in next couple of hours? [16:38:21] Maybe because it just changes JS or because it just changes an extension not running on WMF; the latter would be a misunderstanding on Brion's part [16:39:26] nkomura: It could also be that Brion took my advice and just reviewed the current state of CollapsibleTabs rather than each individual rev [16:39:44] The latter typically involves watching people fall and get up a few times [16:40:18] nkomura: possibly [16:40:47] I am a bit behind on my normal work, but I can probably take a look [16:41:03] It's like 5 revs [16:41:05] thanks werdna [16:41:10] i see three [16:41:12] The revs marked as new and tagged as acaifix [16:41:23] There's three pages of acaifix revs [16:41:28] There's two more near the end [16:42:44] 59191, 59185, 58969, 57541 and 57425 [16:45:14] link me up [16:46:12] !r 57425 [16:46:12] --elephant-- http://www.mediawiki.org/wiki/Special:Code/MediaWiki/57425 [16:46:15] !r 57541 [16:46:15] --elephant-- http://www.mediawiki.org/wiki/Special:Code/MediaWiki/57541 [16:46:19] !r 58969 [16:46:19] --elephant-- http://www.mediawiki.org/wiki/Special:Code/MediaWiki/58969 [16:46:22] !r 59185 [16:46:22] --elephant-- http://www.mediawiki.org/wiki/Special:Code/MediaWiki/59185 [16:46:25] !r 59191 [16:46:25] --elephant-- http://www.mediawiki.org/wiki/Special:Code/MediaWiki/59191 [16:48:30] eek, big commits [16:48:32] will look later [16:48:38] adam_miller: status of some of your revs are "deferred" such as 58383 [16:48:56] do you recall why has that been deferred? [16:49:51] no, what does that status mean exactly? [16:50:10] Ask brion, he's the one that marked them deferred AFAIK [16:50:40] brion? [16:51:06] generally 'not super-relevant as not used in production or not key/core; can be looked at later fairly safely' [16:51:13] any particular reason you ask? [16:51:22] http://www.mediawiki.org/wiki/Special:Code/MediaWiki/58383 [16:51:30] we are ready to deploy acaifix [16:51:49] and some revs by adam's have "deferred" status [16:52:12] brion: It changes an extension that wasn't used on WMF /at the time/ , but we're adding that extension today [16:52:27] as they address certain bugs, i hope we can proceed deploying "deferred" revs [16:52:34] right, so if you're adding it new then someone should review its current state :) [16:53:03] i don't recall any low-level issues (database access, core xss safety) that i happened to notice, it looks like just some js-level rearrangement [16:53:28] so you guys need to check it against its requirements, make sure it does what it's supposed to and works on the browsers you're testing [16:53:46] If you like i can take a quick peek at the current code state [16:54:00] yay! :-) [16:55:56] brion: Yeah we had extensive cross-browser feature testing done [16:56:05] whee [16:56:12] It's really just adding JS AFAIK, so it shouldn't be that scary [16:56:18] right :D [16:56:27] Basically, adam_miller's deferred revs in the following tag need to be reviewed http://www.mediawiki.org/wiki/Special:Code/MediaWiki/tag/acaifix [16:56:39] sorry we didn't catch this sooner [16:56:54] ... except if they only concern CollapsibleTabs (which I think is the case), brion can skip them and just review CollapsibleTabs's current state [16:57:26] mostly tab related revs, yes [16:58:10] http://www.mediawiki.org/wiki/Special:Code/MediaWiki/5791 [16:58:59] nkomura: Typo? [16:59:30] http://www.mediawiki.org/wiki/Special:Code/MediaWiki/57918 [16:59:38] sorry the last digit got cut off [16:59:44] hrm [17:00:00] did you guys restructure UsabilityInitiative? cause i don't see collapsibletoolbar in there [17:00:17] brion: Yes. You can review the copy in /branches/usability/acaifix branch insted [17:00:40] ... [17:00:54] ok i don't understand your branch structure :D [17:01:30] /branches/usability/acaifix is a copy of /extensions/UsabilityInitiative in the state we're deploying it this morning [17:02:32] is there a reason the stuff you're working on doesn't seem to exist in dev trunk? [17:03:34] Yes, we restructured the crap out of it [17:03:49] It's now in /extensions/UsabilityInitiative/Vector/Modules/CollapsibleTabs [17:03:53] Or at least part of it is [17:04:02] The part that didn't become obsolete due to the restructuring [17:04:57] isn't this the path for collapsible tabs? [17:05:10] /trunk/extensions/UsabilityInitiative/CollapsibleTabs [17:05:42] i see adam's deferred revs there for acaifix [17:06:22] well there ain't a lot to look at in http://svn.wikimedia.org/viewvc/mediawiki/branches/usability/acaifix/CollapsibleTabs/ [17:06:38] the stub loads a bit of .js which does some jquery magic to move some bits around [17:06:45] what do you need me to comment on? [17:07:36] for example 58383 [17:07:46] can we proceed with deployment? [17:08:25] all revs tagged as acaifix have gone through cross browser testing [17:08:38] so we do not see any issues from usability perspective [17:09:04] but if we release without it, we will be worried the state of the user experience [17:11:13] brion - the meat of it is in http://svn.wikimedia.org/viewvc/mediawiki/branches/usability/acaifix/js/plugins/jquery.collapsibleTabs.js [17:12:31] that js stuff is gonna be all on your end; i don't have the background to review it in full [17:14:02] Yeah it's really just JS, not much to look at for you guys there [17:15:33] brion: how do you recommend us to proceed? [17:16:19] so the potential issue here is that deferred revs may not be compatible with the rest of the code base? [17:16:40] nkomura: proceed as you wish. you guys are the only people who can tell if that code's ok or not [17:16:57] so either decide it's ok or decide it's not [17:17:07] if it's not ok, then fix it and repeat? [17:17:53] so as long as those are working for us, it is safe to assume they will work for us? [17:18:43] if those revs are only affecting us, i feel confident to proceed [17:19:14] Yeah, it'll be OK [17:19:54] i'm a little concerned from deployment procedure point of view though [17:20:13] in theory, "deferred" revs are not OK to deploy right? [17:20:37] nkomura: They're revs that are not reviewed for some reason [17:21:32] so proceed at your own risk ? [17:21:42] Yes [17:21:48] k [17:21:58] But there's a hardly a risk here, JavaScript can hardly make the site explode [17:21:59] i feel better then ;) [17:22:47] the toolbar was once turned off because of the number of javascript files, remember? [17:22:50] anyways [17:22:55] i'm glad we can proceed [17:23:20] we still needs your code reviewed for IE though, RoanKattouw [17:23:32] *werdna runs [17:23:34] nkomura: Yeah the number of them, that's true. But fundraising did the same thing yesterday (number of images in the thermometer) and the site held this time [17:23:50] nkomura: I had werdna and adam_miller test all of that yesterday, and it worked [17:24:13] There were mysterious bugs that happened for werdna and adam_miller but not for Trevor or Nimish, which was kinda weird [17:24:28] did you fix the thing where clicking *anywhere* made the page jump and not work [17:24:31] brion: thank you so much for your advice! [17:24:37] RoanKattouw - want me to run through those again today? [17:24:51] Nah, no worries, the code didn't change [17:25:03] werdna: I think so [17:25:05] :D [17:25:06] RoanKattouw: i'm confident that the revs fix the problem [17:25:19] but the code need to be in the state to be deployed [17:25:57] so let's be nice to werdna, so that he can review the last five revs for us ;) [17:26:13] It's deployment-ready. There may be some issues left on IE, but that's not very different from the current state [17:26:54] ops folks will not like it if we deploy unreviewed code [17:27:05] whether we feel confident or not is a different story [17:27:39] we need to respect the standard practice [17:28:10] If it's just JS it'll be OK [17:29:24] OK I gotta step out for a while, I should be back in time [17:31:55] i see the five revs are mixture of js and php [17:32:12] and also see tim's comment [17:40:20] TrevorParscal and nimish_g, I'm telecommuting today as you must have noticed already [17:43:49] TrevorParscal: there are five revs which have not passed code review yet [17:44:18] RoanKattouw_away and i talked to brion about adam_miller's deferred revs [17:44:26] for those we can proceed [17:45:22] but i don't feel confident proceeding deployment without reviews of five revs [17:46:22] http://www.mediawiki.org/wiki/Special:Code/MediaWiki/tag/acaifix [17:49:32] nkomura: The only PHP involved is trivial, like changing 54 to 55 [17:50:02] if it is trivial, let's ask werdna to review them [17:52:25] is he on right now? [17:52:29] werdna: ping [17:52:29] TrevorParscal: You sent me a contentless ping. This is a contentless pong. Please provide a bit of information about what you want and I will respond when I am around. [17:52:38] ha ha ha [17:52:56] werdna: contentful ping [17:53:16] TrevorParscal: werdna said he will try to get to it [17:53:35] but he is not on it right now [17:58:00] fixing a stupid annoying heisenbug :) [17:58:58] werdna: Is this the same IE7 debugger thing? [17:59:06] no, it's a LiquidThreads bug [18:05:16] OK it's 10 o' clock, you guys ready to roll? [18:05:40] no we are not [18:06:14] the deployment is on hold until the code review is complete [18:06:41] *RoanKattouw pings werdna [18:15:23] bah [18:15:26] let me look [18:20:18] they all look fine [18:20:23] assuming they've been tested [18:20:30] and you know it all works correctly [18:20:38] If you'd like me to do some testing in my VM, I can do that [18:21:27] thanks a lot werdna! [18:21:39] have you changed the status to "ok"? [18:24:49] He hasn't [18:24:58] So can we roll now? [18:26:52] so you are planning to deploy all revs except deferred for nimish? [18:27:09] werdna: i appreciate if you can change the status of five revs you reviewed [18:27:13] I asked Tim, and he said it was OK to deploy those to as long as ClickTracking remains disabled [18:28:04] fair enough [18:28:23] one more question before we roll [18:28:29] how do you treat "fixme" [18:28:30] ? [18:28:36] are any more fixed? [18:28:45] nkomura1: it means a problem has been discovered [18:28:54] it will be marked as resolved if a fix has been made and verified [18:28:58] so we are not deploying those [18:29:33] or do we deploy and fix them? [18:29:47] I'm working on the clicktracking stuff now... [18:29:54] well, they should be fixed before they are deployed :) [18:30:16] there are two "fixme" for catrope [18:30:31] Yeah we deploy ClickTracking but it stays disabled, we only reenable it when it's fixed [18:30:40] nkomura1: Yeah that was Tim wondering if the fixes really worked, I think [18:31:00] Oh wait there's the middle-button issue [18:31:03] TrevorParscal: You around? [18:31:11] yes [18:31:13] TrevorParscal: R [18:31:15] sorry - distracted [18:31:17] Read the comments at http://www.mediawiki.org/wiki/Special:Code/MediaWiki/58974 [18:31:20] reading backlog [18:31:50] werdna: when you have a moment, can you change the status of five revs you reviewed to "ok"? [18:31:59] if you want to go back to click, I'm OK with that [18:31:59] i want to keep records for ops folks [18:32:26] the reasoning for using mousedown was very minor [18:32:33] TrevorParscal: Yeah that's probably best [18:32:35] it was just improving precieved performance [18:32:37] no worries [18:32:46] OK I'll just change it back to click real quick [18:34:30] hi, hannes-tank [18:34:59] hi [18:38:44] adam_miller_lunc: Woot, NTOC collapsing looks smooth now [18:39:05] where's it deployed at? [18:39:06] yeah, but there's a host of other problems [18:39:10] sandbox#? [18:41:12] TrevorParscal: i don't think it's made it's way onto a sandbox yet [18:41:37] it's in trunk [18:41:57] I'll svn up s3 [18:42:06] i'm updating my localhost [18:42:38] I'm gonna do s5 too [18:42:42] it looks ass-y right now [18:43:39] adam_miller: the trick I use in the toolbar is to make the "shrinking" or "growing" box position:absolute while it's animating [18:43:55] that way you don't get strange reflow bugs / performance issues [18:44:19] Yeah I was looking into doing that this morning [18:44:20] also, in the toolbar, I animate the container separately from the content [18:44:33] that will help make it much smoother looking for sure [18:46:28] also, we should make the area that you can use to drag-and-resize on the right of the scrollbar wider [18:46:33] like, 8 px or something [18:46:39] maybe 5px... [18:49:23] It's 7px right now [18:50:06] the cursor doesn't seem to be properly indicating where it's at right now though [18:56:07] it may be overlapped by something else [18:56:10] mostly [18:56:49] hmm [18:56:59] i see it in firebug - looks right [18:57:02] hi parutron [18:57:04] OK guys, let's roll [18:57:09] but yeah, the cursor seems to be acting up [18:57:15] I merged everything, now I'm gonna put it up at test.wikipedia.org [18:57:20] hi nkomura1 [18:57:23] *TrevorParscal takes down the terrorists [18:57:26] wait, what? [18:57:31] glad you had an easy commute this am ;) [18:57:44] parutron: thedj was looking for you [18:57:55] thedj? [18:57:56] yeah easy commute, but working from 7am [18:58:21] time spent working seems better than time spent commuting! [18:58:55] hello thedj - nice to meet you - hear you are looking for me. [18:59:08] RoanKattouw: so what happened to r58974? [18:59:52] thedj maybe offline by now, he asked me to ask you if we found any insights from documenting how to use search from the usability study [18:59:58] but we can talk offline about it [19:00:04] as we need to move on to the release [19:00:10] nkomura1: I resolved that but forgot to comment, commented now [19:00:56] and r58873? [19:01:47] Tim said it didn't work for him but that sounds like a magic iframe bug. It did work on the deployment candidate [19:02:20] the sensitive highlighting stuff [19:02:25] and nimish's fixme? [19:02:32] That's ClickTracking, right? [19:02:47] right the timestamp stuff [19:03:01] Yeah that's ClickTracking, Tim said I could deploy that even though it wasn't clean, we can clean it up later [19:03:11] k [19:03:13] I'm working on refactoring some of the code as per Tim's request [19:03:19] Good [19:03:42] so, we're still deploying it? [19:03:51] OK guys I updated http://test.wikipedia.org/ to the deployment code, please hit it [19:04:04] *TrevorParscal hits it [19:04:10] nimish_g: we are deploying your stuff, but they will be disabled [19:05:00] wow - rounded corners [19:05:04] still see opt-out survey in "welcome to beta" page [19:05:26] that was fixed on prototypes [19:05:54] RoanKattouw: this is not the candidate build [19:06:05] it doesn't even have watch/unwatch [19:06:08] rounded corners? [19:06:16] ok, I can work on fixing the front-end but aside from one escaping issue (where I added quotes and escaped manually instead of using DB escapes) I don't see any reason why the clicktracking backend can't still be deployed and gather data [19:06:20] watch is in the drop-down [19:06:21] Hmm [19:06:35] test.wikipedia.org looks so 3 months ago [19:06:49] nimish_g: The thing is that it's not clear the code is secure. You should rewrite the code so it's obvious it's secure [19:07:00] *nkomura1 shift-refreshes three times [19:07:35] *TrevorParscal wonders why the latest software is not on test... :) [19:07:50] Ah turns out you have to configure watch/unwatch [19:07:50] *RoanKattouw configures [19:10:28] opt-in flow was not up-to-date either [19:11:06] RoanKattouw: if I understood Tim's message right he had quite a few issues with how I handled DB code in the *front end* stuff, none of which is actually active right now. I'm fine with rewriting it to make the security more clear, but as far as the back end part, ie the part collecting data, the only issue he seemed to have was using addQuotes instead of manually quoting and escaping, which I just fixed. I think turning off the data gather [19:11:28] Aah [19:11:38] (your message got cut off after "data gather") [19:12:02] it said: I think turning off the data gathering while waiting for exclusively front end fixes is unnecessary [19:12:57] Ah [19:12:59] *RoanKattouw is stupid [19:14:49] I think commenting out this line: [19:14:50] so nimish_g, we can continue track the data although the front-end needs some modification before turned on? [19:15:03] $wgSpecialPages['ClickTracking'] = 'SpecialClickTracking'; will be fine to turn off the front end [19:15:08] nkomura1, I sure hope so [19:15:28] or if you want to be thorough, there's like 4 more lines you can comment out [19:15:36] nimish_g: OK, I'll look at this after the general deployment [19:18:38] watch star is back on test.wikipedia.org :-) [19:18:57] Yup [19:19:03] OK our code is now running on test for real [19:19:05] Let's hit it for a bit [19:19:07] the toolbar is old though [19:19:25] it is not the new toolbar we intended to release [19:19:56] i thought the deployment branch was staged on prototypes [19:19:58] Enable dialogs in your prefs? [19:20:17] It's a "hidden preference" [19:20:20] bold, italic icons are old [19:20:32] indentation is missing [19:20:56] Huh [19:20:57] signature icon is missing [19:20:58] Weird [19:21:27] it's old stuff [19:22:01] Yeah I see that, trying to figure out why [19:22:18] RoanKattouw: what is the difference of the code between the one on prototype and the test? [19:22:26] I'm trying to figure that out now [19:26:37] brion: What magic do I do to update styles&JS on test? [19:27:22] brion: I svn up'ed some JS changes on zwinger, but they're not being picked up, while the PHP changes are [19:27:41] RoanKattouw: svn up should do it, as long as you've got sane versino markers [19:27:46] I do [19:27:55] I even changed the version marker by hand and that didn't fix it [19:27:58] ok [19:28:03] - sorry [19:28:09] was helping with a video shoot [19:28:24] brion: I guess I'll just sync-common srv124 [19:29:09] TrevorParscal: thanks for bringing in the video camera [19:29:14] Yeah I see /apache/common not matching /home/wikipedia/common on srv124 [19:29:29] WCJ organizers will be thrilled to receive jimmy's video message [19:40:28] RoanKattouw: please ping me when you want to resume [19:40:39] i will go to take care of a few other things [19:40:47] nkomura1: Will do. Sorting out some tech problems with werdna and brion in #wikimedia-tech [19:41:20] glad to hear you are getting good support in sorting things out [19:42:48] hey adam, collapsing the ntoc on sandbox5 is pure fun! [19:46:53] Woot, test.wikipedia.org works now [19:47:02] nkomura1: Test is working now [19:47:09] Let's go hit it (third attempt) [19:47:14] cool [19:47:20] the third is the charm [19:48:18] looking good now [19:49:38] Huh collapsible tabs isn't working for me [19:49:42] Maybe I need to turn it on :) [19:50:48] not working for me either [19:51:43] opt in worked better this time [19:54:06] mr. TrevorParscal: the screenshots are not quite up-to-date yet... [19:54:28] I was waiting to update them until we actually deploy [19:54:37] cause it's just an image on commons [19:54:43] shall I do it now? [19:54:43] k [19:54:55] will u update the screenshot for the toolbar too? [19:56:16] yeah [19:56:19] working on that right now [19:56:26] merci [19:57:09] OK collapsibletabs is working now [19:57:19] I forgot to put it in the Makefile so it wasn't in the combined JS files [19:58:20] OK it's all working nicely for me [19:58:31] If you guys could confirm that I can go and roll this out to the rest of the cluster [19:58:45] arr.. [19:58:52] (may need Shift+Refresh) [19:58:54] let me refresh and try again [19:59:19] hannes-_-: Are you still getting that IE bug you e-mailed me about, even on http://test.wikipedia.org/ after a Shift+Refresh? [19:59:44] still getting the old toolbar [20:00:02] I'm not [20:00:22] *RoanKattouw increments style version [20:00:54] collapsing tabs are working for me [20:01:00] Try again? Maybe on another page? [20:01:02] but the toolbar is not current [20:01:19] i created a page, but didn't work [20:01:22] will try again [20:02:13] http://commons.wikimedia.org/wiki/File:UsabilityNavigation.png [20:02:17] http://commons.wikimedia.org/wiki/File:UsabilityToolbar.png [20:02:22] both are now updates [20:02:24] *updated [20:03:11] i ran into a page with "Draft" and "Edit draft" [20:03:33] i guess it is with flagged rev [20:03:34] Yeah Drafts is enabled on test, but nowhere else [20:03:42] where do I turn on the tabs [20:04:00] werdna: You don't. Vector should have collapsible tabs automatically. If not, hard refresh [20:04:19] how do I know if they're on or not? [20:04:21] i'm not getting the new toolbar [20:04:23] I can see the new watch button [20:04:40] now i got it [20:04:43] Good! [20:04:55] OK let's roll this thing out then [20:05:20] RoanKattouw: how do I know if the tabs are on or not? [20:05:23]