[15:09:49] hey adam_miller [15:09:56] hey hannes-_- [15:09:59] I just sent u two mails [15:10:00] got both those emails [15:10:09] thanks for the feedback [15:10:17] are u still the one who is working on that? [15:10:36] Trevor is doing some work on the capsules now [15:10:42] well....I d like to add that it s a awesome work! [15:11:10] looks like all I do is critizing it - but those are really details [15:11:19] ha thats your job [15:11:33] some of these things are easy changes, others not so much [15:11:47] like the baseline alignment....not so easy to do [15:11:57] hmm too bad [15:12:07] but that doesn't mean we're not trying [15:12:29] did u do the page vacuuming stuff as well? [15:12:30] just don't expect to see that fixed in the next two hours [15:12:39] or was it nimish baby? [15:12:43] initially yes, has there been more work done on it? [15:12:52] i don think so [15:12:54] i havent touched it in a few months [15:13:00] but I was on my way to review iot [15:13:03] review it [15:13:34] ok - so maybe it s a good thing if I review it and give you some feedback [15:13:46] yeah that'd help me get started back into that again [15:14:45] also...if you need any help with graphical stuff or anything else I can do: feel free to fire work at me ;) [15:15:26] (didn t work at all last week - so I have to catch up) [15:17:11] hey RoanKattouw [15:17:31] Hey hannes-_- [15:18:01] *RoanKattouw wonders which of elephant and elephant2 are active [15:18:03] @trusted [15:18:03] --elephant-- [c-67-170-206-100.hsd1.ca.comcast.net, wikimedia/Catrope, adsl-70-137-172-57.dsl.snfc21.sbcglobal.net] [15:18:03] [c-67-170-206-100.hsd1.ca.comcast.net, wikimedia/Catrope, adsl-70-137-172-57.dsl.snfc21.sbcglobal.net] [15:18:07] Ha [15:20:30] hannes-_-: what do you mean by new buttons exactly? [15:21:11] we designed our own buttons for "save" "cancle" etc.... [15:21:45] currently we use the jquery buttons in the dialogs and system-buttons for "save" and "cancle" in the editor [15:22:25] I remember you were almost done with that [15:22:36] and you even had rounded corners [15:23:05] remember? [15:24:32] yeah i do, but i thought those did go live [15:24:45] where are you doing your review? [15:24:57] on prototype? [15:25:12] S7 [15:25:16] wrong? [15:25:36] s-7 may be out of date [15:25:41] I'm not sure when it was last updated [15:25:52] and your not seeing the rounded buttons on there? [15:26:23] they look exactly the same to me as the buttons I have locally [15:26:58] hannes-_-: if you're not happy with the buttons on s7, send me some changes then because I thought those were our final version [15:27:47] adam_miller: I see windows systems buttons on s7 [15:27:54] in the dialogs? [15:27:54] you don t? [15:28:03] no on the editor [15:28:11] oh beneath the editor [15:28:25] right [15:28:42] yeah on dialogs I can see our buttons [15:29:15] I ll resend a mockup to you for the buttons beneath and above the editor [15:29:26] we wanted to habe a gray and a blue button [15:29:34] to emphasize the save-button [15:29:44] hi there [15:29:50] we could use that for the dialogs as well [15:29:59] hi guillom [15:31:05] so adam_miller and RoanKattouw is s7 up to date? how can we find out? [15:31:26] i'll update it [15:31:31] to make sure [15:31:37] cool [15:31:38] thx [15:31:58] adam_miller: Please update it on tesla (208.80.152.248) as well in that case [15:33:18] RoanKattouw: do i have access to tesla? i only half payed attention to that change [15:33:49] Yes [15:34:10] Excpet that there's no external DNS alias yet, so you need to use the IP [15:34:54] Aw crap [15:35:01] do i need to update it on prototype.wikimedia.org as well? [15:35:06] I just spent half an hour writing what I thought was a performance update [15:35:07] Yes [15:35:16] *performance improvement [15:35:36] Turns out that the case where it harms performance is much more common (thousands to one) than the case where it helps [15:37:16] *RoanKattouw destroys all his local changes [15:38:12] hannes-_-: UP TO DATE. [15:38:20] adam_miller: brilliant [15:40:00] uhh what happend to the ntoc on s7 [15:40:10] hmm nevermind....that s not what 27 is about [17:19:10] I need to create a pop-up in which users can upload an image, previewx it, and provide a caption. Is there any MW stuff I could use for this purpouse? If not, anyone aware of some awesome jQuery plugin that would be well suites for this? [17:20:48] hi JeroenDeDauw. have you tried add-media-wizard gadget michael dale created? [17:20:58] you can enable it from user preferences [17:22:47] weird, template wrapping is not happening after my latest update [17:25:04] pdhanda: that's a good thing no? [17:27:08] no i mean when you type a template the little widget does not appear any more [17:27:13] i think it broke in r63530 [17:27:38] it works in r63526 [17:27:39] oh [17:32:42] nkomura: No I have not - any docs on that? [17:34:21] both michael and I blogged about it [17:34:32] i can walk you through how to enable it or send a link to the blogs [17:35:27] nkomura: Is this widget available on wikimediafoundation.org? [17:36:05] yes, the gadget is available through any Wikimedia projects [17:38:02] nkomura: Awesome. Links to the relevant blog posts would be great. [17:38:57] one sec, let me look it up [17:39:36] adam_miller: some of these changes that hannes mentioned - I did yesterday [17:39:43] some of them i'm unsure I agree with [17:39:49] i thought you might say that [17:39:58] which one did you change yesterday? [17:40:24] the baseline thing is solved - that was never intended afaik - i just took on this work so it was an artifact of whoever set it up before me [17:40:39] the alignment with the bottom is solved for good afaict [17:40:53] i got failry creative coming up with a way to get that done right [17:41:22] oh? it looked pretty broken when i updated this morning so i just committed another fix for baseline stuff [17:41:24] I cleaned up the icons allot in the past few days [17:41:31] really? [17:42:04] yeah man, things were all disconnected and misaligned [17:42:15] what were you testing it in last night? [17:42:17] might have been commits after me? [17:42:18] lookin [17:42:22] localhost/prototype [17:42:36] browser, i mean [17:43:18] FF3 [17:43:21] my understanding is that FF3 is our target right now, so i was judging it based on that [17:43:24] the only browser we need to worry about right now [17:43:35] k cool [17:43:42] FYI wikiEditor.highlight.js Comment out lines 315 through 317 for new template wrapping to work again. Added a comment to code review [17:43:54] yeah, now there's a problem, the capsule moved when you click it [17:44:28] and it's taller than the line-height so when it is in a body of text, it make the line it's on taller [17:44:45] pdhanda: What's on those lines? [17:44:58] *RoanKattouw has to run out and has local changes, just wants to know real quick [17:45:04] TrevorParscal: send me a screenshot [17:45:15] it looks perfect for me in FF3.6 [17:45:42] JeroenDeDauw: i've got distracted by email and other IM [17:45:48] RoanKattouw: [17:45:49] if(cBreak){ [17:45:49] return; [17:45:49] } [17:45:53] looking up the link for you now [17:46:09] after var ca1 = startNode, ca2 = endNode; [17:46:10] if ( ca1 && ca2 && ca1.parentNode ) { [17:48:04] Oh Nimish's new code [17:48:06] I didn't even look at that yet [17:48:15] I have some nice performance improvements in there now [17:48:24] (there = highlight.js and wikiEditor.js) [17:48:30] I'll commit them when I get back [17:50:54] good timing nimish_g [17:51:25] http://www.mediawiki.org/wiki/Special:Code/MediaWiki/63530#c6177 [17:51:39] ^ for nimish_g [17:52:02] k [17:52:08] TrevorParscal: http://i44.tinypic.com/nmn41i.jpg and isnt the moving of the capsule a result of it's display being changed from inline-block to block? [17:52:40] adam_miller: one sec - sorry [17:56:29] adam_miller: basically, your changes make it break in my browser [17:56:40] what version of ff? [17:56:45] do you have local changes? [17:57:03] no, and 3.5.8 [17:57:04] JeroenDeDauw: here are the links to the blogs [17:57:15] when I revert you changes to jquery.wikiEditor.html everything starts working again as it should [17:57:22] http://techblog.wikimedia.org/2009/10/new-media-features-gadget/ [17:57:25] could it be that different between 3.6 and 3.5.8? [17:57:31] http://blog.wikimedia.org/2010/01/26/multimedia-usability-project-underway/ [17:57:51] nkomura: Thanks [17:57:55] *JeroenDeDauw looks at them now [17:58:19] adam_miller: it's possible, I know it's not perfectly aligned in chrome (a couple px off) [17:58:56] what's the deal with taking the gradient out of the expand image? [17:59:02] why is that better / a good idea? [17:59:16] ask hannes_away [17:59:29] hannes_away: ??? [17:59:50] TrevorParscal: why dont you email him and cc me since he's not here [18:00:23] and paru as well [18:00:32] this can be our new closed tabs vs. open tabs [18:00:53] JeroenDeDauw: this blog about subtitles by michael dale may interests you too [18:00:54] http://techblog.wikimedia.org/2009/11/fun-with-subtitles/ [18:08:17] TrevorParscal: I just downloaded FF3.5.8 and it's rendering the capsules the same way as 3.6 was for me. Not seeing any brokenness [18:09:09] nkomura: The gadget is totally awesome. Unfortunately, it's not quite doing what I need. I have a submission for looking like this [18:09:10] http://upload.wikimedia.org/wikipedia/meta/e/e0/Wikimedia-storyboard-mockup02.png [18:09:10] When a user hits an upload button there, he/she should get a similar pop-up, but with simple upload from the user's machine (this is for avatars). [18:09:10] Any way to re-use stuff of what the gadget is doing? [18:09:28] adam_miller: what kind of test page are you using? [18:09:50] Boeing 787 [18:11:13] TrevorParscal: i can send you the wikitext if you want. Is anyone else seeing brokenness after my last commit? [18:12:01] JeroenDeDauw: i think re-use is of AMW (add-media-wizard) is a possibility [18:12:19] michael (mdale) usually hang around in this room [18:12:47] he hasn't appeared yet [18:13:02] adam_miller: pastebin it [18:13:23] nkomura: So you're saying I should ask him about the technical details? [18:13:45] JeroenDeDauw: yes. you can also email michael at mdale at wikimedia dot org [18:14:20] JeroenDeDauw: you must be working on the storyboard project [18:14:30] i heard about it from erik [18:14:33] nkomura: Yeah [18:14:48] TrevorParscal: http://pastebin.com/uq4aUUKs [18:17:53] JeroenDeDauw: the storyboard project sounds awesome [18:18:16] adam_miller: in the Production section [18:18:22] that template that's on it's own line [18:18:33] nkomura: It is. It's my first WMF project though, so I hope I don't mess up to much things :) [18:18:43] when you expand and contract it, in your browser does it move vertically a few pixels? [18:18:48] k car is ready, will be back soon [18:19:29] JeroenDeDauw: no worries, mediawiki developers will help you out [18:19:57] TrevorParscal: totally. i misunderstood you before. is that the only issue you're seeing? [18:20:04] yes [18:20:18] if you revert your change it doesn't do that [18:20:26] however, I'm sure there's other things you change fixes [18:20:30] I'm just looking at the diff now [18:21:06] using vertical-align with pixels is very likely the problem [18:21:20] ooh [18:21:27] no, i'm changing the bottom margin on expansion [18:21:29] margin bottom - that probably doesn't work how you think [18:21:42] nkomura: Who are you anyway? :d [18:21:55] yeah, it should in theory bring content rendered after it up/down [18:21:57] if I disable that and instead change the top margin on the wikitext it fixes it [18:22:13] My name is Naoko Komura, i lead user experience programs at Wikimedia Foundation [18:22:38] instead, because it's inline, it adjusts the box up/down within the line [18:22:41] i report to erik moeller. that's why i heard about the project [18:23:26] adam_miller: cool - just commit that then [18:23:48] Jeroen, she is a wonderful lady :) I respect her a lot [18:24:36] *nkomura blushes [18:25:04] TrevorParscal: done. [18:25:05] adam_miller: maybe use this as a test page as well -> http://pastebin.com/zQxMM2i5 [18:25:10] thats what I've been using [18:25:17] I have your test page on my wiki as well [18:25:48] aweome [18:26:12] it's still making the line 1px too tall - but I'm totally fine with that for now [18:26:21] we need to focus on the bigger fish at this point :) [18:26:41] nkomura http://stats.m.wikipedia.org/ mobile is taking off [18:28:11] 20 million page views yesterday - am I reading that correctly? [18:28:13] amazing! [18:28:54] isn't it tomsz showed it to me last week [18:29:14] i'm really curious to see the origin of the traffic [18:29:16] that is more then the Dutch Wikipedia and less then the Portuguese on the basis of Februaru [18:29:20] origins i mean [18:29:36] I am working hard to grow mobile traffic [18:29:45] but this is beyond my wildest dreams [18:30:02] this is aggregate of currently available languages right? [18:30:07] yes [18:30:42] the number of languages supported by an adequate localisation is growing fast [18:31:11] so i heard, thanks to your and siebrand effort [18:31:24] and translatewiki contributors [18:31:47] ..the big problem is support for fonts in for instance the Indian languages [18:32:15] also the Japanese mobile main page is not configured ... so the room for growth is huge [18:32:57] (Japanese being the second biggest in traffic with a mobile crazy country :) ) [18:34:27] yeah [18:38:12] TrevorParscal and RoanKattouw, can we chat a little about Citron staging, QA and the release? [18:38:47] Sure [18:38:48] ... [18:38:50] On IRC or on the phone [18:38:51] sure [18:39:02] IRC should be fine [18:39:05] Bleh, is Danese in the office? [18:39:12] no [18:39:17] Grr [18:39:26] Is the also sick? [18:39:29] *she [18:39:51] probably just at a meeting of some sort [18:40:09] so i'm still holding us accountable for deploying Citron on March 15th [18:40:29] working the schedule backward, we need to be in the QA state [18:41:07] i do understand that we are still facing major issues with templates and our focus has been making that component work for the usability study [18:41:41] *RoanKattouw looks up which features Citron consists of again [18:41:53] citron also accomplish one more important objective, which is to restore dialogue feature for beta users who opted in for that feature [18:42:23] Yes [18:42:37] basically we took away features at the end of babaco enhancements release and we still have not stored back [18:42:42] So those dialogs are ready for QA, just get Calcey going on them [18:43:01] I need to make one fix so it won't turn on the iframe when dialogs aren't enabled [18:43:05] it is important that existing beta users have experience with dialogues as this feature is slated for "go default" feature [18:43:11] Totally [18:44:06] my suggestion is to start testing dialogue without iframe in the next couple of days [18:44:30] and deploy the changes even if the template work is not ready [18:44:52] Yes [18:44:55] TrevorParscal: how do you feel about that? [18:44:58] That was my thinking as well [18:45:22] Yes, we should be testing the dialogues with textarea [18:45:34] So I suggest Calcey test all three dialogs with the textarea [18:45:51] Where the priorities are 1. link, 2. table, 3. search&replace [18:45:52] all three? or just links and tables? [18:46:03] And if we discover s&r is broken we'll split it out [18:46:10] k [18:46:13] It'd become its own preference too [18:46:25] so S&R does not work well with text area? [18:46:31] Or we just move the dialogs from the dialogs pref to the toolbar pref [18:46:37] I have no idea how well it works, I didn't try [18:46:40] k [18:46:55] It's just that the table dialog is a bike, the link dialog is your car, and S&R is a truck [18:47:06] RoanKattouw: will you stage the build on all prototypes? [18:47:08] And old bridge will definitely hold the first two [18:47:09] Yes [18:47:22] *An [18:47:42] users of S&R do not mind driving druck [18:47:47] they adore it [18:47:48] ;) [18:47:50] I know [18:48:00] But the old bridge they're driving on has to hold [18:48:41] we'll build the new wide and fast bridge soon [18:49:22] although the switch over to tesla is completed, we are serving prototypes from two hosts right now [18:49:41] RoanKattouw: so will you make sure that both hosts are updated? [18:50:16] RoanKattouw: please write test instruction to bhagya and 'cc trevor, mangala and me [18:50:17] Yes [18:50:22] OK [18:50:32] they should go through all babaco test cases for regression [18:50:42] First I need to make some final tweaks, I'll probably have that e-mail out by lunch [18:50:50] great [18:56:12] RoanKattouw: the code that got rid of the extra line breaks was 295 in highlight, correct? [18:56:22] Ask Trevor [18:56:27] He did that [18:56:39] TrevorParscal: same q =) [18:56:44] what about it? [18:56:58] those conditions that check for

[18:57:01] I'm doing all kinds of things at the moment, I'm halfway a massive performance improvement in the highlighter and now I'm poking at dialogs first [18:57:15] yeah [18:57:24] i was thinking it's possible to have

text
text

and that condition still be met [18:57:58] we should check to see that .text() doesn't contain any non-whitespace characters [18:59:24] It won't [18:59:33] $('

foo
bar

').text() --> 'foobar' [18:59:49] Oh NON-whitespace, nm [18:59:52] :) [19:07:07] hey TrevorParscal [19:07:15] hi [19:07:29] I kind of agree with u - the problem is that the expanded area looks like it is above the capsule [19:07:53] so that s a conflict [19:08:09] above? [19:08:18] like, rendering bug? [19:08:29] no it s a gr4aphical thing [19:08:33] let me try to explain [19:08:36] shoot [19:08:55] the capsule has a gradient - the template is flat (pure #f5f5f5f) [19:09:26] but the gradient fades to that color [19:09:33] no [19:09:35] which is the middle color [19:09:38] no? [19:09:41] for the normal capsule it doesnt [19:09:57] for the new down-status it does [19:09:58] in the image I sent in my email [19:09:59] right? [19:10:19] the shaded version, the depressed expand/collapse button has a shaded background [19:10:30] it goes from the darker gray to the middle gray [19:10:35] right [19:10:42] the middle gray is the gray we are using for the editable area [19:10:48] but for the normal status the gradient is the otherway around [19:10:50] right? [19:10:55] thus, it fades to a middle (flat) shade [19:11:04] thus, it's depressed [19:11:22] i know i know... my point is a bit somewhere else [19:11:32] the dark color indicated the surface is at a downward angle (the light coming from the top) [19:11:40] if you take a look on it without the arrow-area [19:11:54] only the capsule with the gradient from light to dark [19:12:01] the middle gray is a surface which is perpendicular to the vewer (flat with the screen) [19:12:09] and underneath the template with flat gray [19:12:24] than it looks like the capsule is rounded [19:12:29] and the template is flat [19:12:40] which is intended [19:12:45] what's wrong with that? [19:12:56] thats intended, right [19:12:59] yes [19:13:02] and the bottom of the capsule is a bit deeper than the template [19:13:09] agree with that? [19:13:30] deeper? well no, it ends up at the same depth as the template [19:13:53] it's just that the vertical center of the capsule is closer to the viewer [19:14:07] and at a similar angle as the editable area [19:14:08] but it is darker - so for me that indictes that it is deeper [19:14:11] thus they are the same color [19:16:43] i ll do a screenshot to explain [19:21:20] nkomura, TrevorParscal: After fixing two obvious flaws, S&R actually seems to work quite nicely (in FF3 at least) [19:21:43] RoanKattouw: great to hear that [19:22:18] hannes-_-: http://img521.imageshack.us/img521/4022/lightq.jpg [19:23:04] A has a postive angle, the second B (oops, I meant C) has a Negative angle to the light [19:23:07] so typing out new templates is still broken [19:23:15] the first B and D have the same angle [19:23:19] thus, they should be the same shade [19:23:23] A should be lighter [19:23:28] nimish_g: How? [19:23:30] C should be darker [19:23:56] make sense? [19:24:06] not the best illustration - sorry, I'm trying to be really quick [19:24:56] so your point is that the depressed state should be all the color B/D [19:24:57] flat [19:25:04] RoanKattouw: I was trying to detect whether a 'realchange' is a change where the user is typing inside an existing template or not. I figure that means the start position will already exist and the end position would change. I look for anything where the start position is the same and I just don't touch it, which works fine [19:25:12] which could physically make some sense [19:25:27] BUT that also means that if you're typing a brand new template it won't get wrapped [19:25:32] nimish_g: Ah that, yeah I'm behind on CR, I haven't looked at anything committed after lunch yesterday [19:25:44] my point is that I want it to look more depressed and stand out a bit more so I wanted to use shading to make it more dramatic [19:25:49] Right so you need to accommodate for that use case [19:26:14] OK so at least this breakage isn't my fault :) now I can continue my dialogs work without guilt [19:26:31] RoanKattouw: ya pdhanda left a msg on one of the code commits about how it breaks wrapping newly created templates. it also broke the old 'enter' behavior...I'm working on getting bindings to stay when you hit enter now [19:26:37] lol yeah, not *this* time =) [19:26:49] :) [19:27:27] I see pdhanda is trying hard to replace me as the light-speed code review hawk :) [19:28:26] nono i would not dream of that.. i am too much of a slacker ;) [19:30:06] TrevorParscal: good graphic. i used it and sent it back to u to explain what I mean [19:32:39] when you say B you mean the first B (sorry about that typo) [19:32:46] http://img85.imageshack.us/img85/1228/profilesf.jpg [19:33:54] I'm proposing using B in this illustration for the name and dialog button in the capsule and C for the expand/contract button [19:34:07] hmm ok I see the difference about you and me sight if it [19:34:19] it does not look like B to me [19:35:36] Well, the name and dialog button of the capsule when expanded are < light - middle - dark >|< middle > [19:36:16] so? [19:36:31] we are just deciding if the expand/contract button should be < middle > | < middle > or < dark - middle > | < middle > [19:36:43] I think we both have some points [19:36:57] actually your graphic is right. but it does not look like the profile you made [19:37:21] yes I think so too.... I know what you mean - but I just can t look at it like you mean it [19:37:32] it s linke one of those switch-view-imgaes [19:37:52] oh, an optical illusion? [19:37:52] ha ha [19:38:15] RoanKattouw: we don't have a method to let us know what element the cursor is in, do we? [19:38:43] No [19:38:54] And be careful writing one, it needs to be O(1) not O(n) [19:39:00] That requires different approaches in different browsers [19:39:07] hannes-_-: it would be a little better if the editable area was white [19:39:14] http://rlv.zcache.com/young_woman_or_old_lady_optical_illusion_flyer-p2449584234590172452mcvz_400.jpg [19:39:25] Stuff like getCaretPosition() -> getOffset() will be O(n) in Firefox or IE (don't remember which offhand) [19:39:31] yes, optical illusions are fun - ha ha [19:40:17] if I want to throw up random technical notes for the benefit of mostly internal users, where would that go? [19:40:28] i.e. which wiki [19:40:33] hannes-_-: we can just leave it be for now [19:40:42] I think we've discussed it plenty [19:40:49] agreed [19:40:52] I'm not THAT worried about it :) he he [19:41:07] no me neither [19:41:07]