[01:13:46] hey guys. is it possible to, er, enable micheal dale's media wizard gadget on our project wiki? [01:20:01] i see. [01:20:03] you can all run [01:20:05] but you can't hide [01:20:19] colloquy can tell me where you are [01:40:55] parutron: usability wiki now features add-media-wizard [01:41:18] it's like media wiki christmas! [01:41:33] i am gonna be the first user. [01:42:38] mdale1: do i need to enable it [01:42:43] or is it installed on the whole site? [01:42:49] parutron: nope should be installed for the whole site [01:43:08] and it starts with the "embed file" button? [01:43:16] correct [01:43:18] reload [01:43:19] thanks [01:43:26] shift-reload [01:45:11] i'm sure this doesn't help BUT it might be nice to have a link to leave feedback. [01:45:22] just putting it out there [01:48:41] for add-media-wizard? [01:50:19] there is: http://www.mediawiki.org/wiki/Extension:Add_Media_Wizard and its talk page [18:00:28] adam_miller: how's it going? [18:00:49] good, fixed the ntoc this morning [18:01:01] for firefox at least [18:01:25] right on [18:01:35] trying to get the list to build correctly in safari now [18:01:53] have we figured out why the cursor seems to change from the pointer to the arrow? [18:02:00] and - we need it to highlight... hmmm [18:02:07] yeah still a lot to fix [18:02:13] but at least clicking works again [18:02:20] you were saying the highlight comes from the cursor moving though... [18:02:22] yes [18:02:24] awesome [18:02:54] the highlight does come from cursor position [18:03:01] we need to figure out a way to make the TOC get smaller to avoid wrapping the advanced, special characters and help section links when the window is resized [18:03:19] oh interesting [18:03:36] maybe like a minimize function [18:03:58] when it hits a threshold we send it to it's min width before it will collapse [18:04:21] yes, but also just adjusting the width [18:05:01] it would be ideal if it was elastic, so as it adjusts it, it's still remembering the wider setting, so that if the window is sized large again that wider setting is eventually restored [18:05:28] gotcha, so as it's resized, it maintains the proportions but shifts the widths [18:05:30] this logic can all just go in the evt.resize handler [18:06:02] k i'll add a todo for that [18:06:55] like, if the user sets it to 100px wide.. then resized the browser so that the TOC needs to scale down to 80px to prevent the section links from being wrapped, then if the browser is sized back up, the TOC will get wider until it gets to 100px [18:07:20] k that should be simple enough [18:07:22] basically we want to store the "user's intent" width somewhere, which is always attempted but not forced [18:07:25] :) [18:07:26] i'm already storing those values [18:07:30] right on [18:07:39] yeah. I see like in the cookies and such [18:07:49] for when someone collapses the ntoc and then reopens it [18:08:00] yes [18:08:11] do you have any code not checked in right now? [18:08:21] nothing of consequence [18:08:40] just a lot of debugging stuff as i'm trying to sort out webkits issue with textnodes in the iframe [18:08:47] I want to move configuration things for modules into a cfg member of the module to encapsulate them [18:08:53] should be simple diffs [18:09:03] have at it [18:23:47] adam_miller: I haven't looked at anything yet, but if you really did fix the NTOC, I owe you some stroopwafels ;) [18:24:35] TrevorParscal: Are you aware of the wikiEditor configuration bug? *looks up bug #* [20:18:01] TrevorParscal: Ugh, you've probably caused a conflict for me in templateEditor now :( [20:26:58] No worries, resolved now [20:33:55] k [20:34:21] there's still lots of things in that module that could be written in less code and use less memory [20:34:32] FYI, I'm now working on debugging an algorithm I wrote that wraps stuff in the iframe DOM [20:34:44] As opposed to using setContents() that just tears down the entire iframe DOM and builds it anew [20:34:50] hows the branch thing going? [20:35:01] awesome [20:35:12] Haven't started yet. Basically you just want the current deployment code plus a few extra extensions, right? [20:35:18] Shouldn't be much trouble [20:35:51] yeah [20:37:23] you like those combines and bumps on those revs though don't you! [20:37:31] Yeah dude they're awesome [20:37:35] ha ha ha [20:53:23] mdale1: are you a replicant or a human? [20:54:05] it does not matter whether ~ i think~ i am replicate or human .. [20:54:24] ha ha ha [20:54:47] after all could have memory rewritten or what not [21:17:15] TrevorParscal: css ? [21:18:53] css? [21:19:10] conflicting style sheet.. want do you have a second to take a look at that [21:19:24] yeah - so... it would be freakin sweet if we could have a way to not load the CSS for jquery ui if it's already there.. [21:19:55] Why not just make each style sheet specific to only wikiEditor dialogs and AMW dialogs respectively? [21:21:18] ~ they both extend jquery ui ~ [21:21:34] but they could extend it in specific ways. [21:22:08] ~ I try not to override any of the base jquery ui stuff ~ [21:22:47] which is why you can have two players in the same page: http://prototype.wikimedia.org/s-2/js2/mwEmbed/example_usage/Player_Themable.html [21:22:53] both using jquery ui base theme stuff [22:46:14] TrevorParscal: i'm having some scrollTop iframe cross browser issues [22:46:27] have ye any experience with such issues? [22:46:41] hmm, not specifically, what's going on? [22:47:12] adam_miller: Did scrolling the html work for you? That's weird, I swear scrolling the body worked for me [22:47:36] i finally got safari rendering the ntoc - had to parse the contents and do some replacement on it to get it to be the same between webkit and gecko [22:48:00] BUT it seems the trick i used to get scrolling working again in firefox does nothing in safari [22:48:36] by trick, i mean i just was setting scrollTop on iframe > html, instead of iframe > body [22:48:41] Yeah [22:48:47] I thought the latter worked in Firefox too [22:49:00] It'd be very weird if it didn't; after all, I wrote that code, and I only had Firefox at the time :) [22:49:07] weird thing is when i switch it back to scrollTop on iframe > html, then safari scrolls it, but it does it inconsistently [22:49:57] i said that backwards -- switched it back to iframe > body, then safari scrolls [22:50:28] RoanKattouw: FF wasn't doing anything on click and I tracked it down to that code, when i switched it to scrolling on html instead of body it started working perfectly again [22:50:35] Hm [22:50:38] Which FF version? [22:50:42] maybe there's something more complex going on? [22:50:43] 3.5.5 [22:50:46] Could be [22:50:48] on a macintosh computer [22:50:49] No idea [22:50:51] Aaah [22:50:57] Could it hurt to just scroll both? [22:51:08] You could try that and see what happens [22:51:35] ha that was my first hack attempt [22:51:45] i dont think it does hurt actually [22:53:14] safari is just screwed up with the iframe selection stuff still [22:53:15] i'm gonna blame it on that [22:54:49] like i double click a word, and it highlights, the deselects itself a second later :) [22:55:27] Probably because there's DOM manipulation going on [22:57:12] well maybe i'll commit this code and get out of there [22:57:15] let you and trevor have your way with it [23:06:05] TrevorParscal: is there some timestamp checking going on the makefile? [23:06:12] Yes, make does that for you [23:06:19] Use make -B to force a full rebuild [23:06:26] cool thanks TrevorParscal [23:06:29] Most common case is where you change stuff, then svn up [23:06:39] i love how closely you watch this RoanKattouw [23:06:41] ha ha ha [23:06:44] Cause svn touches all the timestamps [23:06:50] I was just gonna say something to Trevor :) [23:06:53] it's much more helpful then Trevor [23:07:07] *TrevorParscal is useless in RoanKattouw presence [23:07:35] TrevorParscal: I didn't do the branching thing today, sorry about that. Could you write me an e-mail detailing what you want in that branch (and really do it this time xD)? [23:07:46] Then I'll get it done tomorrow [23:09:35] just committed: hot shit. [23:10:53] i now have a iframe induced headache [23:11:33] Does it by any chance involve multiple newlines in succession being collapsed to one? [23:11:38] soprry [23:11:41] a iheadache? [23:11:46] sorry.. was rewriting algorithm [23:11:51] If so, read the CR comment I just submitted [23:12:55] by hot shit i meant terrible, poorly-regexed roan comment bait [23:13:44] but you know, progress nonetheless [23:14:20] No worries, I'm also committing code with some TODOs and FIXMEs left in [23:17:12] nimish_g: Did you know that your template parser was fundamentally broken for pages with multiple templates on it? It assumed the token stack was sorted by occurrence, but due to the way it was built and a lack of sorting, it was really sorted by token type [23:17:31] RoanKattouw: I just rewrote it... [23:17:40] Oh right [23:17:41] but the occurance issue is not fixed yet [23:17:52] I'm committing a crapload of code right now that also fixes that [23:17:58] RoanKattouw: hm... I tested it with multiple templates the other day and it worked fine... [23:18:03] Weird [23:18:05] RoanKattouw: grrr [23:18:10] wait [23:18:13] which [23:18:19] the order of tokens == yay! [23:18:22] parser == boo [23:18:23] :) [23:18:27] M UsabilityInitiative.hooks.php [23:18:29] AM css/wikiEditor.highlight.css [23:18:30] M js/plugins/jquery.wikiEditor.templateEditor.js [23:18:32] M js/plugins/jquery.wikiEditor.highlight.js [23:18:33] M js/plugins.combined.js [23:18:35] M js/plugins.combined.min.js [23:18:49] you just commited? [23:19:00] No, going to in a few seconds [23:19:06] well, hold off [23:19:11] cause my changes are much more simpl;e [23:19:25] it will be easier for you to merge yours them with mine I think [23:19:31] Dammit [23:19:39] It's already running, no stopping it now [23:19:47] grrr [23:19:50] Sorry dude [23:20:04] You can send me a diff of your changes if you want, I'll integrate it [23:21:10] Just don't update and run svn diff [23:22:52] i just committed a small fix to that regex with recombined plugins script - hope that didnt cause problems [23:22:54] adam_miller: Does that work? Shouldn't the brackets be removed? [23:23:12] The way I read this that'll match either a CR, a LF or a ? [23:23:23] oh maybe they're unnecessary [23:23:27] it does work this way [23:23:42] I think they even break stuff; try messing with an article that contains ? characters [23:23:54] TrevorParscal: What were your changes then? [23:24:27] hmm... Error: WRONG_DOCUMENT_ERR: DOM Exception 4 in wikiEditor.highlight [23:24:47] that's breaking things and preventing me from testing it now [23:24:55] Which browser? [23:25:33] RoanKattouw: I'm just merging my stuff by hand [23:25:53] safari [23:25:59] firefox seems fine [23:26:04] adam_miller: Comment out lines 32, 33, 45, 46 of jquery.wikiEditor.highlight.js [23:27:13] right, it replaces ? with
[23:27:29] recommiting w/o the [ ] [23:28:18] RoanKattouw: cursor keeps being set to the begining of the iframe [23:28:36] Before or after my commit? [23:29:37] after [23:29:38] Also, would your page source happen to begin with a template?> [23:29:40] looking into it [23:30:01] It's probably got to do with the template being wrapped ad infinitum [23:30:09] At every keypress, it's being wrapped in a fresh

[23:30:14] yeah [23:30:15] that's bad [23:30:18] So my DOM now has like ten layers of

s, then the template [23:30:25] Yes, and there's a TODO in the code about that [23:30:28] but it's doing it on source without the template in it [23:30:32] I was planning on tackling that tomorrow [23:30:37] Hm [23:32:54] That's weird [23:33:10] Can you tell whether it's wrapping anything? [23:33:28] Specifically, whether it's manipulating the DOM near the top [23:33:40] If all else fails, just comment out lines 32,33,45,46 of highlight.js for now [23:35:48] goodnight gentlemen - gotta head home [23:38:34] RoanKattouw: check out my latest commit [23:38:43] please check for merge errors [23:39:04] http://www.mediawiki.org/wiki/Special:Code/MediaWiki/r60106 [23:40:44] i wiped out nimish's stylize [23:42:09] Heh [23:42:25] Not that it worked, though, there were some glaring errors in there [23:42:39] Well, two [23:43:12] + // Collect matching level 0 template call boundries from the tokenArrray [23:43:13] ok [23:43:17] nobody commit anything! [23:43:18] There's a weird U [23:43:24] TF-8 char in there [23:43:35] hmm [23:43:37] oops [23:43:59] And boundry is written boundary, there is a function right near there that should've shown you that [23:44:56] ha ha [23:45:52] Meh I'll review tomorrow, too much code at 12:45 AM [23:47:33] TrevorParscal: What language is /that/? [23:47:39] http://www.mediawiki.org/wiki/Special:Code/MediaWiki/60108 [23:47:41] ha ha ha [23:47:46] Oh wait [23:47:52] *RoanKattouw fails at phonetic reading [23:49:58] OK guys I'm off to bed now. I'll fix the remaining issues with my code tomorrow, until then you can comment out the lines I mentioned and it should work fine [23:50:02] k [23:50:18] And sorry again for the coordination fail :)