[14:27:32] YaronK: I did the promised rewrite of the TemplateInfo API module, and also fixed stuff elsewhere [14:33:29] roan ... Malayam ... is er any chance of us supporting webfonts ? [14:33:52] I don't know, I think there was a discussion on wikitech-l about that some time ago [14:34:03] Simetrical might know more about it, he was active in that thread IIRC [14:34:14] But I recommend you just dig up the thread in the list archives [14:38:45] the result was one of "not now, we do not have a good case for it" [14:39:15] ... with Malayam we have one of the top 50 biggest languages benefitting [14:39:37] benefitting in a way that is otherwise not feasible [17:41:43] *werdnus waves [17:43:36] howdy [17:56:54] TrevorParscal: I'm looking to filter some wikitext/HTML to allow only simple stuff in style rules like font-size, color, background-color, and other stuff that makes sense in a signature. [17:57:01] any thoughts on how I might do that? [17:57:34] I assume you already considered using regular expressions [17:57:52] there's too much to match against, isn't there? [17:58:23] I would also say, what you do is detect attributes in HTML tags, and only allow some, then in the content of a style attribute only allow specific rules to be set [17:59:07] you mean writing regex for css is a problem? [17:59:11] well, we already have some code written to validate style attributes [17:59:16] s/style/HTML/ [17:59:28] hmm [17:59:38] At this point I'm looking specifically to add additional validation to the style attributes. [17:59:52] the rest is handled in the wikitext parser (expression() and what-not) [18:00:47] I'm not sure that I can add additional filters to style validation in the parser [18:00:56] so you suggest just regexing :/ [18:01:29] I wonder if there are any valid places for ; to appear other than in comments (for css) [18:01:57] attribute regex: /\ [a-z\-]+\=[\'\"]?[.*]+[\'\"\ \>]/gi [18:02:07] TrevorParscal: did you write the autoEllipse plugin? [18:02:18] He adapted it from some other source [18:02:18] roan and I did together [18:02:26] yes, that too [18:02:52] cool, is there a way to run it on text that it's already been run on, to update it when the space changes? [18:02:58] I only refactored it to support multiple places for the ellipsis and be a jquery plugin [18:03:08] the orig [18:03:15] i've got the ntoc drag-resizing and want to update the truncation when the drag completes [18:03:24] I'm not sure the plugin supports that right now, but it'd be easy to add [18:04:00] the original text is in the title attribute [18:04:01] always [18:04:08] oh [18:04:11] not always [18:04:12] No, only if ( options.tooltip) [18:04:18] We can cram it in a data() [18:04:24] *RoanKattouw does it [18:04:29] we should [18:04:33] I don't see that we do [18:04:35] but we could [18:04:46] then you could always re-esllipse it [18:05:25] maybe it'd be worth just adding a check to the plugin to see if that variable is set in data [18:05:56] if it is, run the auto ellipsing on the variable, if not run it on the text in the element [18:06:27] yes, but also provide an override in the options [18:07:06] Override for what [18:07:19] Current behavior for re-ellipsing is crappy anyway [18:07:22] because you could potentially get in a situation where you change the text some other way, then re-ellipse it (another logical reason for re-ellipsing the text of course) and then the data with the old string would end up getting used [18:07:29] Hah [18:07:31] Good one [18:07:44] heya [18:07:50] Evening nkomura [18:08:09] TrevorParscal: I'll just check if .text() is a prefix or origText (disregarding the trailing ...) [18:08:11] sorry i'm late [18:08:31] is everyone online [18:08:44] RoanKattouw: that could still cause the same problem right? [18:08:45] Hi [18:09:14] hi parutron_ [18:09:17] TrevorParscal: Theoretically, yes. But only if you replace the text with a prefix of itself which is still too large, then resize [18:09:34] strange, i appear to have an underscore hanging on. i guess i've grown an IRC tail! [18:09:53] Plus, in practice, NTOC doesn't change .text(), it just regenerates the whole element IIRC [18:10:13] adam_miller, are you there? [18:10:18] I mean, if the text is "Something here", and it gets ellispsed to "Something..." and you change it to "Something there" - you fall under the same issue [18:10:23] nkomura: i am [18:10:33] True [18:11:00] hi adam_miller [18:11:04] i guess we are missing hannes [18:11:11] TrevorParscal: Basically, if ellipse(a) == ellipse(b) && a != b , you have a problem [18:11:45] hi nimish_g [18:12:00] shall we get started, team? [18:12:21] Yup [18:12:57] hi hannes-tank [18:13:11] so the agenda is: [18:13:23] *Status of Acai enhancement release (Adam, Roan) [18:13:23] (With the current state of the code review and November 11th is a holiday for WMF, so let's shoot for November 18th) [18:13:23] *Beta feedback overview (Howie) [18:13:23] Will need to discuss high traffic languages which are not reaching average retention rate.?? (German, Japanese, French, Italian, and Polish) [18:13:23] *Babaco status (Trevor, Roan) [18:13:25] *Usability Study Feedback (Parul) [18:13:27] *Toolbar re-organization (Hannes, Adam) [18:13:29] *Planning for early December release (all) [18:13:31] *Planning for the usability study for content folding in early December [18:13:39] howie can't join this morning though [18:13:51] so let's get started on acai enhancement status [18:13:52] are we talking in this channel? [18:13:58] sorry for beeing late by the way [18:14:03] np [18:14:17] OK [18:14:40] do you folks mind using this channel? [18:14:44] works for me [18:14:46] likewise, I was a little late too =( [18:14:57] fine for me too [18:14:58] In terms of acai enhancements, we're in pretty good shape, since Brion reviewed pretty much everything yesterday [18:14:58] aha! [18:15:01] we can use it... i dont mind [18:15:03] I think we sort of own it - as much as one could own an IRC channel on someone else's server... [18:15:06] cool! [18:15:11] yeah nothing looked too explody [18:15:22] thanks brion! [18:15:26] We've got a grand total of 2 unreviewed revisions in acaifix [18:15:48] so all adam_miller's stuff checked out too? [18:15:56] Yup, except one (his latest) [18:16:05] which was? [18:16:11] Not reviewed [18:16:14] i had one thing to fix, and did yesterday which is the what Roan is talking about [18:16:25] http://www.mediawiki.org/wiki/Special:Code/MediaWiki/58451 is unreviewed [18:16:27] what kind of fix is it? [18:16:32] That's a rather trivial fix [18:16:35] minor fix to toolbar config [18:16:44] altering the way the indent button behaves [18:17:00] Same story for r58441, fixes an autoellipse bug in the NTOC, also trivial [18:17:19] how is the indent altered? [18:17:45] changed how it behaved when the cursor is sitting in the middle of a line [18:18:00] is it viewable on prototype? [18:18:06] before it was just inserting ':', now it inserts a new line starting with ':' [18:18:15] Yes, and on prototype/deployment too (link in channel topic) [18:18:31] Like tables and such, the indent icon now inserts its stuff on a line by itself [18:18:41] k, will check out after the meeting [18:18:49] yeah, that's the ownline: true part [18:19:11] any open bugs for acai enhancements? [18:19:32] *RoanKattouw looks [18:19:40] Might be a few IE or Safari-specific ones in there [18:20:02] There's https://bugzilla.wikimedia.org/show_bug.cgi?id=20867 , weird bug on an old Safari version [18:20:32] There's a couple in there that I don't really believe but can't reproduce because they're IE-specific [18:20:40] is this "jumping" effect when you press a toolbar button on your buglist? [18:20:54] Yes [18:20:57] the indent button seems to "jump" the text entry box [18:21:02] tee hee, jinx. [18:21:19] bold does too... etc [18:21:34] That bug and a load of other technically related bugs will be fixed with something we call the "magic iframe" [18:21:40] Trevor can tell you more about that [18:21:47] It's magic [18:21:58] Basically, instead of using a textarea [18:21:59] that's all i need to know ;) [18:22:00] the magic iframe is a wonderous thing, I think me and trev will spend a good part of today testing it [18:22:10] We use an iframe in design mode [18:22:29] which is the same thing we are going to be doing for Citron, only in citron we will be doing much more complex things with it [18:22:54] It will get rid of quite a few ntoc bugs [18:23:01] i'm guessing this means no more trying to measure things (like with ntoc) [18:23:02] Yes. We basically decided that there was no reason we can't put the new editor thingy in place earlier, even if it doesn't have syntax highlighting just yet [18:23:06] Exactly [18:23:14] we will have more control over the rendering and interaction of the contents of the iframe, and none of the strange browser behaviors that we get with the textarea (like scrolling the page when it's focused or whatever) [18:23:16] No more guesses and fragile calculations [18:23:19] are iframes already supported by all common browsers? [18:23:30] already? I thought iframes were old? [18:23:41] so guys [18:23:44] hannes-tank, yes [18:23:47] thought its new.-- any way it seems to be save [18:23:56] In short, it's likely that we will support MORE browsers, not less [18:23:57] introduction of iframe will not be rolled out until the release after [18:24:01] Yes [18:24:12] so can we stick to acaienhancements discussion for now [18:24:22] we'll discuss babaco enhancements later [18:24:25] But we're leaving the associated bugs open for now because it's not worth the effort fixing them [18:24:32] right [18:24:33] *TrevorParscal completes his magical tour of magic iframes [18:24:40] so this safari bug for the link status [18:24:46] we can defer it [18:25:13] The thing is the whole link dialog is gonna change, we don't know if that bug will occur for the new dialog in the same way, or at all [18:25:23] right [18:25:40] we are not including ling dialogue in acai enhancement release, right? [18:25:44] Right [18:25:48] no [18:25:54] i originally wanted, but will need to test some more [18:25:57] In fact, I've already closed a bug with "Fixed in new dialog, not bothering to fix the old one" [18:26:08] and look & feel need to be worked on [18:26:16] Yes, the new table dialog has some alignment weirdnesses, I need to shove that in Trevor's face some time [18:26:40] yes...and i think it'll be better if we consolidate the look and feel enhancement in one release. [18:26:41] you can just show it ;) no need to shove it [18:26:48] tee hee [18:26:50] right [18:26:53] i like meeting over IRC [18:27:15] so wrapping up acai enhancement release [18:27:17] me too much easier to understand everything ;) [18:27:29] Roan is welcome to shove UI in my face, but that's about it... [18:27:36] I've got one more thing to say about the acai enhancements [18:27:41] adam_miller: did you check non-english rendering too? [18:27:49] shoot RoanKattouw [18:28:00] With pretty much all of that stuff reviewed, I think we may be able to deploy earlier than Nov 18 [18:28:27] let's discuss the schedule in a bit [18:28:35] there is that one issue with the lines on the tabs on the german prototype...i'm making an note to check that out today [18:28:38] hannes_tank, also I need to give you a tour of the clicktracking tool, as it will be available now [18:28:48] ok cool [18:28:53] YAY [18:28:53] lets do that after the meeting? [18:29:01] adam_miller, it would be really helpful if you can check out arabic prototype too [18:29:16] so schedule wise [18:29:54] shall we shoot for the 10th? [18:29:57] the thing you noticed about the line in german prototypes has to do with a userscript [18:30:01] just so you know [18:30:12] TrevorParscal: really? [18:30:21] yeah i spotted that a while ago, but had no idea how to change it [18:30:37] Roan or I can help you [18:30:47] not sure we are talking about the same line problem [18:30:58] but i do not run any userscript in my browser [18:30:59] TrevorParscal and I are [18:31:08] I think TrevorParscal means site JS? [18:31:25] well either way we'll talk about it later today and get it fixed [18:31:35] cool [18:31:37] not nec. for everyone to be involved in that discussion [18:31:47] sounds boring. [18:32:09] (just kidding) [18:32:10] The 10th (next Tuesday) sounds good to me [18:32:11] can you guys check acai enhancements in RTL? [18:32:25] hinting -> adam_miller [18:32:43] You guys won't be around the day after (Veterans Day), but I will, so at least there'll be someone to clean up any mess [18:32:45] RoanKattouw: yeah, it's a site style [18:32:46] yeahi have been [18:32:59] kool [18:33:17] i had to fix collapsible tabs to work in RTL [18:33:57] has the fix for RTL reviewed yesterday by brion? [18:34:06] yeah that was committed a while ago [18:34:12] nice [18:35:19] RoanKattouw: are you planning to do the deployment on the 10th? [18:35:42] Sounds like a good dat [18:35:44] e [18:36:12] I don't feel it's my call, but I think it'll work, and I'll be around that day [18:36:34] TrevorParscal and adam_miller, does the 10th work for you? [18:37:04] um [18:37:06] we can touch base around 10am(PST), and roan can start if we are good to go [18:37:08] i'm around all day on the 10th, works fine for me [18:37:12] with a holday the next day? [18:37:25] yup, wednesday is a holiday [18:37:51] we deploy on the 10th, then take a holiday :( I mean, we can do that... sure [18:37:53] we(roan) will revert back if the release does not look good [18:37:54] sounded like RoanKattouw was going to be standing by and ready to put out fires? [18:38:01] why not deploy on the 9th? [18:38:12] That's a Monday :) [18:38:23] Works for me but I'm not sure how the ops folks feel about that [18:38:46] I'm fine with the 10th if you all are [18:39:05] i'm arriving monday afternoon, so i'll be in transit [18:39:23] we can shoot for the 12th [18:39:27] which is thursday [18:39:52] does anyone remember when the fundraising start? [18:39:57] at least we are there on the 13th as well [18:40:00] rand sent out the revised schedule last week [18:40:01] checking [18:40:19] 10th [18:40:26] 4pm [18:40:31] yeah [18:40:34] 4 pm? [18:40:44] Why that late? [18:40:47] i thought that time was so late [18:40:53] no idea [18:41:10] k, let's not definitely plan on the 10th [18:41:24] so the 12th then? [18:41:27] 12th has my vote [18:41:52] Fine by me [18:42:00] adam_miller, does the 12th work for you? [18:42:20] i'll be out in the afternoon [18:42:28] oh [18:42:42] so not really, but i'm not sure how important it is to have me on hand anyway? [18:43:23] mostly just so we can blame things on you :P [18:43:35] RoanKattouw: do you feel comfortable even adam_miller is not around on Thursday [18:43:38] ? [18:44:03] Yes [18:44:08] I'll be doing it in the morning anyway [18:44:23] yeah, but that's adam's afternoon already [18:44:33] Ah yes of course [18:44:55] so let's plan on starting 10amPST on the 12th [18:45:02] i'll put on the calendar [18:45:11] OK [18:45:12] is there ops mailing alias? [18:45:24] private-l, I'll do a post after the meeting [18:45:29] thanks [18:45:39] i'm not on it, that's a mysterious list [18:45:49] OK, next topic [18:46:10] howie shared the beta survey analysis [18:46:25] so we suck with Japanese, German, French, Italian and Polish [18:46:43] we know that small fonts are pointed out a lot form Japanese survey respondents [18:46:50] well i wouldn't say *suck* [18:47:04] in 60's... [18:47:20] I'm running translations on that feedback right now...some of it's not so great =\ [18:47:25] we are not performing well [18:47:31] ok that's so not the emoticon I wanted [18:47:33] for IE6.. [18:47:42] 80s for other browsers [18:47:58] suck == < 50% [18:48:02] :) [18:48:22] k, my word was too harsh. i apologize [18:48:23] just eyeballing it, most of them say "it's too small to read well" in some form or another [18:48:35] nimish_g: for japanese? [18:48:42] or others too? [18:48:44] nimish_g: what language?? [18:48:47] Anyway, nkomura suggested Ryan_Lane set up prototypes for it and pl [18:48:53] Sounds like a good idea to mee [18:48:54] and fr [18:48:55] text-size: 0.8em becomes text-size: 0.9em - done [18:48:57] chinese for sure, it looks like arabic too [18:48:58] i didn't know that was getting translated. [18:49:08] non roman fonts [18:49:15] generally maybe? [18:49:21] japanese wikipedia was using large font to start with [18:49:31] so the change might have been too much for readers [18:49:40] Yeah; they probably raised the font size in Monobook.css [18:49:42] I haven't looked at japanese yet, just flipping thru the top of the file "ar" and the bottom "zh" =) [18:49:45] Which means Vector isn't getting it [18:49:57] nimish_g: can you point me to where you're seeing the translated feedback? [18:50:01] TrevorParscal: are you implying universal change? [18:50:09] or just a specific language? [18:50:09] parutron: I'm running translations on it now [18:50:22] you could do it on the project sites [18:50:27] with some CSS [18:50:36] btw, nimish_g, thanks for helping with the translation [18:50:39] that's really great [18:50:47] http://de.wikipedia.org/wiki/MediaWiki/Vector.css [18:51:02] add some line that overrides a specific text-size rule, and you're done [18:51:17] adjustments where needed is normal [18:51:22] happens with monobook as well [18:51:31] every project has made tweaks here and there over time [18:51:33] it looked like the german lower retention was notably less in IE6 too [18:51:44] that sentence made no sense! [18:51:57] well, IE6 users don't get the edit tools, but they do get the new skin [18:52:09] ah. [18:52:29] parutron, I translate that as: "It looks like fewer Germans kept it, even using IE6" [18:52:30] so, they only got part of the beta [18:52:33] TrevorParscal: i'm not following you [18:52:37] we need to take that into consideration in the data analysis plz [18:52:43] i mean about the text size change [18:52:54] i'll mention it to howie. [18:53:03] nkomura: we can just discuss it later then [18:53:11] after reading his analysis i thought that was one of the most actionable items. [18:53:20] if that's OK - it just is sort of technical / out of scope [18:53:45] i don't think we have enough data for de.wp that text size is the issue [18:53:53] from the feedback on the usability wiki [18:54:10] overalapping tabs were more raised than text size for de [18:54:34] so hopefully with the next release, adam's cascading tabs will make it better [18:54:46] which will be solved by collapsing tabs and simplesearch [18:54:47] i mean make german editors happier [18:54:53] that gets a YAY [18:54:59] right [18:55:19] so i'm a little worry to adjust german text size as precaution [18:55:30] that's why i was asking TrevorParscal what adjustment he is implying [18:55:37] i haven't ready ANY translations but i would be too [18:55:44] (translations of feedback) [18:56:06] nimish_g is working with howie on semi-machine translation [18:56:14] just a hunch that more of the font size issues are for non-roman languages [18:56:20] i only read japanese so far [18:56:44] no data for korean or chinese yet as far as i know [18:56:47] nkomura: I'm saying, if people say some particular text is too small - we can go onto the project they experienced that on, like chineese or arabic or whatever and adjust the text size for that particular element on that particular wiki [18:56:47] nihongodewa doo desuka? [18:56:52] though i'm sure people in *every* launguage have opinions on the font size...;) [18:56:56] but nimish_g mentioned something about chinese [18:57:54] yeah, I'd just been testing with chinese feedback and most of the translations were about how the font was too small, but the ones I looked at didn't go into specifics [18:58:34] parutron: what do you recommend about catering to those needs [18:58:51] if people are opting out because of the font size [18:59:09] i think we should look at it on per-wiki. [18:59:13] I liked TrevorParscal's suggestion about just changing site-specific CSS, maybe we could do that for specific languages [18:59:14] we are not meeting the needs of the language community [18:59:22] I say we check whether there's an actual difference, and if there is, we fix the site CSS [18:59:27] agreed! [18:59:47] sounds like a plan [18:59:49] but lets not just do this based on a few translations [18:59:59] yeah [19:00:00] lets wait until we have some solid survey data.... [19:00:06] it is too early to make conclusion yet [19:00:13] agreed! [19:00:20] k [19:00:28] so french, italian and polish [19:00:40] i have no idea why the adoption rate is lower than average [19:00:57] so i suggest 1) we build prototypes for these languages [19:01:10] and 2) reach out to these language communities and find more feedback [19:01:22] the survey translation will be helpful too [19:01:31] are you guys have any insight for these languages? [19:01:35] i can try and spearhead no2. but i think the first place to start would be the survey data [19:01:53] yup [19:02:08] i'll get caught up from howie + nimish_g on translation.... [19:02:09] and howie are placing priority for those languages [19:02:15] cool [19:02:28] once we get something from that - i can try and gather more information where needed... [19:02:42] we'll dig in more [19:02:45] maybe they are culturally resistant to change - in comparison to other more happily adapting cultures? I have no idea, but this is a possibly interesting factor... [19:02:58] i saw an email about a lack of space. can we make more space for those language prototypes?? [19:03:17] rob needs to order more space [19:03:17] parutron: that should be solved by now [19:03:21] not sure if that order got placed [19:03:31] RoanKattouw deleted the video from wikimania [19:03:36] Freed up like 3G [19:03:46] nice. [19:03:46] maybe french and italian didn't like the fact we didn't build prototypes ;) [19:04:17] anything on beta, folks? [19:04:18] tee hee [19:04:30] ? [19:04:52] moving on to the next topic [19:04:55] babaco - [19:05:05] TrevorParscal and RoanKattouw, you are up [19:05:14] so share what you guys are working on [19:05:18] State of babaco enhancements? [19:05:25] right [19:05:26] magic iframe.... [19:05:34] sort of already covered that [19:05:43] Yeah [19:05:47] that should help fix allot of things / make things better [19:05:55] Other things are the new link and table dialog, which you've probably already seen [19:05:57] I need to cleanup the ui on the new table dialog [19:05:59] so this change applies to NTOC [19:06:00] ? [19:06:13] or for dialogues as well? [19:06:14] it's accuracy, yes [19:06:28] it's magic indeed... [19:06:36] dialogs won't be affected by the magic iframe [19:06:54]