| 00:18 | <Hixie> | http://www.whatwg.org/demos/ updated a bit (no new demos) |
| 00:50 | <MikeSmith> | does Firefox already have an extension similar to the Chrome "background pages" feature? |
| 00:50 | <MikeSmith> | http://code.google.com/chrome/extensions/background_pages.html |
| 00:50 | <MikeSmith> | and http://supercollider.dk/2010/04/chrome-extensions-for-web-hackers-part-%E2%85%A1-background-pages-255 |
| 00:50 | <MikeSmith> | and http://julianapena.com/2010/01/how-to-build-a-chrome-extension-part-4-background-pages-and-scheduling-requests/ |
| 00:51 | <othermaciej> | Safari extensions have an equivalent capability |
| 00:51 | <othermaciej> | Firefox extensions can do almost anything |
| 00:58 | <MikeSmith> | ok |
| 01:21 | <roc> | MikeSmith: if you're writing a Firefox extension, you could look at the Add-on SDK, which has a similar model by default --- http://code.google.com/chrome/extensions/background_pages.html |
| 01:21 | <roc> | erg |
| 01:21 | <roc> | wrong URL |
| 01:21 | <roc> | https://jetpack.mozillalabs.com/sdk/1.0b1/docs/ |
| 01:21 | <MikeSmith> | roc: thanks |
| 01:21 | MikeSmith | looks now |
| 01:23 | <MikeSmith> | roc: does it have a mechanism already for generating notifications? |
| 01:24 | <MikeSmith> | I mean platform notifications |
| 01:24 | <MikeSmith> | like Growl stuff |
| 01:24 | <roc> | https://jetpack.mozillalabs.com/sdk/1.0b1/docs/#module/addon-kit/notifications |
| 01:24 | <MikeSmith> | aha |
| 01:24 | <MikeSmith> | cheers |
| 01:24 | <MikeSmith> | cool |
| 01:26 | <MikeSmith> | btw, John Gregg recently updated the Web Notifications draft |
| 01:26 | <MikeSmith> | http://dev.w3.org/2006/webapi/WebNotifications/publish/Notifications.html |
| 01:26 | <MikeSmith> | http://lists.w3.org/Archives/Public/public-web-notification/2011Jan/0002.html |
| 01:26 | <jamesr_> | roc: fyi webkitRequestAnimationFrame is on in chromium ToT |
| 01:26 | <jamesr_> | i'm working on adding a timestamp parameter to the callbacks now |
| 01:27 | <roc> | great |
| 01:27 | <jamesr_> | and i'm gonna see if we can help evangalize it |
| 01:27 | <roc> | that's on elements, with per-element visibility checking? |
| 01:27 | <jamesr_> | it's on window, with an element parameter to request..() |
| 01:27 | <roc> | hmm |
| 01:27 | <roc> | why'd you do it that way? |
| 01:27 | <jamesr_> | it's cancellable |
| 01:27 | <jamesr_> | it returns an id and you can do window.cancel..() |
| 01:27 | <jamesr_> | like setTimeout |
| 01:28 | <roc> | ok but why did you put it on 'window'? |
| 01:28 | <jamesr_> | so having it on the element seemed odd, because you might want to cancel in cases where you didn't have a handy reference to the element |
| 01:28 | <jamesr_> | also it gets really weird if you move an element between documents |
| 01:28 | <jamesr_> | i'm not too attached to having it on the window vs having it on the element |
| 01:29 | <othermaciej> | you could return an object with a cancel() method instead of a dumb ID |
| 01:29 | <othermaciej> | that's a cleaner API design anyway |
| 01:29 | <roc> | I can see why canceling would be nice, but honestly no-one has yet complained about lack of cancellation as a problem |
| 01:30 | <roc> | did you have people asking you for cancellation? |
| 01:30 | <jamesr_> | yeah |
| 01:30 | <jamesr_> | othermaciej: agreed |
| 01:30 | <roc> | what for? |
| 01:30 | <jamesr_> | roc: same as for setTimeout(). sometimes you just don't wanna run the callback any more, and it's arguably nicer to have a cancel rather than checking some flag when the callback fires |
| 01:31 | <roc> | I guess |
| 01:31 | <jamesr_> | one consideration is that i want to make it super easy for developers to convert their setTimeout(.., 0) animations over to this API |
| 01:31 | <jamesr_> | because i want to get as many pages as possible off of doing setTimeout(..., 0); as possible |
| 01:33 | <Philip`> | http://blog.jgc.org/2011/01/code-injected-to-steal-passwords-in.html - that's terrible |
| 01:33 | <Philip`> | They should have used feature detection for XHR, not navigator.appName |
| 01:34 | <jamesr_> | roc, othermaciej: i'm not claiming to be an expert on making web APIs that developers want to use and am completely open to changing the way the API looks to be something nicer |
| 01:34 | <roc> | what about element.cancelAnimationFrame(callback), like add/removeEventListener? |
| 01:35 | <jamesr_> | what if you pass the same function in to multiple elements? |
| 01:35 | <roc> | then you'd have to cancel it on those elements? |
| 01:36 | <jamesr_> | ah, so [element, callback] tuples are assumed to be unique |
| 01:36 | <jamesr_> | that'd work for me |
| 01:37 | <roc> | they don't have to be unique, you just have to count them |
| 01:37 | <roc> | it's just like add/removeEventListener |
| 01:37 | <jamesr_> | right |
| 01:38 | <othermaciej> | if cancel is needed, I'd do element.requestAnimationFrame(..) that returns an object with a cancel() method |
| 01:38 | <othermaciej> | then you don't need to keep a reference to the element or even the window to be able to cancel |
| 01:38 | <jamesr_> | you always have window |
| 01:38 | <othermaciej> | and you can register the same function multiple times |
| 01:38 | <othermaciej> | and don't have to worry about ID |
| 01:38 | <othermaciej> | well, if you have same-origin iframes, you might be dealing w/ multiple different windows |
| 01:39 | <jamesr_> | true |
| 01:40 | <jamesr_> | so i think what should happen is heycam (or someone) posts http://people.mozilla.org/~cmccormack/anim-timing/Overview.html somewhere and we hash this stuff out on public-webapps/public-fx/whatnot |
| 01:40 | <roc> | I don't want to add a new object here for cancellation. As well as the API surface, that would mean we'd always have the overhead of creating that object. |
| 01:40 | <jamesr_> | as there's obviously some diversity of opinion :) |
| 01:40 | <othermaciej> | is this a method that would be called a lot in practice? |
| 01:40 | <roc> | and like I said, so far cancellation hasn't been an important requirement for our users |
| 01:40 | <jamesr_> | othermaciej: it'd be called once per frame |
| 01:41 | <othermaciej> | I don't know enough about the API to know if cancellation is important |
| 01:41 | <jamesr_> | roc: it was mentioned as a 'nice to have' and i've used it in my test pages |
| 01:41 | <roc> | ok |
| 01:41 | <othermaciej> | I see - why isn't there a way to just set it and have your callback fire for all future frames? |
| 01:41 | <jamesr_> | but we could start without cancellation and add it |
| 01:41 | <roc> | literally no-one has mentioned cancellation to me as a requirement before :-) |
| 01:41 | <othermaciej> | (if it's for one frame, cancellation seems rather inessential) |
| 01:41 | <jamesr_> | seems kind of weird to have a way to ask for a callback and no way to say "never mind" |
| 01:41 | <roc> | it's for one frame |
| 01:42 | <jamesr_> | othermaciej: it's setTimeout() not setInterval() |
| 01:42 | <roc> | auto-repeating is a reasonable alternative design |
| 01:42 | <roc> | but the level of complexity is about the same |
| 01:43 | <jamesr_> | with auto-repeating you'd definitely need a cancel |
| 01:43 | <roc> | and auto-repeating seems slightly more prone to runaway callback mistakes |
| 01:43 | <jamesr_> | roc: yeah, like all the goddamn setIntervals() on my gmail all the time |
| 01:43 | <roc> | cough cough |
| 01:43 | <jamesr_> | people just forget to clear them |
| 01:47 | <jamesr_> | i'll try to collect more opinions from webdevs around here as well |
| 03:02 | <MikeSmith> | http://thechangelog.com/post/2910383164/webdis-a-redis-http-interface-with-json-output |
| 07:55 | <hsivonen> | Hixie: I know it's against your principles, but it would be really helpful to use the Microsoft-specific syntax to turn of the IE8/9 XSS filter on the live DOM viewer |
| 08:44 | zcorpan | makes the script in html5-elements not work in older browsers |
| 08:48 | <Evet> | zcorpan: https://github.com/Modernizr/Modernizr/wiki/HTML5-Cross-browser-Polyfills |
| 08:50 | <zcorpan> | Evet: array.forEach isn't listed there. anyway, i don't care that it doesn't work in old browsers |
| 09:06 | <zcorpan> | i didn't know it was possible to do <a href=javascript:foo() target=bar> to make foo() execute in bar's context |
| 09:07 | <annevk> | does not seem possible in Chrome |
| 09:08 | <zcorpan> | doesn't seem to work in opera either |
| 09:09 | <zcorpan> | but works in ie8 |
| 09:09 | <hsivonen> | zcorpan: in IE7, too |
| 09:09 | <hsivonen> | didn't test IE6 |
| 09:10 | <zcorpan> | should we change to match ie6/webkit/opera and leave <base> alone? |
| 09:10 | <hsivonen> | actually, I tested <base target>, not <a target> |
| 09:10 | <zcorpan> | s/ie6// |
| 09:10 | <hsivonen> | zcorpan: we as in spec? |
| 09:10 | <zcorpan> | yeah |
| 09:10 | <zcorpan> | dunno what the spec says currently |
| 09:10 | <hsivonen> | zcorpan: afaict, Opera 11 does what Firefox 3.6 did |
| 09:11 | <zcorpan> | wrt <base target>? |
| 09:11 | <hsivonen> | zcorpan: spec says what Chrome and Firefox nightlies do |
| 09:11 | <hsivonen> | zcorpan: spec says honor first <base target> in the document order |
| 09:11 | <zcorpan> | yeah |
| 09:11 | <hsivonen> | doing the IE7 thing breaks hyperlatex output |
| 09:12 | <hsivonen> | and doing the IE7 thing for <base target> but the Chrome thing for <base href> would be inconsistent |
| 09:12 | <hsivonen> | and not doing the Chrome thing for <base href> breaks online check-in for non-U.S. citizens on united.com |
| 09:12 | <hsivonen> | which is likely to make people who are in a busy situation very unhappy |
| 09:12 | <annevk> | zcorpan, it does influence Opera |
| 09:13 | <annevk> | zcorpan, we do not always execute the script |
| 09:13 | <zcorpan> | annevk: oh. indeed |
| 09:14 | <hsivonen> | when iPad with it's locked browser gets traction in the enterprise, I hope the enterprisey vendors react by making a single cross-browser code path |
| 09:14 | <hsivonen> | and don't add a new one for WebKit |
| 09:14 | <zcorpan> | <base target=bar><a href=javascript:foo()> does nothing in opera |
| 09:15 | <zcorpan> | but <a href=foo()></a><base target=bar> executes (in the current window's context) |
| 09:15 | <hsivonen> | I bet enterprisey software would suck less for Firefox and IE developers, too, if the enterprisey software was written to work across Trident, Gecko, WebKit and Presto instead of being Trident/Gecko only |
| 09:17 | zcorpan | guesses there are sites that are broken in opera/webkit because javascript:/target is not supported |
| 09:18 | <othermaciej> | there is a tragic amount of WebKit-only content out there |
| 09:19 | <othermaciej> | enough that it is sometimes painful to fix bugs |
| 09:19 | <othermaciej> | the fact that WebKit is an API for other apps creates extra lock-in risk beyond just that of web sites |
| 09:21 | <hsivonen> | of late, Firefox 4 incompatibilities with Google and Microsoft services have been cases where stuff works if faking the UA to look like a WebKit UA :-( |
| 09:22 | <othermaciej> | we have a lot of compat bugs where faking a Firefox UA fixes it |
| 09:23 | <annevk> | Hixie, why is WebGL on the media-independent layer? |
| 09:23 | <annevk> | Hixie, in http://wiki.whatwg.org/wiki/Specifications_that_apply_to_Web_browsers |
| 09:24 | <hsivonen> | annevk: looks like an advocacy mind trick |
| 09:28 | <othermaciej> | does that page aim to at some point be comprehensive? |
| 09:28 | othermaciej | can't figure out the inclusion criteria |
| 09:29 | <annevk> | othermaciej, criteria are browser support |
| 09:30 | <annevk> | othermaciej, we are just adding things as we go so if you have anything please add |
| 09:30 | <othermaciej> | some of the more obvious things missing include DOM Events, XHR, other CSS3 modules (Selectors?) |
| 09:30 | <hsivonen> | looks like HTML is missing from the list |
| 09:30 | <hsivonen> | oh. it's there as WA 1.0 |
| 09:32 | <MikeSmith> | EcmaScript |
| 09:33 | <othermaciej> | I don't understand the categories enough to know where ECMAScript would go |
| 09:34 | <MikeSmith> | maybe it needs re-categorization |
| 09:34 | <annevk> | Infrastructure I think |
| 09:35 | <jgraham> | (FWIW we get both the bugs where you have to fake Firefox UA and the bugs where you hvae to fake the WebKit UA) |
| 09:35 | <annevk> | that or media-independent |
| 09:35 | <othermaciej> | we have had bugs where we needed to fake the UA string of an older Safari version |
| 09:36 | <othermaciej> | does URL/URI/IRI go in there, or are works of fiction excluded? |
| 09:38 | <hsivonen> | othermaciej: as I understand it, works of fiction are excluded |
| 09:39 | <annevk> | for that it seems better to wait for Adam Barth's draft |
| 09:39 | <annevk> | DOM Events might also get obsoleted by DOM Core |
| 09:46 | <jgraham> | Hmm, someone should tell Hallvord that WebIDL ought to define the behaviour of functions with the wrong number of arguments |
| 09:47 | <jgraham> | s/ought to define/defines/ perhaps |
| 10:41 | <annevk> | so I only read hybi casually but wtf |
| 10:41 | <annevk> | they are now at the point where browsers speak something different from anyone else? |
| 10:41 | <annevk> | Hixie argued that from the start |
| 10:41 | <annevk> | omg |
| 10:42 | <jgraham> | No, they are at the point where browsers and other things both speak "websockets" but browsers use masking but others don't |
| 10:42 | <jgraham> | At least that seems to be the point that Greg is at |
| 10:44 | <annevk> | same deal |
| 10:44 | <annevk> | imo |
| 10:44 | <jgraham> | Not at all |
| 10:44 | <jgraham> | Having other things not speak websockets is sane |
| 10:45 | <jgraham> | Piling features onto websockets to support non-browser use cases is not sane |
| 10:45 | <annevk> | fair enough |
| 10:46 | <jgraham> | But the browser people have more-or-less all retreated from the hybi group |
| 10:46 | <jgraham> | As far as I can tell |
| 10:46 | <annevk> | doing this at the IETF was in retrospect not a smart move |
| 10:47 | <annevk> | TLS-handshake and some simple framing and maybe some non-TLS handshake should have been sufficient |
| 10:50 | <annevk> | I wish Josh would provide some reasoning for his assertions on what "living standards" do to a11y |
| 10:50 | <annevk> | We almost never implement standards that are stable... |
| 10:51 | <Philip`> | Are the browser people still going to implement whatever the hybi group produces, or are they going to implement nothing, or are they going to restart their own spec development process, or does nobody have any idea? |
| 10:51 | <annevk> | In fact, there is not really any stable standard that I can think of... |
| 10:51 | <jgraham> | Philip`: I really have no idea |
| 10:52 | <jgraham> | Philip`: I guess the most likely outcome is that we will still implement for lack of an alternative, concrete, proposal |
| 10:52 | <jgraham> | Which is a bit worrying |
| 10:52 | <annevk> | Philip`, as far as I can tell Mozilla has not given up (in part because they hired someone more closely involved with the IETF); Google has not given up (they provide the editor, and Adam Barth is doing a bunch of stuff) |
| 10:52 | <jgraham> | Even abarth seems somewhat exhausted by the WG |
| 10:53 | <annevk> | Philip`, I guess we still have some involvement |
| 10:53 | <jgraham> | and he has an admirable capacity to navigate difficult standards groups |
| 10:53 | <annevk> | abarth should start the byebyeWG |
| 10:57 | <othermaciej> | it's draining to engage in the hybi WG |
| 10:57 | <othermaciej> | server people outnumber browser people, so they can create "consensus", but the costs of a poor protocol design and the costs of the constant thrashing will be disproportionately borne by browsers |
| 10:58 | <othermaciej> | also the chairs do not seem to understand the technical issues well enough to try to actually mediate instead of holding votes |
| 10:59 | <jgraham> | Yes |
| 11:00 | <jgraham> | That seems like a fair summary |
| 11:16 | <annevk> | btw |
| 11:16 | <annevk> | they have been doing this thing from the beginning |
| 11:17 | <annevk> | they want to share some kind of handshake because they think browsers will be able to get the intermediaries along (I thought browsers were outnumbered by all those other clients, but not here I guess...) but other than that they do not really care about browsers |
| 11:17 | <annevk> | or the security model of browsers, or anything really |
| 11:17 | <annevk> | but yeah, what othermaciej says |
| 11:17 | <annevk> | said |
| 11:18 | <othermaciej> | it's weird because I think at least some of the server people implement servers that will likely be connected to by browsers |
| 11:22 | <annevk> | othermaciej, Roy Fielding does that as well |
| 11:23 | <annevk> | othermaciej, or used to, yet I have not see him show any interest (other than mocking and getting angry) |
| 11:28 | <hsivonen> | annevk: not caring about browsers in the Web Socket context makes absolutely no sense |
| 11:28 | <hsivonen> | the whole point of the thing is to integrate with the browser security model |
| 11:28 | <hsivonen> | why on earth do people who don't have a browser as one end point care about Web Socket at all? |
| 11:29 | hsivonen | wishes the protocol had been done at the W3C |
| 11:29 | <annevk> | they think it makes sense; so maybe it does |
| 11:29 | <jgraham> | I'm not sure why they think it makes sense tbh |
| 11:29 | <annevk> | I know I don't care and do not want the browser part crippled |
| 11:30 | <jgraham> | But they keep getting all these requirements like "what if we want to speak websockets inside our datacenter so we can use the same protocol as we use to talk to clients" |
| 11:30 | <annevk> | what they want is bidirectional HTTP |
| 11:31 | <hsivonen> | putting Web Socket together with the bidirectional HTTP folks was such a mistake |
| 11:31 | <annevk> | to replace all kinds of proprietary protocols with hopefully better overall support |
| 11:31 | <jgraham> | And then they go "but to support that use case we don't want all the protection that clients need, so we should make all the protetcion optional" |
| 11:31 | <annevk> | it is quite different from what we want with Web Sockets |
| 11:31 | <annevk> | but they seem to have the upper hand |
| 11:31 | <annevk> | or more people shouting |
| 11:32 | <hsivonen> | jgraham: http://i938.photobucket.com/albums/ad222/powderny/picard-facepalm.jpg is still appropriate here :-( |
| 11:33 | <hsivonen> | on another topic, am I the only to whom the "People Are Making Stuff!" bit on the W3C HTML5 logo page has an astroturf flavor? |
| 11:34 | <othermaciej> | unfortunately it appears "rough consensus" just means "majority vote with some subjective and unstated supermajority requirement" |
| 11:34 | <foolip> | abarth, more nitpicking: the right border for text/xml in http://tools.ietf.org/html/draft-abarth-mime-sniff-06 is mis-aligned |
| 11:34 | <annevk> | I just noticed the logo page sets alt and title to the same thing... |
| 11:36 | <othermaciej> | http://www.w3.org/html/logo/ might be one of the least accessible pages on w3.org |
| 11:39 | <othermaciej> | the two most notable content images on the page are CSS background images |
| 11:42 | <hsivonen> | othermaciej: walking the walk is hard |
| 11:43 | <zcorpan> | othermaciej: while the decoration stars are <img>s, both with alt="HTML5 Navbar" |
| 11:43 | <zcorpan> | FAIL |
| 11:44 | <othermaciej> | well, the visual design is also atrocious, so I guess everyone can have a bad experience |
| 11:45 | <hsivonen> | I wish the designer had gone with the starfish idea |
| 11:46 | <othermaciej> | people would name the starfish |
| 11:47 | <jgraham> | starfish? |
| 11:47 | <hsivonen> | jgraham: as seen in the intermediade drawings the designer posted |
| 11:47 | <jgraham> | hsivonen: pointer? |
| 11:49 | <hsivonen> | jgraham: http://ocupop.com/html5 |
| 11:50 | <hsivonen> | (and I didn't mean the starfish variant that looks like a 5-pronged Confederate flag) |
| 11:50 | <foolip> | abarth, oh, and "An Ogg Vorbis audio or video signature." should be "An Ogg stream signature" or some such, since there's no guarantee that it's Vorbis or any other audio/video Ogg stream (could be Ogg Kate) |
| 11:51 | <annevk> | hmm, how do you find out signatures? |
| 11:51 | <hsivonen> | (nor the one that looks like Chrysler star) |
| 11:51 | <annevk> | I want to find them for fonts |
| 11:53 | <jgraham> | Oh the CSS thing was supposed to look like a 3! |
| 11:54 | <zcorpan> | also, read teh text in slide 9 |
| 11:55 | <zcorpan> | i dunno, but the logo doesn't come across as a keystone to me |
| 11:55 | <Philip`> | annevk: http://www.microsoft.com/typography/otspec/otff.htm (the "Offset Table"), if that's what you mean |
| 11:55 | <hsivonen> | zcorpan: the keystone idea has some merit, but the association with a keystone got lost in the shield |
| 11:56 | <hsivonen> | zcorpan: also, the keystone metaphor fails for i18n |
| 11:57 | <jgraham> | zcorpan: Clearly in the end it lost the keystone and went entirely with "shield" |
| 11:59 | <annevk> | I know have this |
| 11:59 | <annevk> | "We check for the first four bytes being 'ttcf' (TTC), 'OTTO' (OTF) or 0x00010000 (TTF)" |
| 11:59 | <annevk> | I don't have WOFF though |
| 12:01 | <annevk> | 0x774F4646 'wOFF' |
| 12:01 | <annevk> | yay |
| 12:03 | <Philip`> | I doubt 0x00010000 is uniquely TTF, but I don't see any more precise simple indicators :-( |
| 12:03 | <zcorpan> | oh, so that's why lasse refers to it as 'wOFF' |
| 12:03 | <karlcow> | http://wiki.eclipse.org/Orion |
| 12:03 | <karlcow> | >Orion is a browser-based open tool integration platform which is entirely focused on developing for the web, in the web. Tools are written in JavaScript and run in the browser. |
| 12:11 | <annevk> | quite nice that they all use four bytes |
| 12:11 | <annevk> | emailed websec |
| 12:13 | <hsivonen> | fwiw, regarding mp4, the ftyp part is the magic for the ISO container and the next four bytes are the subtype |
| 12:13 | <hsivonen> | JPEG 2000 has ftyp but the next four bytes differ from video/audio |
| 12:50 | <zcorpan> | Hixie: why don't you combine these into one url? "Specification: http://www.whatwg.org/specs/web-apps/current-work/complete/tokenization.html Section: http://www.whatwg.org/specs/web-apps/current-work/#scriptEndTag " |
| 13:41 | <annevk> | http://www.flickr.com/photos/danja/5386955591/ |
| 13:41 | <annevk> | <script> support sucked back then |
| 13:44 | <hsivonen> | annevk: wow. they must have had the Host HTTP header supported back then already |
| 13:45 | <karlcow> | annevk: this is honestly better than the current homepage, except for the grey background |
| 13:45 | <ben_c> | karlcow: current homepage needs rounded corners, then it's perfect |
| 13:46 | <hsivonen> | karlcow: don't you like the HTML5 CSS3 box-shadows? |
| 13:46 | <karlcow> | <cough repeat="3"/> ;) |
| 13:47 | <annevk> | karlcow, your sarcasm is very close to normal statements :) |
| 13:48 | <annevk> | I retweeted the screenshot on the WHATWG account |
| 13:48 | <karlcow> | the lovely box-shadow, the magnificent color scheme, the huge size of the footer ;) |
| 13:48 | zcorpan | tries border-radius:10em |
| 13:48 | <ben_c> | http://dl.dropbox.com/u/38899/ROUND.png |
| 13:48 | <zcorpan> | looks pretty! |
| 13:48 | <annevk> | very 2009 |
| 13:49 | <zcorpan> | 10em border-radius is better |
| 13:49 | <zcorpan> | except it becomes a bit girly |
| 13:49 | <karlcow> | hmmm I miss bgsound and birds on hr |
| 13:50 | <annevk> | that's like missing <marquee> and <blink> |
| 13:51 | <annevk> | it's exciting if you come across it these days, but you do not want to see it on every single site |
| 13:51 | <ben_c> | http://dl.dropbox.com/u/38899/wobbly.png ? |
| 13:52 | <jcranmer> | no, <blink> is still an annoying thing to see |
| 13:52 | <karlcow> | oh and maybe, http://www.fusion-productions.com/Images/under-construction-icon.jpg |
| 13:52 | <zcorpan> | ben_c: win |
| 13:52 | <jcranmer> | especially if the text being blinked is a long, important sentence |
| 13:52 | <ben_c> | I like under construction in a marquee |
| 13:52 | <karlcow> | holy cow, colored potatoes |
| 13:53 | <ben_c> | an alternating marquee is particularly exciting |
| 13:53 | <zcorpan> | you should make the potatoes rotate |
| 13:53 | <karlcow> | I knew it that whatwg was creating chemical food junk. not edible. |
| 13:53 | <jcranmer> | the potato shapes look slightly discomforting |
| 13:53 | <ben_c> | all jokes aside, what can we do to make it prettier? I'm happy to spend some time on it next week if anyone is up for it |
| 13:54 | <jcranmer> | some of the horizontal lines are still there, which is what is hurting my eyes, I think |
| 13:57 | <annevk> | ben_c, better |
| 13:58 | <karlcow> | ben_c the first thing to make it better right now, is to remove features :) Simplicity wins |
| 13:59 | <wilhelm> | Desaturation, please. (c: |
| 14:13 | <eighty4_> | trying to understand what html (5) element to use when wrapping a list of "related content" is the <aside> ok or should I just use a <section>? |
| 14:14 | <gsnedders> | eighty4_: aside |
| 14:14 | <eighty4_> | gsnedders: nice, I've been using that before but realize I might be over using it |
| 14:15 | <hsivonen> | please ship the ROUND.png version |
| 14:18 | <eighty4_> | gsnedders: and shouldn't you be at school studying right now? |
| 14:21 | <hsivonen> | isn't it so that the only attributes in SVG or MathML that take URLs are xlink:href and xml:base? |
| 14:22 | <hsivonen> | and plain href in MathML? |
| 14:28 | <gsnedders> | eighty4_: I could be in student union in the same room as people playing Magic :P |
| 14:33 | <david_carlisle> | hsivonen: definitionurl but the browser isn't expected to follow that so perhaps you don't count it |
| 14:34 | <hsivonen> | david_carlisle: ah, right. the only MathML attribute with capital letters in it. Thanks. |
| 14:34 | <david_carlisle> | sorry before my time (ie before 1999) |
| 14:36 | <david_carlisle> | mathml also has a few others (look for anyURI in the rnc) eg altimg (a fallback for if you don't do mathml) |
| 14:38 | <annevk> | I updated a few WHATWG wiki pages including the front page |
| 14:38 | <annevk> | could still be made clearer though |
| 14:39 | <hsivonen> | david_carlisle: thanks |
| 14:40 | <erlehmann> | gsnedders, ey, mtg is a nice game against computer scientists. it screams of OH EXPLOITABLE |
| 14:41 | <annevk> | hmm, the rel="up up up" thing is about to expire |
| 14:41 | <gsnedders> | erlehmann: Meh, I don't play it (or rather haven't in years, and then only stole friend's cards), and off the top of my head none of the people here except for me do CS |
| 14:43 | <erlehmann> | gsnedders, i played it when i was 13? 14? and recently i got some of my friends to remember it and pull out their cards :) |
| 14:44 | <gsnedders> | erlehmann: heh, I had a few friends who played it at school, especially people I was around in my final year (when I was 17, most of them 18), and have now ended up at uni around people who play it incesently |
| 14:44 | <erlehmann> | i noticed the 3d shadow effect on whatwg.org |
| 14:44 | <erlehmann> | can i haz anaglyphs? ;) http://blog.fefe.de/?css=http://daten.dieweltistgarnichtso.net/src/fefe-anaglyph-css/anaglyph.css |
| 14:49 | <annevk> | I have now written a second Change Proposal |
| 14:50 | <hsivonen> | argh. I was planning on writing one today but forgot. |
| 14:50 | <hsivonen> | 5 minutes to telecon |
| 14:53 | <annevk> | I never got a reply to my ISSUE-126 I noticed |
| 14:54 | <annevk> | email /\ |
| 14:54 | <annevk> | so I wrote a no change proposal instead |
| 14:54 | <annevk> | I hope I did it correctly... |
| 15:08 | <kennyluck> | hsivonen, I actually have written quite a bit for conforming-u -> http://www.w3.org/html/wg/wiki/ChangeProposals/UShouldBeConforming . I am not sure whether I am making sense, though. |
| 15:09 | <hsivonen> | kennyluck: cool. I'm on a telecon right now and can't review |
| 15:09 | <kennyluck> | me too, actually. |
| 15:09 | <MikeSmith> | kennyluck is on a telcon too actually :) |
| 15:32 | <karlcow> | http://humanstxt.org/ |
| 15:34 | <zcorpan> | "I'm not a robot" |
| 15:36 | <annevk> | so how much does "tree model" explain? |
| 15:36 | <annevk> | I guess not much and I should actually define ancestor node, parent node, child node, descendant node, etc. |
| 15:36 | <Ms2ger> | Yes please :) |
| 15:37 | <annevk> | you know |
| 15:37 | <annevk> | to me when someone says "tree" everything follows from that :) |
| 15:37 | <Ms2ger> | "See Wikipedia"? |
| 15:37 | <annevk> | the Web Platform should have a definition for this |
| 15:37 | <annevk> | it comes back all over the place |
| 15:38 | <annevk> | and you just say this implements a "tree model" |
| 15:39 | <annevk> | and readers would go like |
| 15:39 | <annevk> | ooh right |
| 15:39 | <annevk> | and everyone would be enlightened |
| 15:39 | <Ms2ger> | I approve of enlightenment |
| 15:41 | <Philip`> | Is it a problem of teaching people the fundamental concepts of trees, or just of defining precisely the terminology you use? |
| 15:41 | <annevk> | can it be a combination? |
| 15:41 | <Philip`> | Maybe |
| 15:42 | <annevk> | my way out was starting with defining Events :) |
| 15:42 | <Ms2ger> | "A tree is a perennial woody plant." |
| 15:42 | <Ms2ger> | Or wouldn't that work as a definition? |
| 15:43 | <annevk> | The DOM is a tree of nodes. In the sense of a family tree, you get the idea. |
| 15:43 | <annevk> | Would work for most people :) |
| 15:43 | <Philip`> | For the fundamental concepts it seems best to point people at an introductory CS text that describes it in proper detail, rather than trying to make up a new detailed description |
| 15:44 | <Philip`> | and hopefully that'll use the same terminology as web specs do so you won't need to redefine anything |
| 15:48 | <jgraham> | annevk: Family trees are more like DAGs though |
| 15:51 | <Philip`> | Family trees usually have edges between married couples so they're not really D or A |
| 15:52 | <jcranmer> | probably a better way of putting it is that they're not really `G' |
| 15:53 | <jcranmer> | although I wonder how you encode children born via extramarital affairs or out of wedlock |
| 16:00 | <annevk> | see |
| 16:00 | <annevk> | because some people think about this too hard I need to provide airtight silly definitions |
| 16:02 | <Philip`> | That's why you need to point to a resource written by people who've already thought about it too hard :-) |
| 16:03 | Philip` | doesn't have any specific suggestions though, since the only resource he can think of is on paper |
| 16:03 | <annevk> | I think I'll end up borrowing from either CSS or the HTML bit on browsing contexts |
| 16:11 | jgraham | was just considering each child to have two parents and not encoding the relationship between the parents, fwiw |
| 16:14 | <Philip`> | So you're assuming the human population increases exponentially as you go back through the generations? |
| 16:14 | <Philip`> | Oh, or do you just mean not encoding the co-parenting relationship but still encoding shared ancestors of those parents? |
| 16:16 | <jgraham> | The latter |
| 16:27 | <Lachy> | Philip`, the population would only exponentionally increase as you go back through generations if each couple only had 1 child, which is typically not the case |
| 16:28 | <Lachy> | well, based on the average being greater than 2 in most places around the world |
| 16:28 | <Philip`> | I was assuming jgraham was referring to a family tree rooted at a single child and only showing ancestors |
| 16:28 | <Lachy> | ok |
| 16:29 | <Philip`> | because otherwise there's no chance it'll ever look anything remotely like a tree |
| 16:30 | <Ms2ger> | "IMO browsers should implement <link>." |
| 16:31 | <annevk> | first <zelda> |
| 16:32 | <Ms2ger> | <wario>! |
| 16:33 | <erlehmann> | <wario>HRRRRGGGRRRR</wario> |
| 16:33 | <annevk> | what is the link bewteen Zelda and Wario? |
| 16:33 | <annevk> | aah |
| 16:33 | <annevk> | super smash brothers! |
| 16:33 | <erlehmann> | The Nintendo Markup Language matures. |
| 16:33 | <Ms2ger> | I want Wario Markup Language - WML |
| 16:33 | <annevk> | <snake> |
| 16:33 | <f1lt3r_bocoup> | lol |
| 16:34 | <Ms2ger> | <badger> |
| 16:34 | <wilhelm> | WML certainly looks like something Wario could have made. |
| 16:34 | <annevk> | Ms2ger, I guess I should have used a namespace, I meant Solid Snake! |
| 16:34 | <jgraham> | <kirby> |
| 16:34 | <kennyluck> | <kuppa> |
| 16:34 | <erlehmann> | <mushroom> |
| 16:34 | <erlehmann> | <badger> <badger> <mushroom> |
| 16:34 | <Ms2ger> | Oh |
| 16:34 | <kennyluck> | s/<mushroom>/<toad> or <kinopio>/ |
| 16:35 | <erlehmann> | that looks more like an obscure SGML feature than a regular expression |
| 16:35 | <Ms2ger> | <snake xmlns="http://www.albinoblacksheep.com"; /> |
| 16:35 | <annevk> | clearly I am not getting this nodes model down |
| 16:44 | <foolip> | http://esw.w3.org/HTML_XML_Use_Case_07 is weird |
| 16:45 | <foolip> | 'The user aspires to create well-formed XHTML5 ... would be disuaded from this if normal "hard fail on wellformedness-error" XML processing were used' |
| 16:45 | <foolip> | sounds like the user doesn't want to create well-formed XHTML5 |
| 16:46 | <Ms2ger> | He wants to, but he's a bozo |
| 16:49 | <annevk> | look at what you did to twitter foolip |
| 16:49 | <annevk> | http://twitter.com/mattur/status/29944802837528578 |
| 16:49 | <annevk> | I wonder how more levels we can go |
| 16:50 | <foolip> | annevk, I should have made that tweet 140 characters long to prevent that :) |
| 16:50 | <foolip> | is everyone else using an awesome client that hides the dupes? |
| 16:52 | <Lachy> | I don't see the problem with dupes in twitter. I read twitter infrequently enough that anything I see is more than likely going to be new for me anyway |
| 16:53 | <annevk> | No wonder then... |
| 16:54 | <karlcow> | foolip: I don't understand the manual retweet… you mean that for retweeting I would have to go to the Web UI of twitter? |
| 16:54 | <Lachy> | karlcow, or use a client that supports retweet |
| 16:55 | <karlcow> | Are there special meta which are sent along the text for retweet? |
| 16:56 | <annevk> | yeah, the ID of the tweet you are retweeting (I think) |
| 16:56 | <Lachy> | a manual retweet by convention just begins with "RT @username...". Otherwise, the automated ones just use something in the twitter API with an ID to retweet it as-is |
| 17:00 | <annevk> | Ms2ger, time for some brainstorming? |
| 17:00 | <annevk> | if you do |
| 17:01 | <annevk> | DOM is a tree of Node objects (simple called <dfn title=concept-node>nodes</dfn>) |
| 17:02 | <annevk> | each <span ...>node</> has exactly one <dfn title=concept-parent-node>parent node</dfn> with exception of the <code>Document</code> object |
| 17:02 | <annevk> | i.e. basing it on http://www.w3.org/TR/CSS21/conform.html#doctree |
| 17:03 | <annevk> | and introducing "node" as a term so we don't need <code> for it all the time |
| 17:03 | <annevk> | maybe it does not need to be a <dfn> term as it is used so commonly |
| 17:03 | <annevk> | or a referenced <dfn> term that is |
| 17:04 | <annevk> | if anyone else has comments feel free |
| 17:04 | <gsnedders> | annevk: a parent node can be null, that needs to be there, no? |
| 17:04 | <annevk> | oh right, there's not always a tree |
| 17:04 | <annevk> | hmm |
| 17:05 | <annevk> | okay so yeah parent can be null |
| 17:09 | <annevk> | bah |
| 17:09 | <david_carlisle> | annevk: XDM isn't DOM but they had to face the same issues of parentless element nodes etc, you might (or might not) be able to steal ideas from |
| 17:09 | <david_carlisle> | http://www.w3.org/TR/xpath-datamodel/#terminology |
| 17:11 | <MikeSmith_> | http://twitter.com/#!/tmpvar/status/29762642008612864 "WANTED: xml/namespace/dtd guru to answer noob questions and explain how the w3c level2/core tests are not contradictory." |
| 17:12 | <annevk> | http://www.w3.org/TR/xpath-datamodel/#DocumentNode seems similar to our constraints |
| 17:12 | <Philip`> | Are they not contradictory? |
| 17:12 | <annevk> | although it is a bit confusingly worded |
| 17:12 | Philip` | wouldn't be particularly confident in that |
| 17:16 | <MikeSmith> | seems he's running into the typical fun with namespaces that most implementors encounter |
| 17:17 | <MikeSmith> | he clearly just doesn't realize how simple it all is to implement |
| 17:17 | <MikeSmith> | no trouble at all |
| 17:25 | <MikeSmith> | anyway, tmpvar is the force behind jsdom |
| 17:53 | <annevk> | writing this text is boring |
| 17:54 | <annevk> | I should have done the event model |
| 17:54 | <othermaciej> | what text are you writing? |
| 17:54 | <annevk> | I also have the feeling it's not going well so the distracting email is nice |
| 17:54 | <annevk> | explaining how the tree nodes form works |
| 17:55 | <annevk> | i.e. that nodes have a parent node, that Document does not, that there's such a thing as child nodes |
| 17:55 | <annevk> | ancestor nodes, blahdieblah |
| 17:58 | <TabAtkins_> | gsnedders: Let us know when you want to call in. |
| 19:06 | <annevk> | maybe http://www.456bereastreet.com/archive/201101/html_beyond_html5/ is worth spending another post on |
| 19:06 | <annevk> | if someone has the time |
| 19:06 | <annevk> | well, makes the time |
| 19:06 | <annevk> | it quite clearly illustrates some of the problems people are seeing |
| 19:07 | <annevk> | of course it does not admit that using "HTML5" you will have all these problems because it has features that are not ready yet, but... |
| 19:07 | <annevk> | that could be something for the post to point out |
| 19:11 | <annevk> | Ms2ger, put some WIP in src.html |
| 19:11 | <Ms2ger> | Will check |
| 19:12 | <annevk> | might be back in an hour or so, but have to see |
| 19:15 | <_bga> | omg http://www.thespanner.co.uk/2009/01/29/detecting-browsers-javascript-hacks/ |
| 19:16 | <Ms2ger> | http://dvcs.w3.org/hg/webevents/raw-file/tip/touchevents.html < Now that's a pretty old draft |
| 19:19 | <annevk> | _bga, hmm yeah, I hope browsers break those expectations soon |
| 19:19 | <annevk> | _bga, otherwise that is gonna suck |
| 19:19 | <paul_irish> | _bga: i think ie9 broke the \v hack. |
| 19:20 | <paul_irish> | all of them are pretty damn brittle |
| 19:20 | <annevk> | oh, I guess I might be online from the train |
| 19:21 | <_bga> | btw im member of "feature detection" religion :) |
| 19:21 | <hsivonen> | those JS hacks are a terrible idea to put in production |
| 19:21 | <paul_irish> | _bga: i know. :) me too. though you might have seen my Undetectables page.. basically gives some reasoning why we can't go without UA sniffing. |
| 19:51 | <Ms2ger> | Anne, I pushed a few small changes, looks good to me otherwise |
| 20:01 | <Ms2ger> | If you haven't checked the logs yet, <Ms2ger> Anne, I pushed a few small changes, looks good to me otherwise |
| 20:01 | <annevk> | everything went better than expected |
| 20:10 | <annevk> | haha |
| 20:10 | <annevk> | for a moment The Official Google Blog said: "This is the body of the blog post. I can include HTML tags." |
| 20:15 | <annevk> | pushed to W3C |
| 20:15 | <annevk> | have to go |
| 20:18 | <Hixie> | moved webgl |
| 20:24 | <MikeSmith> | Hixie: moved? |
| 20:30 | <annevk> | not bad |
| 20:31 | <annevk> | fifteen minutes from train to home |
| 20:31 | <annevk> | including some relaxing |
| 20:31 | <annevk> | MikeSmith, on the Specs page |
| 20:32 | <annevk> | http://wiki.whatwg.org/Specs |
| 20:32 | <MikeSmith> | ah |
| 20:32 | <annevk> | make that http://wiki.whatwg.org/wiki/Specs |
| 20:32 | <annevk> | I guess we should point to ArrayBuffer too |
| 20:32 | <annevk> | and I should update the XHR spec with that... |
| 20:34 | <MikeSmith> | the Ecmascript link could go to http://es5.github.com/ |
| 20:35 | <annevk> | whoa that page makes my browser crawl |
| 20:35 | <MikeSmith> | annevk: yeah |
| 20:35 | <Hixie> | i sent http://damowmow.com/temp/456bereastreet.com-archive-201101-html_beyond_html5-response.txt to roger using his web form |
| 20:36 | <Hixie> | if anyone wants to take bits of that to update the faq, be my guest. |
| 20:36 | <MikeSmith> | is is not ironic that "Web Applications 1.0" is the the only spec on the spec page that has a version number? |
| 20:37 | <annevk> | it is |
| 20:38 | <annevk> | but removing the 1.0 makes it too vague :) |
| 20:38 | <annevk> | or maybe it is just a vague title :) |
| 20:39 | <MikeSmith> | maybe we need a "1.0" logo |
| 20:40 | <nimbupani> | \|/ |
| 20:44 | <annevk> | if I'm allowed to bikeshed a little |
| 20:44 | <annevk> | we could call it WHATWG C |
| 20:44 | <annevk> | WHATWG HTML and WHATWG C |
| 20:44 | <annevk> | short: C |
| 20:45 | <annevk> | Hixie, pretty good response |
| 20:45 | <Hixie> | the 1.0 is for nostalgic reasons |
| 20:46 | <karlcow> | just call it "woot!" :) |
| 20:46 | <MikeSmith> | let's do six blades and just change it to a symbol instead of words |
| 20:46 | <MikeSmith> | like Prince |
| 20:48 | <annevk> | Hixie, maybe your response should be put on the blog? |
| 20:49 | <Hixie> | nah |
| 20:49 | <Hixie> | it just prolongs the story if we put it on the blog |
| 20:49 | <annevk> | yeah true, there's more interesting things happening |
| 20:50 | <Hixie> | btw, that screenshot is awesome |
| 20:50 | <Hixie> | backwards compatibility "ftw", as they say |
| 20:50 | <Hixie> | gotta say, i kinda like our new home page |
| 20:51 | <Steve^> | your name is in very, very small text |
| 20:51 | <Hixie> | i considered removing that line altogether |
| 20:51 | <Hixie> | but instead i put analytics on it and we'll see if anyone's using it |
| 20:53 | <Steve^> | Google Analytics? |
| 20:53 | <Hixie> | yeah |
| 20:53 | <Steve^> | So the click on your email, you get that information with an event? |
| 20:53 | <Steve^> | I've had very little success with GA and off-site linking |
| 20:54 | <Hixie> | annevk: thanks for the wiki cleanup, looks good |
| 20:54 | <Hixie> | Steve^: oh i just meant the page as a whole, and if the page is getting hit a lot, i'll try playing with the link |
| 20:54 | <Steve^> | ah, ok |
| 20:54 | <Hixie> | maybe i should change the e-mail address and see if i get mails to it |
| 20:54 | <Hixie> | other than spam |
| 20:55 | <Steve^> | in theory you can tag it and GA will let you know, but there were no docs last time I checked |
| 20:55 | <Hixie> | ah |
| 20:55 | <Steve^> | they are transitioning to asynchronous and the docs aren't.. in sync |
| 20:56 | <Hixie> | ok i changed the e-mail address, we'll see if i get any organic traffic from it |
| 20:56 | <Hixie> | lunch time |
| 20:57 | <annevk> | maybe the WHATWG can do some deals with Nigeria |
| 20:57 | <annevk> | lots of cash there |
| 20:57 | <annevk> | or Niger |
| 20:57 | <annevk> | then I can finally get an.ne o_O |
| 20:57 | <Steve^> | :D |
| 20:58 | <Steve^> | Mix with the Princes of Nigeria and they'll accuse WHATWG of having government influences |
| 20:59 | <Steve^> | Come to think of it, it's been so long since i had any spam, I've forgotten how the good ones go |
| 20:59 | <webr3> | omg have you seen the registration services site for .ne ? http://www.intnet.ne/ |
| 20:59 | <Steve^> | Here we go, Chairman of the Electoral Commission Ghana |
| 21:00 | <annevk> | Steve^, heh, we can prolly ask Shell for tips |
| 21:01 | <Steve^> | webr3, why is the Nigerian telecoms company based in Niger? |
| 21:02 | <Steve^> | Pardon me, dodgy translation by Google |
| 21:13 | <annevk> | Hixie, btw the problem at the bottom is that Mosaic renders the contents of <script> |
| 21:13 | <annevk> | Hixie, so you would need to use <!-- and --> in there if you really want to hide that |
| 21:14 | <annevk> | Hixie, on whatwg.org |
| 21:14 | <annevk> | Hixie, also, ga.type = 'text/javascript'; is not needed |
| 21:17 | Dashiva | is confused by the HTML XML use cases |
| 21:17 | <Dashiva> | How can "Using an XML toolchain to consume HTML" and "Forgiving error-handling with XHTML-style markup" ever coexist? |
| 21:18 | <annevk> | By improving the XML toolchain? |
| 21:19 | <annevk> | But yes, not everything is possible. |
| 21:19 | <Dashiva> | It seems to me that any XML toolchain that's capable of consuming non-XML probably doesn't need HTML XML |
| 21:21 | <annevk> | https://labs.ericsson.com/developer-community/blog/beyond-html5-peer-peer-conversational-video |
| 21:23 | <MikeSmith> | ConnectionPeer |
| 21:23 | <annevk> | it is happening |
| 21:24 | <annevk> | time to go |
| 21:37 | <karlcow> | http://stackoverflow.com/questions/4798291/applying-css-transform-rotation-to-an-element-using-font-face |
| 21:39 | <Rik`> | karlcow: I've seen that without font-face on Firefox and Opera |
| 21:39 | <roc> | It's graphics API limitations mostly |
| 21:41 | <Rik`> | I thought I was CCed on one bug like that, but I can't find it… |
| 21:43 | <roc> | seems to be basically fixed with FF4 on Windows 7 |
| 21:43 | <roc> | using Direct2D |
| 21:44 | <roc> | I think screen pixellation makes it impossible to look 100% right |
| 21:45 | <Rik`> | http://software.hixie.ch/utilities/js/live-dom-viewer/saved/798 |
| 21:45 | <Rik`> | this looks ugly on Mac OS |
| 21:46 | <roc> | yeah, it's a Quartz issue |
| 21:46 | <roc> | there's some secret API that's supposed to fix it, but we couldn't get it to work or something |
| 21:46 | <roc> | I forget the details |
| 21:47 | <Rik`> | Safari is ok with this |
| 21:47 | <roc> | yes, they got the secret API to work |
| 21:49 | <Rik`> | https://bugzilla.mozilla.org/show_bug.cgi?id=492214 was the bug I was thinking about |
| 21:50 | <karlcow> | it works well with opera too. it is interesting how resizing the window makes the letter in firefox "dancing" |
| 21:50 | <Rik`> | karlcow: I tried in Opera, not looking good at all |
| 21:51 | <Rik`> | the letters are fine but the anti-aliasing is bad |
| 21:55 | <karlcow> | http://www.la-grange.net/2011/01/25/rotate-tests this is what I see |
| 21:55 | <karlcow> | ooops mistake |
| 21:55 | <karlcow> | :) |
| 21:55 | karlcow | modifies |
| 21:56 | <Rik`> | two Firefox screenshots |
| 22:02 | <karlcow> | http://www.la-grange.net/2011/01/25/rotate-tests |
| 22:02 | <karlcow> | updated |
| 22:04 | <Rik`> | so, what I said |
| 22:05 | <Rik`> | Safari is ok, Opera is doing a bad antialiasing and Firefox dances with the letters |
| 22:06 | <karlcow> | party time! |
| 23:16 | <Figaroo> | Hello folks, I have some interesting input which you might or might not have received already. |
| 23:19 | <Figaroo> | Limiting character input on text fields is a bit cumbersome to do with javascript, currently. |
| 23:22 | <Figaroo> | Firstly, keypress events do not account for pasting in input fields and the change event only fires when the text field is blurred. A propery ontextchange event would be useful here. |
| 23:25 | <Rik`> | Figaroo: there is an oninput event that does that |
| 23:25 | <Figaroo> | the event object passed to the ontextchange field should also contain a property which reveals the text change -- event.oldValue and event.newValue or similar. This would be extremely useful to determining what text is entered from keystrokes and various text based commands such as pasting. |
| 23:25 | <Rik`> | and to limit character input, you have the maxlength and pattern attributes |
| 23:25 | <Figaroo> | Rik`, wonderful! Does it contain the properties I just mentioned? |
| 23:25 | <Rik`> | nope |
| 23:26 | <Figaroo> | Rik`, that's not what I mean by limit character input. I mean limit which characters to be inputted. |
| 23:26 | <Rik`> | but you just need to save the value to get an "old value" |
| 23:27 | <Rik`> | pattern is a javascript regexp, so it won't block people from entering invalid characters, but it will report the input as invalid |
| 23:27 | <Figaroo> | Rik`, save the value? How would you know to save it? |
| 23:28 | <Rik`> | just save it every time there is an input ? |
| 23:28 | <Figaroo> | When there is an input, wouldn't the value would have already been changed? |
| 23:29 | <Figaroo> | *x out one would* |
| 23:29 | <Rik`> | yes but since you saved the value the last time the callback was called, you'll have value and e.target.value |
| 23:31 | <Figaroo> | Rik`, wouldn't it be more convenient to have this value available in the oninput event? |
| 23:31 | <Figaroo> | Also, limiting text input is not the same as limiting valid input. |
| 23:31 | <Rik`> | that's one line of code really |
| 23:32 | <Figaroo> | Rik`, what is? |
| 23:34 | <Figaroo> | What is one line of code? |
| 23:37 | <Figaroo> | I see, so storing the old value is one line of code extra; maybe two because you'll need to define the variable which contains the old value. But I think it would be nice to have this built into the event, I don't see why not. |
| 23:38 | <bga_> | Figaroo https://github.com/dperini/nwxforms |
| 23:41 | <twisted`> | hi maybe someone here can point the me in the right direction... I need to re-polish my web-knowledge (last was xhtml 1.1 strict) but with all the html5 fad going on lately (well ok I've been in hibernation for a while ;)) I'm not sure what exactly is now which spec is most widely supported (with new sweet features) |
| 23:41 | <bga_> | https://github.com/dperini/nwxforms/blob/master/src/nwxforms.js#L585 |
| 23:41 | <inimino> | Figaroo ⋱ What was your issue with <input type=number>? |
| 23:43 | <paul_irish> | twisted`: read diveintohtml5 and watch developers.whatwg.org |
| 23:44 | <twisted`> | diveintohtml5.org ? |
| 23:44 | <paul_irish> | yup |
| 23:44 | <inimino> | twisted` ⋱ The whatwg FAQ page is also a good read. |
| 23:45 | <inimino> | http://wiki.whatwg.org/wiki/FAQ |
| 23:45 | <TabAtkins_> | inimino: What character are you using after people's names? It looks like a diagonal ellipsis on my terminal. |
| 23:45 | <twisted`> | thanks I'll check those docs out |
| 23:46 | <twisted`> | I read a few years ago about xhtml2 which looked very interesting but it seems it's deprecated |
| 23:46 | <twisted`> | important for me is that the sites I write still function in stuff like lynx/links2 but are (without too much js) very nice in modern browsers |
| 23:52 | <franksalim> | TabAtkins_, DOWN RIGHT DIAGONAL ELLIPSIS (U+22F1) |
| 23:53 | <twisted`> | doesn't show up here like the image I can find of "DOWN RIGHT DIAGONAL ELLIPSIS" |
| 23:53 | <twisted`> | hmm weird |
| 23:53 | <twisted`> | terminal.app should be fully unicode |
| 23:53 | <twisted`> | maybe it's my font ;) |
