| 03:24 | <rektide> | i think i've gotten by my Shared Worker impass, thanks to zcorpan |
| 05:18 | <MikeSmith> | seems like the only intuitive way to implement "mouse over" equivalent on touchscreen devices is to also have a proximity sensor on the device that can detect when you are actually hovering over something with your finger |
| 05:21 | <MikeSmith> | there is a big irony in the fact that the prevalent GUI conventions are modeled on physical desktop conventions that can't currently be implemented practically on devices that provide a greater degree of real physical interaction with the UI (through the touchscreen) rather than through the abstraction of a separate pointing device |
| 05:23 | MikeSmith | , with those words of wisdom, finishes off his second breakfast beer, while trying to avoid spilling it on his laptop keyboard |
| 05:24 | <wirepair> | breakfast? |
| 05:25 | wirepair | looks at watch |
| 05:25 | <boblet> | late breakfast methinks |
| 05:25 | <wirepair> | indeed |
| 05:26 | <MikeSmith> | wirepair: actually, late lunch |
| 05:26 | <wirepair> | *nod* <-- in japan as well |
| 05:26 | <MikeSmith> | I didn't have time for my breakfast beers today, so I'm playing catch-up |
| 05:26 | <wirepair> | heh |
| 05:27 | <MikeSmith> | wirepair: where in Japan? |
| 05:27 | <wirepair> | oota-ku |
| 05:27 | <MikeSmith> | ah cool |
| 05:27 | <wirepair> | yourself? |
| 05:28 | <MikeSmith> | i'm normally in Shinjuku or nearby, or at the W3C office at Keio U. near Fujisawa, or out and about elsewhere |
| 05:28 | <MikeSmith> | at the moment I'm actually in Mitaka |
| 05:29 | <MikeSmith> | at an NTT tech expo that's going on today |
| 05:29 | <wirepair> | ah cool, didn't realize w3c had an office here heh |
| 05:29 | <MikeSmith> | yeah, smallish office |
| 05:29 | <MikeSmith> | in terms of people, currently |
| 05:29 | <wirepair> | i used to be at symc, did some browser security assessments and been hooked on browsers ever since, so decided to hang out here see what cooky stuff you guys are cooking up ;> |
| 05:29 | <wirepair> | er symc = symantec |
| 05:30 | <MikeSmith> | excellent |
| 05:30 | <MikeSmith> | I would guess you might have some good insight into client-side security issues |
| 05:30 | <MikeSmith> | which the community could definitely use more of |
| 05:30 | <wirepair> | yeah i've been passively reading ;> |
| 05:30 | <wirepair> | mainly seeing in fighting though ;> |
| 05:32 | <MikeSmith> | wirepair: you may want to read up on iframe/@srcdoc and get involved in recent discussion of that |
| 05:32 | <wirepair> | got a link for that? |
| 05:33 | <MikeSmith> | http://dev.w3.org/html5/spec/text-level-semantics.html#attr-iframe-srcdoc |
| 05:33 | <wirepair> | thanks |
| 05:33 | <MikeSmith> | and see related discussion on that whatwg mailing list |
| 05:33 | MikeSmith | will try to get a link for that too |
| 05:33 | <wirepair> | appreciate it |
| 05:34 | <MikeSmith> | I don't know if you aware already but another thing that's been in the works for some time now is a mechanism for enabling cross-origin (cross-site) XHR (and other forms of cross-origin requests) |
| 05:34 | <MikeSmith> | that is definitely worth taking a loot at too |
| 05:34 | <MikeSmith> | Cross Origin Resource Sharing |
| 05:34 | <wirepair> | yeah, i've read a little about it |
| 05:35 | <wirepair> | couldn't see anything immediately wrong with the preflights, but i think it may boil down to browser implementation |
| 05:37 | <MikeSmith> | wirepair: you've read up on the alternative Uniform Messaging proposal as well? |
| 05:37 | <wirepair> | hmm, no i have not i've honestly been busy building my testing framework heh |
| 05:39 | <MikeSmith> | wirepair: about recent iframe discussions, http://lists.whatwg.org/pipermail/whatwg-whatwg.org/2009-December/thread.html#24487 is a good place to start if/when you have time |
| 05:40 | <wirepair> | yeah i should be testing this web app ehe, i'll check it after work |
| 05:40 | <MikeSmith> | はい |
| 05:40 | <wirepair> | blarg, stupid utf-7 term |
| 06:05 | <MikeSmith> | othermaciej: in your response to Larry about normative reference to IRIbis, the URL you gave is a file URL for your local copy of the issue-status file |
| 06:05 | <othermaciej> | MikeSmith: oops |
| 06:06 | <MikeSmith> | no biggie |
| 06:06 | <MikeSmith> | the dev.w3.org version is linked to in plenty of other messages, so I'm sure people can put 2 and 2 together |
| 06:10 | <MikeSmith> | I notice that nobody seems to have responded yet to Aryeh and Henri's message questioning the utility of warning about absence of lang |
| 06:11 | <MikeSmith> | http://lists.w3.org/Archives/Public/public-html/2010Feb/0160.html and http://lists.w3.org/Archives/Public/public-html/2010Feb/0171.html |
| 06:11 | <MikeSmith> | I guess I should try to see if I can get someone to respond to those points |
| 06:16 | <JonathanNeal> | Hey all, how's everyone tonight (or whatever time it is where you exist) ? |
| 06:16 | <othermaciej> | MikeSmith: does the spec currently require that? |
| 06:18 | <MikeSmith> | othermaciej: I was thinking maybe the change proposal was advocating for having it be a warning, but maybe in fact it doesn't and that suggestion only came up on the e-mail thread |
| 06:18 | MikeSmith | goes to looks now |
| 06:19 | <othermaciej> | my brain hurts from sending all that email |
| 06:19 | <othermaciej> | and I still need to write an agenda |
| 06:26 | <MikeSmith> | I need to write an agenda for the accessibility TF call for this week too |
| 06:26 | <MikeSmith> | which is less work |
| 06:27 | <MikeSmith> | but I believe we will have two actual proposals from the TF to move forward to the wider group of the a11y TF telcon this week |
| 06:27 | <MikeSmith> | the canvas adom proposal and the summary proposal |
| 06:28 | <MikeSmith> | *move forward to the wider group for further discussion |
| 06:28 | <MikeSmith> | which reminds me also that I wanted to check in with Silvia |
| 06:28 | <MikeSmith> | nessy: you around and have time to chat a bit about status on the media proposals? |
| 06:35 | <MikeSmith> | OK, I see that Richard's change proposal for issue 88 in fact says nothing about warning on absence of lang |
| 06:35 | <MikeSmith> | so nm |
| 06:41 | <MikeSmith> | http://dvcs.w3.org/hg |
| 06:45 | <hsivonen> | "range of expectation" is pretty narrow as far as desktop browsers go... |
| 07:16 | <othermaciej> | ok, why is it that the various Calls for Whatever I posted today drew so many random follow-up comments? |
| 07:17 | <othermaciej> | most of the things of this type that the Chairs post get no responses at all |
| 07:22 | <othermaciej> | hsivonen: does .@ mean something special in twitter-ese (as opposed to just @) |
| 07:29 | <Philip`> | "So if someone really wanted to propose allowing the rel attribute on <img> elements, that would be in scope (though in my opinion, it would be a silly idea)." - RDFa already makes that allowed |
| 07:29 | <Philip`> | (if I'm not misreading the DTD) |
| 07:39 | <hsivonen> | othermaciej: It means I wanted to broadcast those tweets to all my followers--not just those who also follow masinter |
| 07:39 | <othermaciej> | I did not know twitter made such a distinction |
| 07:40 | othermaciej | has learned something |
| 08:11 | <hsivonen> | Is it just me or does "The requirement is from TimBL's original axioms." come close to the axiomatic proof? |
| 08:20 | <othermaciej> | ok, someone tell me that I shouldn't reply to Roy's latest email |
| 08:20 | <othermaciej> | because, if we're gonna drop ping anyway, there is no point debating it |
| 08:24 | <Philip`> | I don't really understand why anyone would like the idea of simply removing the W3C logo from the feature, when it's equally likely to be implemented if it's in a WHATWG spec |
| 08:24 | <Philip`> | Isn't the idea meant to be to improve the feature? |
| 08:25 | <Dashiva> | Not if you think it's broken by design, I suppose |
| 08:27 | <Philip`> | If you think that, I still don't see why you'd happy to just have it move to a different spec where it'll get implemented just the same and you won't have any chance to provide feedback to make it slightly less broken |
| 08:31 | <Dashiva> | Philip`: You make a clear statement that you're against it, and wash your hands of the consequences |
| 08:31 | <hsivonen> | Philip`: having it only in the WHATWG spec takes it away from the W3C PP, which might have an impact on the willingness of some vendors to implement |
| 08:32 | <annevk> | "I look forward to the more exciting XHTML2 phase of our Working Group" -- hehe |
| 08:33 | <asmodai> | Oh wow |
| 08:33 | <asmodai> | Valve is beta testing webkit for their Steam service |
| 08:33 | <asmodai> | http://store.steampowered.com/uiupdate/ |
| 08:34 | <hsivonen> | annevk: source? |
| 08:35 | Philip` | notes that the name "HTML5 (including next generation additions still in development)" came after a long discussion about how people would get confused if they were pointed to something that people normally call "HTML5" and the document did not call itself HTML5 |
| 08:36 | <annevk> | hsivonen, http-state mailing list, othermaciej being sarcastic about the suggestion that we should replace cookies |
| 08:37 | <othermaciej> | annevk: after he called it "phase 2" I could barely resist |
| 08:37 | <Philip`> | asmodai: It's interesting since they've used IE for years, and have presumably put a lot of effort into understanding it, and only care about Windows support, and still WebKit offers sufficient technical benefits to switch |
| 08:38 | <asmodai> | Philip`: I just updated it, man, does it look slick |
| 08:38 | <Philip`> | ("we swapped out the Internet Explorer rendering engine with WebKit, which gives us a bunch of size, stability and performance benefits") |
| 08:38 | <asmodai> | yeah, it starts up noticeably faster too |
| 08:38 | <Dashiva> | Even without any direct benefits, being able to support a single version of webkit seems better than 3-4 different IE engines |
| 08:39 | <asmodai> | Jesus, I wonder how long they worked on this |
| 08:39 | <asmodai> | If I now click my games it will present me with so much more information |
| 08:39 | <Philip`> | Surely you only need to support one IE engine, and put the others in compatibility mode |
| 08:47 | <hsivonen> | annevk: thanks |
| 08:47 | <hsivonen> | Google FAIL: http://www.google.fi/search?q=http-state+archive |
| 08:48 | <Dashiva> | http-state worked :) |
| 08:49 | <hsivonen> | googling for 'ietf mailing list' and navigating worked, too |
| 08:50 | <othermaciej> | yes I was snarky to Adam: http://www.ietf.org/mail-archive/web/http-state/current/msg00589.html |
| 08:50 | <othermaciej> | I hope he forgives me |
| 08:53 | <Hixie> | i really, really, really wish i could put unique data into the method of the websocket handshake |
| 08:53 | <Hixie> | before the resource |
| 08:53 | <Hixie> | that would make it _so_ much harder to screw up the server-side implementation of this |
| 08:54 | <othermaciej> | definitely seems like it would reduce the attack surface |
| 08:54 | <Hixie> | something like "WS012 /resource/ HTTP/1.1" |
| 08:55 | <Hixie> | where the three digits have to be converted to a number by parsing them as a decimal |
| 08:55 | <MikeSmith> | Hixie: so what is preventing you from doing that? |
| 08:55 | <othermaciej> | annevk: in all sincerity, I am pretty dubious of the prospects for or even usefulness of an incompatible replacement for cookies, but it's probably not *quite* as silly as an incompatible replacement for HTML |
| 08:56 | <othermaciej> | MikeSmith: seems like it would be a pain for HTTP servers to dispatch on, if they dispatch based on method |
| 08:56 | <Hixie> | and then some other number has to be divided by that number, so even if the implementor really screws it up, the number they get will be zero, and the division will crash the server rather than let it be attacked |
| 08:56 | <othermaciej> | MikeSmith: also, I am sure it would make the HTTP gods furious with hot anger |
| 08:56 | <Hixie> | MikeSmith: the wg members are up in arms enough as it is, and i'm using existing methods |
| 08:56 | <annevk> | othermaciej, I'm skeptical as well |
| 08:57 | <annevk> | othermaciej, though maybe if HTTP authentication is made useful enough... but then I'm not sure if that would be sufficient |
| 08:57 | <othermaciej> | annevk: I guess I haven't seen Adam's use cases that would justify a replacement for cookies as the solution |
| 08:57 | <othermaciej> | annevk: being able to tie form-based login to HTTP auth would remove a decent fraction of the true need for cookies, but the way to achieve that is not through a protcol spec |
| 08:58 | <othermaciej> | annevk: but on the other hand, most of the cookies currently in my cookie store are not login cookies |
| 08:58 | <Hixie> | i love that if i telnet to port 80 of hixie.ch and then send "GET / HTTP/1.1" (with no Host:) I get a 400, but if I send "GET / WEBSOCKET" I get a 200 OK with an error message |
| 08:59 | <annevk> | othermaciej, yeah, but those might be replaced I suppose with localStorage and XHR... though maybe not |
| 08:59 | <othermaciej> | annevk: those aren't protocol specs either... |
| 08:59 | <annevk> | true, not sure how much protocol changes would be needed |
| 09:00 | <annevk> | maybe a little, to support logout |
| 09:00 | <Hixie> | hmm... if I include a Host:, they both work |
| 09:00 | <Hixie> | i wonder whether we can just send WebSocket instead of HTTP/1.1 |
| 09:00 | <annevk> | Hixie, maybe with the second one it assumes HTTP/0.9? |
| 09:00 | <othermaciej> | annevk: I guess I should hold my fire until Adam presents the use cases and or design for Phase 2 |
| 09:00 | <Hixie> | annevk: maybe... |
| 09:00 | <othermaciej> | annevk: the other thing Phase 2 makes me think of is... underpants gnomes |
| 09:02 | <MikeSmith> | eh? |
| 09:04 | <MikeSmith> | underpants gnomes? |
| 09:04 | <MikeSmith> | ah, south park |
| 09:05 | <MikeSmith> | btw, news flash: sitting in a chair all day in front of a computer can make you fat and unhealthy: http://opinionator.blogs.nytimes.com/2010/02/23/stand-up-while-you-read-this/ |
| 09:06 | <MikeSmith> | it appears that some scientists have been looking into this |
| 09:07 | <annevk> | this comes to mind: http://or.ly/ |
| 09:07 | <Dashiva> | MikeSmith: That's misrepresenting the article gravely |
| 09:08 | <Dashiva> | "In other words, irrespective of whether you exercise vigorously, sitting for long periods is bad for you." |
| 09:08 | <MikeSmith> | gravely indeed |
| 09:08 | <MikeSmith> | my apologies |
| 09:08 | <hsivonen> | annevk: I assume the site shows the orly owl, but I don't know because it's hostile to my Referer settings |
| 09:08 | MikeSmith | types up a retraction |
| 09:09 | <Philip`> | Dashiva: That sentence seems to be misrepresenting the rest of the article |
| 09:10 | <othermaciej> | dang I sit for long periods |
| 09:10 | <MikeSmith> | me too |
| 09:10 | <MikeSmith> | so I'm switching to lying down |
| 09:10 | othermaciej | reconsiders the idea of a standing height workstation |
| 09:11 | <MikeSmith> | sideways, Roman-feast style |
| 09:11 | <Dashiva> | othermaciej: Even better, walkstations |
| 09:11 | <Dashiva> | Threadmill + workstation |
| 09:11 | <Philip`> | MikeSmith: I'm not sure the Roman feast helps with weight issues |
| 09:11 | <Dashiva> | -h I guess |
| 09:11 | <Philip`> | Dashiva: That's what the article suggests |
| 09:12 | <jgraham> | Presumably it would be treadmill + workstation + need to go fast enough or the computer will die |
| 09:12 | <MikeSmith> | the best solution to the weight issues is probably to combine the sitting with a lot of meth smoking |
| 09:13 | <Hixie> | what exactly is the "bad"? |
| 09:13 | <Dashiva> | jgraham: Can't go that fast, or you get exhausted before the end of the day |
| 09:13 | <jgraham> | Right but presumably the benefits are lost somewhat if you just stand there |
| 09:13 | <Dashiva> | Even standing still is a great improvement over sitting |
| 09:14 | <Dashiva> | I like how the references section is about as long as the main article |
| 09:14 | <Philip`> | Hixie: Obesity, diabetes, heart disease, cancer, and death, which all sound quite bad to me |
| 09:14 | <MikeSmith> | Dashiva: her articles are almost always like that |
| 09:15 | <Hixie> | Philip`: good to know |
| 09:15 | <Dashiva> | A ray of hope in an otherwise desolate journalistic landscape |
| 09:15 | <Philip`> | Sitting down can kill puppies too |
| 09:15 | <MikeSmith> | Dashiva: she's actually more of genuine scientist that an journalist.. some of her prose is just awful |
| 09:15 | <MikeSmith> | but her stuff is always worth reading |
| 09:16 | <MikeSmith> | Hixie: btw, need I ask, but are you perhaps planning to attend the hybi f2f in Anaheim on March 24? |
| 09:17 | <Hixie> | not in person |
| 09:17 | <MikeSmith> | ok |
| 09:17 | <Hixie> | probably on irc and listening, though if it doesn't seem productive i'll drop off |
| 09:18 | <MikeSmith> | hmm, that really makes me think.. there's no fee for calling in to the meeting, right? |
| 09:18 | <annevk> | nope |
| 09:19 | <MikeSmith> | if you were to actually attend in person, I think you'd need to pay 200 USD to do it |
| 09:19 | <annevk> | you don't call in actually |
| 09:19 | <MikeSmith> | ah yeah |
| 09:19 | <MikeSmith> | right |
| 09:19 | <annevk> | you get a live audio feed and there's an IM bachchannel |
| 09:19 | <Hixie> | MikeSmith: ! |
| 09:19 | <Hixie> | $200?! |
| 09:19 | <foolip> | spec experts: is it valid to use percent-encoding even when it isn't necessary? e.g. to encoding a as %61 |
| 09:19 | <foolip> | in URIs that is |
| 09:19 | <MikeSmith> | Hixie: that is what I had to pay to attend Hiroshima for 1 day |
| 09:19 | <annevk> | foolip, yup |
| 09:20 | <foolip> | ok |
| 09:20 | <annevk> | yeah, and USD 600 for a week or so |
| 09:20 | <Hixie> | wow, that's worse than the w3c's tpac $50/day extortion |
| 09:20 | <Hixie> | man i hate f2fs |
| 09:21 | <othermaciej> | MikeSmith: dang, that's expensive |
| 09:21 | <othermaciej> | I was thinking about going or sending a minion but I'm not sure I could justify that cost on top of travel and hotel |
| 09:22 | <MikeSmith> | for the IRC part in IETF meetings, you are allowed to get in the speaker queue and type questions and the chairs read them out, right? |
| 09:22 | <MikeSmith> | othermaciej: it may be there is some other way to attend without paying that much but if so, I don't know about it and it was not offered to me as an option |
| 09:22 | Philip` | had to pay €565 for 5 days at a networking conference |
| 09:23 | <annevk> | MikeSmith, yeah, not necessarily by the chairs though |
| 09:23 | <Philip`> | (Well, my university had to pay) |
| 09:23 | Philip` | concluded that it wasn't worth it |
| 09:23 | <othermaciej> | MikeSmith: there's three sessions of interest in the whole f2f, each on a different day, and nothing I would be remotely interested in for the rest of the day |
| 09:23 | <MikeSmith> | Philip`: not enough massages and mint juleps included in the deal? |
| 09:24 | <othermaciej> | wait, I forgot about IR |
| 09:24 | <othermaciej> | IRI |
| 09:24 | othermaciej | wonders if that is a fourth day |
| 09:24 | <MikeSmith> | othermaciej: luckily, last time time IRI and hybi meetings were on the same day |
| 09:24 | <MikeSmith> | I don't know about this year |
| 09:24 | <Philip`> | MikeSmith: None at all |
| 09:24 | <MikeSmith> | Philip`: then you definitely got gipped |
| 09:24 | <Philip`> | (The hotel was another €500) |
| 09:24 | <othermaciej> | hybi, iribis, httpbis, http-state would be the sessions I might be interested in |
| 09:25 | <MikeSmith> | ah yeah, http-state |
| 09:25 | <annevk> | othermaciej, same here, but I don't think I'll be going |
| 09:25 | <MikeSmith> | I guess abarth will be that one at least |
| 09:25 | <MikeSmith> | there was no http WG f2f in Hiroshima last time |
| 09:26 | <annevk> | Though I'm considered for IRI co-chair apparently there are other candidates as well so it doesn't really seem worth my time to fly over there. On top of that there's something fun going in the Netherlands at the same time :) |
| 09:27 | <MikeSmith> | I wonder if it's permitted to allow people to voip/Skype others in so they can actually ask and answer questions verbally instead of just by IRC |
| 09:27 | <Hixie> | MikeSmith: my understanding is taht it is |
| 09:27 | <Hixie> | or at least, that option was offered to me for anaheim |
| 09:27 | <MikeSmith> | OK |
| 09:28 | <Hixie> | frankly though the idea of _charging_ people to participate in what should be completely open standards development is just insane to me |
| 09:28 | <Hixie> | i hated when w3c did it, and i hate it when ietf does it |
| 09:28 | <Hixie> | especially given how often ietf does it |
| 09:29 | <othermaciej> | so httpbis is Monday, httpstate is Tuesday, hybi is Wednesday, and iri is Friday |
| 09:29 | <MikeSmith> | I think part of the IETF reason is that they othwerwise charge no membership fees |
| 09:29 | <othermaciej> | Hixie: I wish the W3C could figure out sponsorship instead of charing fees to individuals |
| 09:29 | <Hixie> | (if google hadn't paid for a bunch of people to go to the htmlwg meeting, i probably wouldn't have gone) |
| 09:29 | <othermaciej> | for TPAC |
| 09:30 | <Hixie> | i'm really not at all convinced f2fs are that good |
| 09:30 | <othermaciej> | like have some corps pay big bucks to have their logos on a page in the conference program |
| 09:30 | <MikeSmith> | othermaciej: well, me too |
| 09:30 | <Hixie> | people keep saying they're useful because of the social aspects |
| 09:30 | <Hixie> | but that hasn't helped the htmlwg much as far as i can tell |
| 09:30 | <Hixie> | mostly because most people can't go anyway |
| 09:30 | <Hixie> | so why bother |
| 09:31 | <othermaciej> | TC-39 meeting is Wednesday and Thursday, the same week as the IETF meeting, so I'm missing at least hybi |
| 09:31 | <othermaciej> | which is too bad because that's the one where I most expect to be able to have a useful impact |
| 09:32 | <zcorpan> | #whatwg is useful because of the social aspects |
| 09:32 | <zcorpan> | but it hasn't helped the htmlwg much as far as i can tell |
| 09:33 | <Hixie> | we started an irc channel quite a long time into the whatwg being quite productive |
| 09:33 | <Hixie> | iirc |
| 09:33 | <Hixie> | so i don't think irc really helped in that respect |
| 09:33 | <Hixie> | it's fun and helpful, but it's not necessary, imho |
| 09:33 | <Hixie> | f2fs aren't even fun and helpful |
| 09:33 | <Hixie> | they're miserable and generally a waste of time, imho |
| 09:34 | <MikeSmith> | 2010 W3C AC meeting is 22nd and 23rd March in Cambridge, btw. so anybody attending that won't be able to attend the first two days of the IETF meetings |
| 09:34 | <othermaciej> | I feel like I get some value out of f2fs but I have never tried to quantify |
| 09:34 | <Dashiva> | <Dracula> What is a f2f? |
| 09:34 | <Hixie> | face-to-face |
| 09:35 | <Dashiva> | It was a joke, pay it no heed :) |
| 09:35 | <Hixie> | othermaciej: i used to think so, then i tried to measure the net impact on the specs i was working on, and i determined that it didn't unblock any blocked issues, and yet basically inserted a 10-day delay (assuming a 3-day meeting with 2 travel days) |
| 09:35 | <Hixie> | Dashiva: d'oh |
| 09:35 | <MikeSmith> | as I think hsivonen pointed out, the worst part by far of f2f meetings is the travel part |
| 09:36 | <MikeSmith> | especially if you happen to live in a part of the world far away from where most of the f2f meetings take place |
| 09:37 | <othermaciej> | Hixie: the opportunity cost is my biggest worry |
| 09:38 | <othermaciej> | the main value I feel I have experienced is in conveying information about complex technical issues that seemed impossible to transfer in email, and opportunistically learning interesting facts from people I run into |
| 09:47 | <Dashiva> | I don't understand larry's latest blog post. What are these superfluous MUSTs he's concerned about? |
| 09:49 | <Philip`> | Dashiva: You mean all the anti-competitive ones? |
| 09:51 | <zcorpan> | "HTML5 also suffers from problems in Firefox 2 and Camino 1 because these two browsers use the Gecko rendering engine." - http://www.whatcreative.co.uk/blog/?p=410 |
| 09:51 | <Dashiva> | The only one I can think of is that .width and .height aren't allowed to disappear for no reason |
| 09:52 | <Dashiva> | And that just shows how little he cares about usable APIs |
| 09:52 | <zcorpan> | where are .width and .height not allowed? |
| 09:53 | <zcorpan> | or were there too many negatives in that sentence |
| 09:53 | <Dashiva> | The values must remain available once they become available |
| 09:54 | <Philip`> | What happens to .width/.height in e.g. Opera when you disable images after loading a page? |
| 09:54 | zcorpan | assumes they return 0 but hasn't tested |
| 09:54 | <Dashiva> | The images remain with the same size, they just don't render |
| 09:54 | <Dashiva> | The layout is fully preserved |
| 09:55 | <Dashiva> | Hum |
| 09:55 | <Dashiva> | Maybe that's styling interfering, though. Let's check otherwise. |
| 09:56 | <zcorpan> | Dashiva: i get layout changes when i turn off images |
| 09:56 | <zcorpan> | Dashiva: at least when there's alt text |
| 09:57 | <Dashiva> | Yeah, there was a fixed image size where I was checking. |
| 09:57 | <Dashiva> | Without styles, it resizes to fit the alt tex, and the dimensions match that |
| 09:58 | <annevk> | sounds like some kind of bug |
| 10:07 | <MikeSmith> | annevk: was there any kind of resolution at all on the "Backward-compatibility of text/html media type" thread on the TAG mailing list a few weeks back? |
| 10:08 | <MikeSmith> | I think you posted a few messages about that |
| 10:08 | <annevk> | I did and I don't think there was |
| 10:09 | MikeSmith | finds http://lists.w3.org/Archives/Public/www-tag/2010Feb/thread.html#msg23 |
| 10:09 | <annevk> | You sometimes go in a discussion with the TAG but then it takes ages for them to reply |
| 10:09 | <MikeSmith> | I think they may be waiting on some kind of update from the HTML WG |
| 10:10 | <annevk> | I guess in the end we just disagreed |
| 10:11 | <MikeSmith> | I'm trying to find what HTML WG bugzilla bug is associated with that, if any |
| 10:12 | <Hixie> | i randomly picked this mail from that thread: |
| 10:12 | <Hixie> | http://lists.w3.org/Archives/Public/www-tag/2010Feb/0027.html |
| 10:12 | <Hixie> | which asks: |
| 10:12 | <Hixie> | 1. Is there an intention to have at least the vast majority of the older |
| 10:12 | <Hixie> | content work and be considered conforming? |
| 10:12 | Hixie | wonders if Noah realises that the vast majority of the older content _isn't_ considered conforming even under the old rules |
| 10:13 | <MikeSmith> | I guess I first probably should try to figure out exactly what the TAG seems to be asking for around this |
| 10:14 | othermaciej | wonders if Noah realizes that "work" and "be considered conforming" are independent axes? |
| 10:14 | <MikeSmith> | OK |
| 10:14 | <MikeSmith> | I think I understand what the question/request is |
| 10:15 | <jgraham> | Not really independent, but not parallel either |
| 10:15 | <MikeSmith> | I think it's about whether the spec says that documents with an HTML 2, 3.2, 4 or XHTML 1 doctype are considered to be text/html documents or not |
| 10:15 | jgraham | doesn't think that parallel is quite the right word |
| 10:16 | <othermaciej> | in the case of HTML 4.01, there are documents that "work" but are not conforming, and documents which do not work, but are conforming |
| 10:16 | <MikeSmith> | http://www.w3.org/2001/tag/group/track/actions/364 is public right? |
| 10:16 | <othermaciej> | maybe not an orthonormal basis, but definitely not identical |
| 10:17 | <othermaciej> | MikeSmith: is there any way for me to tell, given that I have Member access? |
| 10:17 | <othermaciej> | I guess I could figure out how to delete my http password |
| 10:17 | <MikeSmith> | aren't all the tracker instances world-readable? |
| 10:18 | <MikeSmith> | othermaciej: using curl or wget or whatever, I guess |
| 10:18 | <othermaciej> | it curls |
| 10:19 | <MikeSmith> | OK, so I see there that action is a "give somebody else an action" action |
| 10:19 | <MikeSmith> | and the somebody else in this case is me |
| 10:19 | <othermaciej> | yes |
| 10:19 | <MikeSmith> | or PLH |
| 10:20 | <MikeSmith> | and the action we are expected to do is to write a change proposal |
| 10:20 | <zcorpan> | x3d profiles, hmm |
| 10:20 | <othermaciej> | on something you may or may not believe in |
| 10:20 | <MikeSmith> | yeah |
| 10:20 | <othermaciej> | I would much prefer for Change Proposals to be written by actual advocates for the change in question |
| 10:20 | <othermaciej> | (no need to be a WG member; DanC is signed up for one already) |
| 10:21 | <MikeSmith> | that'd certainly seem to be the best way to get them written, if they are to be written well |
| 10:21 | <MikeSmith> | this is otherwise like some kind of telephone-game thing |
| 10:22 | <othermaciej> | however that issue is also one we decided we do not need to address before LC |
| 10:22 | <MikeSmith> | oh |
| 10:22 | <MikeSmith> | I sorta remember that now, yeah |
| 10:22 | <MikeSmith> | OK, so I guess I can respond to say that |
| 10:22 | <othermaciej> | which is why it is in "HTML5 Spec - PR Blockers" instead of "HTML5 spec" |
| 10:23 | <othermaciej> | which I guess means my open issue count is off by one |
| 10:24 | <othermaciej> | I wish I could look at the issues for a specific product |
| 10:25 | <MikeSmith> | you can, can't you? |
| 10:26 | <othermaciej> | how? |
| 10:26 | <MikeSmith> | http://www.w3.org/html/wg/tracker/products/8 |
| 10:26 | <othermaciej> | how that's cool |
| 10:26 | MikeSmith | reads the recent public-html thread related to this |
| 10:26 | <othermaciej> | it even consolidates open and raised |
| 10:26 | <othermaciej> | this is now my favorite page on the tracker: http://www.w3.org/html/wg/tracker/products/1 |
| 10:27 | <zcorpan> | maybe the title should be "HTML5 (*plus* next-gen additions additions still in development)" |
| 10:27 | <annevk> | hmm appending ,access does not work on dynamic URIs |
| 10:28 | <MikeSmith> | annevk: right, unfortunately |
| 10:28 | <othermaciej> | zcorpan: that would make more sense |
| 10:29 | <zcorpan> | Hixie: ^ |
| 10:29 | <MikeSmith> | othermaciej: can you give me some guidance from your chairs perspective about what update I should give to the TAG about this? |
| 10:29 | <hsivonen> | annevk: cool. I was unaware of ,access |
| 10:30 | <othermaciej> | MikeSmith: I dunno |
| 10:30 | <othermaciej> | MikeSmith: have they made a request of you yet? |
| 10:30 | <MikeSmith> | othermaciej: just, "The group is planning resolve the text/html media-type registration issue after LC" |
| 10:31 | <MikeSmith> | hmm, I don't think that will fly |
| 10:31 | <MikeSmith> | othermaciej: DanC pinged me about it |
| 10:31 | <MikeSmith> | hot potato |
| 10:31 | <othermaciej> | MikeSmith: we previously decided that ISSUE-53 doesn't need to be addressed before LC and had WG consensus on that change |
| 10:31 | <Hixie> | zcorpan: given how long it took to get the title be something everyone in that discussion was happy with, i'd rather not reopen that discussion |
| 10:31 | <othermaciej> | MikeSmith: that doesn't mean we can't make any related changes, but we are unlikely to reconsider and make it an LC blocker |
| 10:32 | <MikeSmith> | OK |
| 10:32 | <zcorpan> | Hixie: i'd be suprised if anyone objected to s/including/plus/ but ok |
| 10:32 | <MikeSmith> | +1 to plus |
| 10:33 | <annevk> | makes sense to me too fwiw |
| 10:33 | <annevk> | (though the term HTML5 would still be used differently, but that seems fine) |
| 10:34 | <hsivonen> | I support s/including/plus/ |
| 11:01 | <Philip`> | "HTML5+" |
| 11:09 | <nessy> | lol |
| 11:11 | <othermaciej> | Philip`: for all those people who miss "HTML+" |
| 11:12 | <othermaciej> | whoah, HTML+ had a figure element! (named FIG, but still...) |
| 11:13 | <Hixie> | html3 had a <math> element |
| 11:13 | <Hixie> | a lot of the stuff in html5 is inspired by older drafts |
| 11:13 | <Hixie> | much like a lot of the stuff in html "6" will be inspired by things we dropped in 5 |
| 11:13 | <Hixie> | like datagrid |
| 11:14 | <othermaciej> | HTML+ also had a LIT element for content to be rendered literally, but in a proportional font, giving a verse example |
| 11:14 | <othermaciej> | that's a neat idea |
| 11:14 | <othermaciej> | though of course you could just style PRE |
| 11:14 | <MikeSmith> | nessy: can you remind me if any of the media subgroup proposals are ready for moving forward within the a11y TF this week? |
| 11:14 | <nessy> | I think they are both, actually |
| 11:14 | <MikeSmith> | OK |
| 11:14 | <nessy> | you mean at Friday's meeting? |
| 11:15 | <MikeSmith> | I will add them to the agenda |
| 11:15 | <nessy> | ah, ok - I thought that had already happened last week :) |
| 11:15 | <Philip`> | "HTML+ documents offer a means for providing hypertext links to a variety of media including images, sound sequences, MPEG movies, Postscript files and other formats." - we still don't have MPEG support :-( |
| 11:15 | <nessy> | discussion will keep going though, but an in-principle agreement that we are on the right way and ready for trial implementations would be good |
| 11:15 | <othermaciej> | nessy: what's the state of the media subgroup proposals? |
| 11:16 | <nessy> | http://www.w3.org/WAI/PF/HTML/wiki/Media_TextAssociations |
| 11:16 | <nessy> | http://www.w3.org/WAI/PF/HTML/wiki/Media_MultitrackAPI |
| 11:16 | <othermaciej> | nessy: the chairs (well, me and Sam) were wondering today when to put out a formal Call for Proposals for ISSUE-9 video-accessibiity |
| 11:16 | <nessy> | oh! |
| 11:17 | <nessy> | well, the latter one is related ot issue-9 |
| 11:17 | <nessy> | I guess I assumed that because the subgroup existed, the video a11y related issues and bugs were sorta "delegated" |
| 11:18 | <nessy> | I think after MikeSmith has had them discussed at the a11y TF meeting, it'd be good to introduce them to the larger html wg |
| 11:18 | <othermaciej> | nessy: everything ultimately has to go through the HTML WG process |
| 11:18 | <nessy> | sure |
| 11:18 | <othermaciej> | nessy: quick question about the first - how do <track> and <trackgroup> interact with <source>? |
| 11:19 | <MikeSmith> | nessy: I think the movement forward on the proposals is pending me putting them to the a11y TF for sign-off that we are ready to formally ask for review from the wider WG |
| 11:19 | <nessy> | not at all |
| 11:19 | <nessy> | MikeSmith, agree - just what I thought |
| 11:19 | <othermaciej> | all the examples use <video src="">, and not multiple <source> elements |
| 11:19 | <othermaciej> | and it is not clear to me how you would use them with <source> elements |
| 11:19 | <othermaciej> | could you add examples like that? |
| 11:19 | <nessy> | that's just because after the source selection algorithm, a single video source will be dealt with |
| 11:20 | <nessy> | I'll add an example, no problems |
| 11:20 | <othermaciej> | do all <track> elements apply to all sources, regardless of which is picked? |
| 11:20 | <nessy> | yes |
| 11:20 | <annevk> | that seems problematic |
| 11:20 | <nessy> | they are external files that get associated to the selected video source |
| 11:20 | <nessy> | how so? |
| 11:20 | <annevk> | there's no guarantee <source>es are of equal length |
| 11:21 | <annevk> | e.g. based on device features you might get a shorter video |
| 11:21 | zcorpan | doesn't think that's especially problematic |
| 11:21 | <nessy> | I think that's an author's problem if he provides alternatives that are not really alternatives |
| 11:21 | <othermaciej> | I was thinking you might want the option to nest <track> inside <source>, but I am not sure there is a compelling use case |
| 11:22 | <annevk> | currently the primary use for <source> is different formats, but I expect that to shift over time |
| 11:22 | <nessy> | yeah, we had such a proposal at one stage |
| 11:22 | <othermaciej> | the example I thought of is, you have one video with no captions, one video with burned-in captions, and external captions in other languages |
| 11:22 | <othermaciej> | I am not sure if that is realistic |
| 11:22 | <annevk> | nessy, why would they not be alternatives? |
| 11:22 | <nessy> | such an example is realistic |
| 11:23 | <nessy> | alternatives is not a good word choice |
| 11:23 | <annevk> | zcorpan, yeah, maybe not |
| 11:23 | <annevk> | othermaciej, nesting inside <source> might no longer be feasible |
| 11:24 | <nessy> | but what I mean is: right now the choice is mostly based on formats, and maybe later on devices (through the media queries) - what is provided to the user should, however, not be a different experience |
| 11:24 | <annevk> | othermaciej, unless we change deployed content and fix our parsers quickly... |
| 11:24 | <nessy> | it would be strange if you went to a site on different browsers and you got different videos |
| 11:24 | <othermaciej> | annevk: we might need to make <source> imply </source> to do that |
| 11:24 | <othermaciej> | yuck |
| 11:24 | <nessy> | annevk: yes, nesting inside <source> was rejected because of that problem |
| 11:24 | <annevk> | othermaciej, still fails for the last though I suppose </video> could close the last one... |
| 11:25 | <othermaciej> | nessy: the basic idea of Media_TextAssociations seems pretty good |
| 11:26 | <othermaciej> | nessy: is that what the SRT vs DFXP vs smilText format war has been about? |
| 11:26 | <nessy> | it tried to really match with the way in which tracks are encoded in MPEG, too, including the track grouping |
| 11:27 | <nessy> | well, the srt - DFXP - smilText discussion is about what external format to support - the markup itself seems rather uncontroversial by now |
| 11:27 | <nessy> | also, it seems only the SMIL guys are for smilText nobody else has spoken up for it yet |
| 11:27 | <annevk> | really, requiring DFXP? |
| 11:27 | <annevk> | omg |
| 11:27 | <nessy> | several voices have been for DFXP - mostly those that want professional captioning |
| 11:28 | <othermaciej> | nessy: the track access API also seems reasonable - I worry slightly that it's more power than needed for the use case, but we may end up going that way anyway |
| 11:28 | <annevk> | DFXP is extremely broken |
| 11:28 | <annevk> | and way too complex |
| 11:28 | <nessy> | but we seem to be agreeing that DFXP needs to define conformance levels, which will make it easier to implement support for it |
| 11:28 | <othermaciej> | I think it would be plausible to not require any format or require only SRT (which I gather is the bare-bones format) |
| 11:28 | <annevk> | SRT has some weird stuff too |
| 11:28 | <othermaciej> | is smilText less broken than DFXP? |
| 11:28 | <annevk> | it would need to be defined |
| 11:28 | <annevk> | I'm not sure we'd want either |
| 11:29 | <annevk> | it seems to me that SRT addresses well over 80% of the use cases |
| 11:29 | <annevk> | for the rest we could provide an API so you can implement stuff yourself |
| 11:29 | <nessy> | we will need to have something more than DFXP since captioning requirements by law require more quality than what DFXP can provide, IIUC |
| 11:29 | <nessy> | (that's US law) |
| 11:29 | <nessy> | SRT addresses subtitles very well |
| 11:30 | <nessy> | quality captioning has actually more indepth requirements, such as colours and formatting and positioning |
| 11:30 | <nessy> | nothing that couldn't be done with a simple subpart of DFXP and a mapping to HTML+CSS |
| 11:31 | <othermaciej> | so DFXP is really complicated, but still cannot meet US legal requirements for captioning? |
| 11:31 | <nessy> | at the lowest level, DFXP could be no more than SRT plus basic formatting markup in the captions plus CSS |
| 11:31 | <nessy> | no, no, DFXP meets all US legal requirements |
| 11:31 | <nessy> | I referred to SRT not meeting them |
| 11:32 | <othermaciej> | I see, ok |
| 11:32 | <nessy> | sorry about the double negation :) |
| 11:32 | <Philip`> | "we will need to have something more than DFXP" - did you mean SRT there? |
| 11:33 | <annevk> | I thought SRT had formatting extensions |
| 11:33 | <nessy> | I think all browser vendors will initially only want to implement srt - and that's fine |
| 11:33 | <othermaciej> | are the track roles defined in Media_MultitrackAPI all things that can be determined in common container formats? |
| 11:33 | <Philip`> | (Otherwise I'm not quite understanding) |
| 11:33 | <othermaciej> | Philip`: i think she did |
| 11:33 | <annevk> | Philip`, I think so |
| 11:33 | <nessy> | I think we have to take it in steps |
| 11:33 | <hsivonen> | nessy: what legal requirements does the U.S. have that .srt doesn't meet? |
| 11:33 | <nessy> | Philip' good catch - I said DFXP where I meant SRT |
| 11:33 | <othermaciej> | nessy: is there any expected use for <track> besides external caption tracks? |
| 11:33 | <nessy> | I meant: we will need to have something more than SRT |
| 11:34 | <othermaciej> | (wondering if the name is too general) |
| 11:34 | <hsivonen> | considering what closed captioning in the U.S. looks like on e.g. CNN, it's hard for me to believe that SRT didn't meet legal requirements |
| 11:36 | <nessy> | hsivonen: I don't know the details but apparently you need to be able to place captions at locations where they don't overlap central activity, you need to be able to mark up in colour and provide italics for certain things and stuff |
| 11:36 | <nessy> | othermaciej: I'm hoping we can use the <track> spec also for lyrics, karaoke and other such text stuff, and ultimately also for other audio and video tracks, though that last one will take a long time before browsers will feel comfortable implementing it |
| 11:37 | <annevk> | it seems we need to define something like Web SRT anyway; adding extensions for a few things shouldn't be too hard and definitely much easier than DFXP |
| 11:38 | <nessy> | this is an interesting read with what features broadcast captions provide: http://www.theneitherworld.com/mcpoodle/SCC_TOOLS/DOCS/SCC_FORMAT.HTML |
| 11:38 | <nessy> | annevk: DFXP is actually in use by a lot of companies already |
| 11:39 | <nessy> | it's not that difficult after all - and extending srt wouldn't serve a purpose here other than those who already do it saying: WTF and those who are doing DFXP saying: WTF |
| 11:39 | <nessy> | here's a typical DFXP file http://www.apple.com/media/us/mac/imac/2009/tours/apple-imac-design_video-cc-us-20091111_cc.xml |
| 11:40 | <nessy> | honestly not that different from SRT |
| 11:40 | <Hixie> | sweet jesus |
| 11:40 | <nessy> | ? |
| 11:40 | <Hixie> | namespace alert |
| 11:40 | <Hixie> | woop woop |
| 11:40 | <Hixie> | namespaced _attributes_ no less |
| 11:40 | <othermaciej> | dxfp looks just enough like HTML to confuse me |
| 11:41 | <nessy> | hehe :) |
| 11:41 | <Hixie> | wait is that a THIRD styling language at the w3c? or is DFCP not W3C |
| 11:41 | <Hixie> | DFXP |
| 11:41 | <hsivonen> | nessy: does the captioning functionality of analog over-the-air TV in the U.S. support italics and color? |
| 11:41 | <annevk> | Hixie, I guess so |
| 11:41 | <nessy> | Hixie: SMIL has the same stuff |
| 11:41 | <othermaciej> | I have a hard type typing or saying DFXP correctly - it's got to be the worst acronym ever |
| 11:41 | <othermaciej> | especially considering that internally it calls itself "Timed Text" |
| 11:41 | <nessy> | the ppl that developed it wanted it to be "working independent of the Web", too |
| 11:41 | <hsivonen> | Hixie: DFXP reinvents large parts of CSS and HTML |
| 11:42 | <Hixie> | can we please not reinvent the wheel |
| 11:42 | <Hixie> | why do subtitles have to be formatted? |
| 11:42 | <hsivonen> | nessy: always a reason to reinvent parts of the Web stack |
| 11:42 | <nessy> | hsivonen: it seems so - if you follow the link that I posted above, it has a list of what US captions are capable of |
| 11:42 | <othermaciej> | I think the cat is out of the bag on reinventing the wheel |
| 11:42 | <annevk> | omg it doesn't even refer to CSS for things like <color> |
| 11:42 | <annevk> | http://www.w3.org/TR/ttaf1-dfxp/#style-value-color |
| 11:42 | <Hixie> | maybe colour and italics, but sheesh |
| 11:42 | <Dashiva> | Aren't there plenty of subtitle formats in existence? |
| 11:42 | <annevk> | definitely do not want DFXP in a browser |
| 11:43 | <annevk> | evar |
| 11:43 | <othermaciej> | note the host in the URL nessy posted |
| 11:44 | <othermaciej> | but indeed this looks like it reinvents more wheels than SVG 1.2 Tiny |
| 11:44 | <nessy> | once you're over your initial shock, can I convince you to just look at it as a format to parse and map into proper html/css/etc? :) |
| 11:44 | <nessy> | http://www.apple.com/imac/ <- is the video, incidentally |
| 11:44 | <annevk> | that would mean writing a new spec for something that is very complex |
| 11:44 | <annevk> | whereas adding some features to SRT is much simpler |
| 11:45 | <Hixie> | why would we ever want to map subtitles to HTML/CSS/etc? |
| 11:45 | <Hixie> | surely we want subtitles to be media-independent and simple |
| 11:45 | <othermaciej> | it's so weird that they reused html element names with strangely different meanings |
| 11:45 | <Hixie> | so they can work with braille displays, speech synth, etc |
| 11:45 | <nessy> | that's what DFXP set out to be |
| 11:45 | <annevk> | Hixie, apparently there are some requirements for formatting |
| 11:45 | <Hixie> | what requirements? |
| 11:46 | <nessy> | read the bottom of this http://www.theneitherworld.com/mcpoodle/SCC_TOOLS/DOCS/SCC_FORMAT.HTML "Closed Caption Style Guid" |
| 11:46 | <annevk> | Hixie, placement, and color/italics (which you can prolly do media-independent in some way) |
| 11:46 | <Philip`> | According to DFXP, rgba(255, 255, 255, 1) is almost entirely transparent white |
| 11:46 | <nessy> | s/bottom/more or less middle/ |
| 11:46 | <Philip`> | whereas CSS says it's solid white |
| 11:46 | <Philip`> | Fun |
| 11:46 | <annevk> | really? |
| 11:46 | <annevk> | lolz |
| 11:46 | <Hixie> | SRT supports placement and color/italics |
| 11:46 | <nessy> | no, no placement |
| 11:46 | <hsivonen> | Hixie: not the kind of SRT that would make sense for the Web |
| 11:46 | <Hixie> | you don't need an inline box model or whatnot |
| 11:46 | <nessy> | italics and bold only for some extensions |
| 11:47 | <hsivonen> | Hixie: color and italics are in the RSS title land |
| 11:47 | <nessy> | I'd rather we use only pure srt and don't do any interpretation on random text, actually |
| 11:47 | <Philip`> | ("the alpha component, if expressed, is maximum (255) at 100% opacity and minimum (0) at 0% opacity" is what it says for all colour syntaxes) |
| 11:47 | <annevk> | nessy, if we can avoid the mess that is DFXP... |
| 11:48 | <annevk> | Philip`, oh wow, it uses 0-255 as scale rather than 0-1 for alpha? |
| 11:48 | <othermaciej> | that's kinda more logical |
| 11:48 | <othermaciej> | but tragically inconsistent |
| 11:48 | <annevk> | Philip`, they are either intentionally incompatible with CSS and all or just stupid |
| 11:49 | <Hixie> | i don't think you need to be stupid to have that kind of design error |
| 11:49 | <Philip`> | annevk: Seems like an easy bug to introduce if you're not being sufficiently careful |
| 11:49 | <Hixie> | just poorly informed |
| 11:49 | <nessy> | no comment ;) |
| 11:50 | <Philip`> | but the more fundamental problem is redefining CSS-like colours at all, rather than deferring to CSS, which would avoid any possible divergence |
| 11:50 | <nessy> | there was a lot of information from the traditional TV side of things, but the world of the Web seemed to have been a bit less relevant |
| 11:50 | <annevk> | o_O |
| 11:51 | <Hixie> | anyway, i'm all in favour of providing ways to define different voices, some of which might even by default be mapped to italics or specific colours, and i'm even ok with providing inline equivalents to <cite>, <em>, <i>, and <b>, but seriously, putting a CSS engine into subtitles is just a bad idea on so many levels |
| 11:52 | <othermaciej> | is what they have really equivalent to a CSS engine? |
| 11:52 | <nessy> | why would we need an extra CSS engine? browser would just use their existing one, once the format is mapped to html + css, no? |
| 11:52 | <nessy> | it's a lot simpler, actually |
| 11:52 | <Hixie> | nessy: not all UAs have a CSS engine |
| 11:52 | <nessy> | but yes, there will be issues |
| 11:53 | <othermaciej> | Hixie: that's a mean thing to say about IE |
| 11:53 | <nessy> | well, in that case you can only do best effort on captions, too, no? |
| 11:53 | <nessy> | ahhh (lol) |
| 11:53 | <Hixie> | i didn't mean IE |
| 11:53 | <Hixie> | i mean like a braille display |
| 11:54 | <Hixie> | or my TV |
| 11:54 | <othermaciej> | I was joking |
| 11:54 | <nessy> | braille people aren't interested in subtitles or captions, I would think ;) |
| 11:54 | <Hixie> | or any number of other environments where embedding an entire CSS engine is overkill on a massive scale |
| 11:54 | <othermaciej> | nessy: a deaf-blind user might be |
| 11:54 | <Philip`> | othermaciej: They would be interested in the text, but not timed and linked to video/audio tracks |
| 11:55 | <Dashiva> | Why not audio tracks? |
| 11:55 | <nessy> | as I said: best effort for the given situation should be acceptable |
| 11:55 | <othermaciej> | Philip`: true, assuming a non-timed transcript was available |
| 11:55 | <Philip`> | Dashiva: Because they're deaf |
| 11:55 | <nessy> | we can't standardise on the lowest common denominator |
| 11:55 | <Dashiva> | Isn't it possible they're only blind? |
| 11:56 | <Philip`> | Dashiva: othermaciej was talking about deaf-blind users |
| 11:56 | <Philip`> | who, by definition, are deaf |
| 11:56 | <nessy> | we can do the lowest common denominator as a baseline, of course |
| 11:56 | <Dashiva> | Oh, didn't see that |
| 11:56 | <othermaciej> | Dashiva: my suggestion was that a deaf-blind user might want to use caption tracks via a braille display |
| 11:56 | <othermaciej> | a blind user would likely most want an audio description track |
| 11:57 | <Philip`> | (although I suppose they could be hearing-impaired such that they can hear music but can't understand speech) |
| 11:57 | <Philip`> | (so it's probably nicer to accommodate such people if possible) |
| 11:58 | <nessy> | so, I really do wonder: is there a problem with just parsing the external file (SRT, DFXP, or whatever format we decide on), mapping it to existing HTML/CSS/JavaScript/ whatever constructs, and displaying it in the browser that way? |
| 11:58 | <Philip`> | (but not if it significantly compromises the majority use cases) |
| 12:00 | <othermaciej> | nessy: my guess is that in our case we'd use QuickTime's existing implementations |
| 12:01 | <Hixie> | nessy: the video playback subsystem should not have to reimplement, or import, an HTML/CSS/JS/DOM/HTTP/etc engine just to render subtitles |
| 12:01 | <othermaciej> | nessy: I expect using the Web engine to present subtitles it would be harder to get the performance good and the synchronization accurate |
| 12:01 | <nessy> | othermaciej: I think that makes sense when the DFXP track is inside a QuickTime file - such as the 3GPP files that use it, but maybe not so much when it comes straight into the browser as a text file |
| 12:02 | <Hixie> | there should be no difference between an embedded file and an external file, imho |
| 12:02 | <othermaciej> | nessy: well I sure don't want two totally separate DFXP implementations |
| 12:02 | <Hixie> | they should be exposed using the same API, and rendered using the same subsystem |
| 12:02 | <nessy> | Hixie: I don't think you can pass an external text file through the media subsystem for parsing and interpretation |
| 12:03 | <nessy> | but ok, that's fair enough |
| 12:03 | <nessy> | the API that we are proposing has been designed to be the same across internal and external tracks |
| 12:04 | <nessy> | and maybe it is possible to reuse the parsing of a media subsystem |
| 12:04 | <nessy> | I'm sure going to have to work at encoding DFXP into Ogg... |
| 12:05 | <nessy> | or at least a sensible subpart of it ... |
| 12:05 | <hsivonen> | nessy: why? does someone want to write an Ogg player with DFXP support? |
| 12:06 | <nessy> | we cannot ignore the professional market with Ogg IMO |
| 12:06 | <nessy> | if we want to be omnipresent, we have to be able to do quality captions |
| 12:07 | <hsivonen> | quality captions I agree with |
| 12:07 | <nessy> | several open source players support basic DFXP already |
| 12:07 | <nessy> | out of MPEG and MOV files mainly |
| 12:08 | <othermaciej> | nessy: I wonder what the deal is with the way the page you linked references the external captions |
| 12:08 | <othermaciej> | the <text> element inside <video> |
| 12:09 | <nessy> | othermaciej: I couldn't get them out of firebug either - must be linked inside the file or something, not sure either |
| 12:09 | <othermaciej> | I don't see any support for that in WebKit |
| 12:09 | <nessy> | Eric showed me that file and since I like collecting examples, I keep using it |
| 12:10 | <othermaciej> | the <video> element has a child element <text type="application/ttaf+xml" src="/media/us/mac/imac/2009/tours/apple-imac-design_video-cc-us-20091111_cc.xml" lang="en"></text> |
| 12:10 | <othermaciej> | which is nonstandard |
| 12:10 | <othermaciej> | I should ask Eric what is up with that |
| 12:10 | <nessy> | oh, I didn't see that (silly me) |
| 12:10 | <othermaciej> | maybe there is also a caption track in the file |
| 12:10 | <nessy> | must be just some javascript then that does the display... |
| 12:10 | <hsivonen> | <text> overlaps with SVG. :-( |
| 12:11 | <othermaciej> | it seems there are also captions in the file |
| 12:11 | <othermaciej> | cause the controls show up even if I go straight to the video file: http://movies.apple.com/media/us/mac/imac/2009/tours/apple-imac-design_video-cc-us-20091111_r848-9cie.mov |
| 12:12 | <othermaciej> | (Safari uses a generated HTML document with a <video> element when you directly link to a video file) |
| 12:12 | <nessy> | ah, I should probably look at it in Safari to get the HTML5 markup!! |
| 12:12 | <nessy> | that would make a difference! |
| 12:12 | <othermaciej> | oh yeah, I got all that with the Web Inspector |
| 12:12 | <othermaciej> | I don't think the <text> element does anything |
| 12:12 | <nessy> | in Firefox I just got an object element of course :) |
| 12:14 | <nessy> | looks like somebody had similar ideas to us with the <text> element :) |
| 12:20 | <nessy> | anyway - looking at the DFXP thing as a long-term possibility and SRT as a quick here-and-now solution is probably where we are going |
| 12:24 | <nessy> | incidentally, neither has been "standardised" as such - we need to write a spec for SRT and register the mime type and the TimedText guys have to get to CR |
| 12:24 | <nessy> | you might also enjoy looking at smilText http://www.w3.org/TR/SMIL3/smil-text.html |
| 12:27 | <nessy> | we'll introduce the spec into public-html soon - maybe should cross-post to whatwg? |
| 12:40 | <othermaciej> | nessy: cross-posting messes up threads - I recommend not |
| 12:41 | <nessy> | yeah, I have had that experience in the past, too |
| 12:41 | <nessy> | so, maybe two posts - on each |
| 13:20 | <annevk> | hsivonen, I'd say no to 1 |
| 13:20 | <annevk> | hsivonen, much like CSS obsoletes features over time, so should it be possible for HTML |
| 13:28 | <hsivonen> | annevk: do you mean that previously valid CSS1 is no longer valid CSS1? |
| 13:38 | <annevk> | hsivonen, yes |
| 13:39 | <hsivonen> | annevk: interesting |
| 13:54 | <annevk> | also, allowing HTML4 makes no sense since we already decided it's broken |
| 13:54 | <annevk> | all kinds of stuff validates as HTML4 and will never be processed as such |
| 13:54 | <annevk> | it makes no sense |
| 13:59 | <hsivonen> | annevk: I agree. |
| 13:59 | <hsivonen> | however, saying that DTD-valid HTML4 no longer is DTD-valid HTML4 is quite revisionist |
| 13:59 | <hsivonen> | but I'm not saying that it's useful to have DTD-valid HTML4 |
| 14:09 | <annevk> | I see |
| 14:10 | <annevk> | In CSS that works because CSS does not have "versions". Just different levels of features. So if something is changed that affects a features it will change in every level... |
| 14:11 | <annevk> | I tried to find a list of the things that changed, but can't see to find it now. At least one of the things is the escape syntax. |
| 14:14 | <Dashiva> | I remember clip changed in 2.1 |
| 15:33 | <boblet> | hey all, when talking about a11y Hixie mentioned “making features more explicitly media-independent” |
| 15:34 | <boblet> | does media-independent refer to eg not just visual browsers? |
| 15:34 | <boblet> | can anyone hazard a guess? |
| 15:38 | <smaug> | Hixie: ping |
| 15:44 | <annevk> | boblet, yup |
| 15:44 | <annevk> | boblet, he gave some examples, e.g. braille readers |
| 15:44 | <boblet> | annevk: thanks |
| 15:44 | <boblet> | I wasn’t around for the mention of braille readers unfortunately :) |
| 16:18 | <leejongwook> | firefox support html5 ? |
| 16:18 | <miketaylr> | sure |
| 16:18 | <leejongwook> | how about iceweasel ? |
| 16:18 | <leejongwook> | @_@ |
| 16:18 | <miketaylr> | depends on what you mean by html5, and which subset of it |
| 16:18 | <leejongwook> | <--- want to learn html5 |
| 16:19 | <leejongwook> | i know nothing but the name html5 |
| 16:19 | <leejongwook> | <--- need to setup my environment :) |
| 16:19 | <miketaylr> | leejongwook: http://diveintohtml5.org/ |
| 16:19 | <leejongwook> | so i need firefox :) and... that site :P |
| 16:19 | <miketaylr> | as well as the spec ;) |
| 16:20 | <leejongwook> | miketaylr, thanks :) |
| 16:20 | <miketaylr> | newer firefox, chrome, safari, opera all support html5 to varying degrees |
| 16:20 | <leejongwook> | i see :) |
| 16:22 | <smaug> | also IE8 supports some html5 |
| 16:22 | <miketaylr> | oh, good call |
| 16:24 | <leejongwook> | :) |
| 16:27 | <leejongwook> | FlowPlayer <--- this is interesting |
| 16:29 | <leejongwook> | Everything you know about XHTML is wrong <-- good, 'cause i don't know xhtml :P |
| 16:48 | <leejongwook> | i doubt my eyes @_@ |
| 16:49 | <leejongwook> | so it's not flash player :P and it plays video @_@ without any kind of plugins |
| 16:50 | <leejongwook> | can't compare this with html4, only seen it with flash |
| 16:58 | <leejongwook> | In other browsers that do not support <video>, it falls back to QuickTime. <--- i see |
| 17:00 | <leejongwook> | It’s important to note that the user is not prompted to install QuickTime if they don’t have it. Fallback is instant and automatic |
| 17:00 | <leejongwook> | cool :) |
| 17:03 | <leejongwook> | http://camendesign.com/code/video_for_everybody <--- i saw this |
| 17:03 | <leejongwook> | nice :) |
| 17:05 | <leejongwook> | so i don't need to learn action script and svg to share videos and to draw something on |
| 17:05 | <leejongwook> | all in one :P |
| 17:07 | <leejongwook> | thanks again :P miketaylr smaug, see you all next time :) |
| 17:07 | <miketaylr> | later |
| 17:07 | <leejongwook> | :) |
| 17:09 | <zcorpan> | isn't drawFocusRing() better than image maps? |
| 17:19 | <Philip`> | Does anybody still seriously use image maps? |
| 17:19 | <Philip`> | I thought server-side image maps went out of fashion in the 90s, and client-side ones in the early 00s |
| 17:26 | <mpt> | Philip`, they'd be a moderately vital part of any accessible replacement for Flash |
| 18:23 | <deltab> | Philip`: I'm already using SVGs with href |
| 18:32 | Philip` | discovers bizarre shadow compositing behaviour in Opera |
| 19:06 | <JonathanNeal> | The <command> element is buggy in Firefox. It is documented and validates only as self closing tag, but Firefox will not close the tag without a </COMMAND>. So, you can pick your poison; you can not use the element all-together, you can markup the element as a self-closing tag and suffer the Firefoxequences, or you can add the closing tag and ignore validation all-together. |
| 19:10 | <AryehGregor> | JonathanNeal, can't you put it in another tag, so when Firefox hits the wrapper's close tag, it will close it? |
| 19:13 | <JonathanNeal> | AryehGregor, <command /> will not close, so the next <command /> will be a child of the previous. |
| 19:13 | <Philip`> | JonathanNeal: I expect he means <div><command></div> |
| 19:14 | <Philip`> | (or similar) |
| 19:14 | <JonathanNeal> | The next element will be a child of <command, until it is closed by some other means. |
| 19:14 | <AryehGregor> | What does <div><command></div><div><command></div> do, or whatever? Does that make sense here? |
| 19:15 | AryehGregor | isn't clear on how or why one would use <command> in practice, so isn't sure if that makes sense. |
| 19:15 | <Philip`> | Why does it matter how Firefox parses it, given that the functionality is not implemented? |
| 19:15 | <AryehGregor> | And right, does it matter how it gets parsed? |
| 19:15 | <JonathanNeal> | Say: <menu type="toolbar"> |
| 19:15 | <JonathanNeal> | <command onclick="insertTag(buttons, 0);" label="strong" icon="bold.gif"/> |
| 19:15 | <JonathanNeal> | <command onclick="insertTag(buttons, 1);" label="em" icon="italic.gif"/> |
| 19:15 | <JonathanNeal> | </menu> |
| 19:16 | <JonathanNeal> | Ah crap, XChat didn't catch the line returns, sorry. |
| 19:16 | <AryehGregor> | So why can't they just be nested in Firefox? |
| 19:16 | <AryehGregor> | I mean, they're invisible anyway, aren't they? Or are you styling them somehow? |
| 19:16 | <AryehGregor> | <div> or <span> wrappers should fix it, anyway. |
| 19:16 | <JonathanNeal> | AryehGregor, nested as in each command being a child of the previous? |
| 19:17 | <AryehGregor> | Yeah. |
| 19:17 | <AryehGregor> | What's making these visible at all? |
| 19:17 | <JonathanNeal> | Well, then the styling would get in the way. |
| 19:17 | <Philip`> | It seems a bad idea to use those elements at all, because the spec might change (given that there's no implementations yet, as far as I'm aware) |
| 19:17 | <AryehGregor> | Yeah, I would very strongly advise that you avoid using anything with zero implementations. |
| 19:17 | <JonathanNeal> | That is a Catch 22. |
| 19:17 | <AryehGregor> | No, it's not. |
| 19:17 | <AryehGregor> | Implementations happen before features are used in authoring. |
| 19:18 | <Philip`> | By the time there's implementations, the parsers will have been fixed |
| 19:18 | <AryehGregor> | Once there's at least one implementation, you can test in that implementation. |
| 19:18 | <JonathanNeal> | It is the developers who have dictated the implementations. |
| 19:18 | <JonathanNeal> | That's how this whole mess started. |
| 19:18 | <AryehGregor> | Huh? |
| 19:19 | <AryehGregor> | If authors deploy features without implementations, then implementers may find that the sites break when the feature is actually implemented. |
| 19:20 | <AryehGregor> | If your site is big enough, that will hold up their implementation. If it's not, then your site will break. |
| 19:20 | <AryehGregor> | In any event, you gain nothing by using features that don't exist yet in *any* browser, so what's the point? Use <input type=image> instead. |
| 19:20 | <JonathanNeal> | AryehGregor, you're referring to Gecko or Webkit implementing <command> ? Okay, I just wanted to use a recommended element as a child of <menu> |
| 19:21 | <Philip`> | Why do you want to use <menu>? |
| 19:22 | <Philip`> | Nobody implements that either :-) |
| 19:23 | <AryehGregor> | What Philip` said. |
| 19:25 | <JonathanNeal> | Accessibility and giving better meaning to page controls. |
| 19:25 | <JonathanNeal> | let me provide an example |
| 19:25 | <JonathanNeal> | http://sandbox.thewikies.com/liferay-rotating-banner/ |
| 19:25 | <JonathanNeal> | The source downlaods are outdated, but you'll see on that page I'm using <menu> and <command> |
| 19:26 | <AryehGregor> | Why not just use <nav> and <a>? |
| 19:28 | <JonathanNeal> | Well, is there some markup that would allow it to be a control but also prevent the screen-reader from listing the control, since it is visual only. |
| 19:28 | <Philip`> | JonathanNeal: I think you'd actually need to use <li><command></li> to make that correct, otherwise the content is interpreted as merely "flow content describing available commands" and not as actual commands |
| 19:29 | <JonathanNeal> | Using a screen reader or keyboard navigation, you can already move through all of the banners, so it didn't make sense for it to then list the controls, since the keyboard and screen-reader already provide those controls. |
| 19:29 | <JonathanNeal> | The controls are for the visually unimpaired only. |
| 19:29 | <AryehGregor> | I have no idea, but I'm pretty sure that using a completely unimplemented feature isn't the best way to do it. |
| 19:29 | <AryehGregor> | (Maybe the validator should raise a warning when someone does that?) |
| 19:32 | <JonathanNeal> | Hey TabAtkins :D |
| 19:33 | <TabAtkins> | yo |
| 20:00 | Philip` | always forgets how many canvas tests he has :-/ |
| 20:00 | <TabAtkins> | 310? 306? Something like that. |
| 20:01 | <Philip`> | My text editor says I'm still only 60% of the way through the test file, and I've not even started writing any new ones yet |
| 20:01 | <Philip`> | I don't mean literally forgetting the number - I can check the index page which says 724 |
| 20:02 | <Philip`> | I just mean forgetting how much work it is to even read through all the tests |
| 20:02 | <TabAtkins> | Ah, kk. |
| 20:03 | <Philip`> | Would anybody mind if I made APNG support necessary in order to pass some tests? |
| 20:03 | <TabAtkins> | Nah, go right ahead. ^_^ |
| 20:38 | <JonathanNeal> | TabAtkins, I had a question for you about accessibility. |
| 20:38 | <TabAtkins> | shoot |
| 20:38 | <JonathanNeal> | I'm trying to create controls visible only to the visually impaired, is there ARIA for that, or an appropriate element? |
| 20:39 | <TabAtkins> | If they can be expressed with normal HTML semantics, but you just don't want them shown to sighted users, use the abspos trick. |
| 20:39 | <JonathanNeal> | I was using <menu> and <command>s but they might be considered unstable, and the validation / browser implementation doesn't match up. |
| 20:39 | <JonathanNeal> | I want them show to sighted users, not unsighted users. |
| 20:40 | <TabAtkins> | Ah, that's the exact opposite of what you just said. ^_^ |
| 20:40 | <JonathanNeal> | You're right, I meant to say "visually paired" |
| 20:40 | <JonathanNeal> | unsighted users get the entire website experience without these additional controls, example @ http://sandbox.thewikies.com/liferay-rotating-banner/ |
| 20:41 | <TabAtkins> | Hmm, actually I'm not sure then. Typically you're needing to expose extra controls to the disabled. Let me see... |
| 20:42 | <JonathanNeal> | The visually impaired wouldn't do well with the controls, but javascript markup seems like a secondary solution, if I can be educated on an elemental or ARIA solution. |
| 20:42 | <JonathanNeal> | "javascript markup seems like a secondary solution" eg including the controls via js. I don't think that would help with screenreaders either. |
| 20:43 | <JonathanNeal> | I can use a series of empty divs and spans, of course, but I thought there was another way. |
| 20:44 | <JonathanNeal> | I was inspired after I saw http://www.ibm.com/developerworks/library/x-html5/#N10405 |
| 20:44 | <TabAtkins> | Oh, duh. role=presentation |
| 20:45 | <TabAtkins> | That conflicts with the allowed semantics in the aria mapping table in html5, though. |
| 20:45 | <TabAtkins> | I'd bring that up on the list. |
| 20:47 | <JonathanNeal> | Wait, so it's good but bad? |
| 20:47 | <TabAtkins> | You want to use role=presentation. But it's officially disallowed right now in HTML5 (only <img alt=""> can have it). |
| 20:48 | <JonathanNeal> | Oh, and I want presentational html. |
| 20:49 | <TabAtkins> | Alternately... Just use empty elements and/or images with no alt for the controls. If the controls don't *contain* any content, they can't be read out. |
| 21:01 | <JonathanNeal> | In the meantime I may add the role="presentational", for possibile forward compatibility. |
| 21:01 | <JonathanNeal> | I wonder who IBM got the idea that <menu> and <command> would work the way they described. |
| 21:01 | <JonathanNeal> | *how |
| 21:04 | <JonathanNeal> | What's a good element to use for controls? |
| 21:05 | <Philip`> | JonathanNeal: I expect they got the idea by reading the spec |
| 21:05 | <Philip`> | although it wasn't "IBM" |
| 21:05 | <TabAtkins> | Depends. <div>, <span>, or <img> likely. |
| 21:05 | <Philip`> | and it wasn't even a person who works for IBM |
| 21:05 | <Philip`> | and it was 2.5 years ago |
| 21:05 | <JonathanNeal> | Right. |
| 21:06 | <JonathanNeal> | I have a section with controls at the top, I guess I could stick them in a <nav> |
| 21:06 | <JonathanNeal> | Since they don't belong to the content, but they do belong to the section |
| 21:27 | <TabAtkins> | Hrm. Just found a rendering bug in FF3.6. |
| 21:34 | <jwalden> | impressive! |
| 21:34 | <jwalden> | ;-) |
| 21:34 | <jwalden> | offered without comment: http://x264dev.multimedia.cx/?p=292#comment-2673 "Hi, that’s ten screenfuls of text, then eleven screenfuls of comments. Can you summarize? Thanks." |
| 22:03 | <supL> | hi |
| 22:04 | <TabAtkins> | yo |
| 22:52 | <JonathanNeal> | a breadcrumb wouldn't imply a <nav> right? |
| 22:54 | <TabAtkins> | I wouldn't personally mark it up as one, but it's arguable. |
| 22:55 | <JonathanNeal> | It's part of the content but not part of any particular section. |
| 22:56 | <JonathanNeal> | I'm referencing http://sandbox.thewikies.com/html5-layout/portal_normal.html |
| 23:06 | <Hixie> | there are examples of breadcrumbs in the spec, if they help |
| 23:10 | <JonathanNeal> | Oh yea? I'll check that out. |
| 23:10 | <JonathanNeal> | Where? |
| 23:10 | <JonathanNeal> | Maybe I'm looking at the wrong thing. |
| 23:11 | <JonathanNeal> | Oooh, I see. |
| 23:18 | <TabAtkins> | shepazutoo: Yo, Shep, you around? Have a question about SVG color interpolation for gradients. |
| 23:42 | <Philip`> | The new documentation in http://blogs.msdn.com/ie/archive/2010/02/24/documenting-standards-in-ie.aspx seems to be pretty boring |
| 23:43 | <Philip`> | (just in case anyone was tempted to read through it all for exciting new information) |
| 23:48 | <jcranmer> | hmm, they do mention hasLayout |