| 02:40 | <wirepair> | in html 4.01~5 the only two elements that start with <! are <!DOCTYPE and <!-- comments, yes? |
| 02:50 | <MikeSmith> | wirepair: yes |
| 02:50 | <MikeSmith> | wtf https://twitter.com/jashkenas/status/215093650789642241 |
| 02:50 | <wirepair> | cool thanks :) |
| 02:51 | <MikeSmith> | «Just now learning that Windows 8 Metro won't treat JavaScript as JavaScript without a "UTF8 BOM" in front.» |
| 02:51 | <wirepair> | ugh |
| 02:58 | <MikeSmith> | jwz http://www.jwz.org/blog/2012/06/i-have-ported-xscreensaver-to-the-iphone/ |
| 02:59 | <MikeSmith> | fun read |
| 02:59 | <wirepair> | what... exactly would you call the ! character in terms of HTML state (think when parsing) |
| 02:59 | <wirepair> | it's not a command processor like it is in XML right? |
| 03:00 | <MikeSmith> | eh? |
| 03:00 | <MikeSmith> | command processor in XML? |
| 03:00 | <wirepair> | isn't it? :> |
| 03:00 | <wirepair> | process instruction or something |
| 03:00 | wirepair | confused |
| 03:00 | <MikeSmith> | ah |
| 03:00 | <MikeSmith> | processing instruction starts with <? |
| 03:01 | <wirepair> | ah yes, processing instructions |
| 03:01 | <MikeSmith> | anyway, what if anything it's call in terms of HTML state is in the spec |
| 03:01 | <MikeSmith> | tokenizer |
| 03:02 | <MikeSmith> | http://www.whatwg.org/specs/web-apps/current-work/multipage/tokenization.html#tokenization |
| 03:02 | <MikeSmith> | http://www.whatwg.org/specs/web-apps/current-work/multipage/tokenization.html#markup-declaration-open-state |
| 03:02 | <MikeSmith> | or wind back to http://www.whatwg.org/specs/web-apps/current-work/multipage/tokenization.html#tag-open-state |
| 03:03 | <MikeSmith> | in "Tag open state", the ! puts the parser into "Markup declaration open state" |
| 03:03 | <wirepair> | cheers |
| 03:04 | <wirepair> | exactly what i needed, thanks! |
| 03:06 | <MikeSmith> | thank Hixie :) |
| 03:08 | <MikeSmith> | ah man jwz is an inspiration |
| 03:08 | <MikeSmith> | http://www.jwz.org/blog/2012/06/i-have-ported-xscreensaver-to-the-iphone/#comment-106843 |
| 03:08 | <MikeSmith> | "I do not find reading diffs in email to be a hardship, but I used to walk uphill both ways to school, so I dunno." |
| 03:10 | <MikeSmith> | "OpenGL 3.1 is basically a back-port of OpenGL ES to desktops. Its what you get when an already-not-very-bright child mates with its grandparent." |
| 03:27 | <MikeSmith> | oh yeah let's bikeshed on the name of the Web Notifications spec that we've been using for 2+ years |
| 03:28 | <MikeSmith> | http://lists.w3.org/Archives/Public/public-web-notification/2012Jun/0050.html |
| 03:29 | <MikeSmith> | "So I wonder if a better name wouldn't be "Browser notifications", reserving the name "Web notifications" for some future work with applicability to the whole Web." |
| 03:29 | <MikeSmith> | "I expected it to be about notifications on the Web, perhaps a backchannel notification extension to HTTP or something like that." |
| 03:29 | <MikeSmith> | here's an idea: how about calling that non-existent thing "HTTP Notifications" instead |
| 03:30 | <MikeSmith> | whatever teh hell a "backchannel notification extension to HTTP" is anyway |
| 03:30 | <MikeSmith> | god what world do these people live in |
| 03:31 | <MikeSmith> | I'm looking forward to Anne's response on that thread |
| 03:59 | <Hixie> | MikeSmith: i almost replied back "Web means Browser" but I didn't want to put Anne in the position of having to tell me to stop trolling :-) |
| 04:04 | <MikeSmith> | Hixie: hah |
| 04:09 | <Hixie> | "OpenGL 3.1 is basically a back-port of OpenGL ES to desktops. Its what you get when an already-not-very-bright child mates with its grandparent." ouch |
| 04:09 | <MikeSmith> | yeah, dude still has a way with words |
| 04:09 | <MikeSmith> | a still a coding madman it seems |
| 04:09 | <MikeSmith> | I thought he had given up on writing software |
| 04:10 | <MikeSmith> | with better things to do |
| 04:10 | <Hixie> | the one thing that's great about being editor for HTML is that if anyone insults your baby, it's unlikely to be anywhere near as bad as any insult you have yourself used about it |
| 04:10 | <MikeSmith> | hahaha |
| 04:13 | <MikeSmith> | HTML spec is like starting from that boy raised by wolves and trying to introduce him back into eating with paper napkins and silverware |
| 04:14 | <Hixie> | and doing so while in the middle of a banquet on a flatbed car on a train going 100km/h |
| 04:14 | <Hixie> | while desperately hoping that someone will invent some better form of transportable meal and make the whole thing irrelevant |
| 04:16 | <wirepair> | ha. |
| 04:16 | <MikeSmith> | hah |
| 04:17 | <MikeSmith> | Hixie: as far as inventing that next thing, I take comfort in knowing that Doug Crockford and Zed A. Shaw are already on top of that |
| 04:18 | <wirepair> | MikeSmith: is that sarcasm? (I don't know who those two people are) |
| 04:18 | <MikeSmith> | though sadly we no longer have Dmitry Turin working on it |
| 04:18 | <MikeSmith> | wirepair: deadly serious. read their wisdom and learn |
| 04:18 | wirepair | googles |
| 04:18 | <Hixie> | hm, i think i missed Zed's attempt |
| 04:19 | <Hixie> | oh |
| 04:19 | <Hixie> | that video recently |
| 04:19 | <Hixie> | yes |
| 04:19 | <Hixie> | can't wait to see what he comes up with that replaces the web |
| 04:19 | <wirepair> | i can't find anything, do you have a link MikeSmith ? |
| 04:19 | <MikeSmith> | Hixie: yeah that |
| 04:20 | <MikeSmith> | wirepair: http://vimeo.com/43380467 |
| 04:20 | <wirepair> | cheers |
| 04:20 | <Hixie> | MikeSmith: though i thought you were going to say Ted Nelson was on it |
| 04:20 | <MikeSmith> | reality is beneath Ted's attention |
| 04:21 | <MikeSmith> | thought experiments are much easier to actually implement |
| 04:21 | <MikeSmith> | wirepair: http://www.crockford.com/html/ |
| 04:23 | <MikeSmith> | wirepair: one thing I am actually serious about, Crockford is a fun speaker. He does speak wisdom about what the problems are, and does it in an entertaining way |
| 04:23 | <MikeSmith> | I wish I could say the same about Zed |
| 04:24 | <MikeSmith> | Zed is like the insult comedian of the Web |
| 04:24 | <MikeSmith> | without the comedy |
| 04:24 | <Hixie> | a lot of what Zed pointed out are perfectly legitimate problems |
| 04:24 | <MikeSmith> | bingo |
| 04:24 | <MikeSmith> | same with Crock |
| 04:24 | <Hixie> | his historical and contextual knowledge is somewhat lacking |
| 04:24 | <Hixie> | but that doesn't make his complaints any less valid |
| 04:24 | <Hixie> | however, we all know the web sucks |
| 04:25 | <MikeSmith> | right |
| 04:25 | <Hixie> | we have hundreds of bugs and thousands of e-mails still pending detailing precisely how it sucks |
| 04:25 | <MikeSmith> | it's when the start talking about the solutions that they go off into la-la land |
| 04:25 | <Hixie> | yeah |
| 04:25 | <Hixie> | i really wish someone would invent something to replace the web |
| 04:25 | <Hixie> | i just have no freaking idea what it could be |
| 04:26 | <MikeSmith> | I don't expect to see it happening any time during my watch on earth |
| 04:26 | <Hixie> | man, that means we're stuck with HTML for decades more |
| 04:27 | <MikeSmith> | yeah but I hope we eventually move on to working on something funner and leave the rest of HTML for some less jaded youths to pick up the torch on |
| 04:28 | <MikeSmith> | or really I just wanna run a nightclub like jwz |
| 04:28 | <MikeSmith> | or else just become a gardener or something |
| 04:31 | <MikeSmith> | at this point expecting to see the Web replaced by something else is like people in the 19th century when movies were first created expecting to right away see movies wholesale replaced by something else within the next few decades |
| 04:31 | <MikeSmith> | instead of course what happened is that movies didn't even have color until like 60 years later or whatever |
| 04:32 | <MikeSmith> | nor for 30 years or whatever did they even get sound |
| 04:32 | <MikeSmith> | etc |
| 04:33 | <MikeSmith> | and then there's where printed books started and how long they stayed the same |
| 04:33 | <wirepair> | this presentation |
| 04:33 | <wirepair> | is absolute gold. |
| 04:33 | <Hixie> | MikeSmith: now you're just bringing me down, man |
| 04:33 | <MikeSmith> | Hixie: sorry brotherman |
| 04:33 | <MikeSmith> | wirepair: yeah on many levels |
| 04:33 | <Hixie> | MikeSmith: i want my paper book -> ebook reader transition to happen quicker than 200 years! |
| 04:33 | <Hixie> | or 500 |
| 04:33 | <Hixie> | or whatever it took with books |
| 04:33 | <MikeSmith> | Hixie: keep hope alive man |
| 04:34 | <MikeSmith> | wirepair: outside of the fact he makes a lot of valid points, there's got to be some clinical term for the particular sort of neurosis exhibited in his presentation style |
| 04:35 | <MikeSmith> | I say that as a fellow neurotic and sometime presenter |
| 04:36 | <MikeSmith> | but at least my presentation neurosis hopefully doesn't make people in my audience want to walk up and kick me in the teeth |
| 04:37 | <wirepair> | i'm sitting here, dieing laughing. |
| 04:37 | <wirepair> | times like this i'm glad i work from home :) |
| 04:39 | <MikeSmith> | wirepair: I forgot where you live. way over in Ota-ku or somewhere? |
| 04:40 | <Hixie> | (i feel it's important to note that while he makes many valid points, they are surrounded by factual errors) |
| 04:40 | <Hixie> | (wrapped in them, even, like by bubblewrap in a package) |
| 04:40 | <MikeSmith> | hey somebody should come up with a way to third-party annotate videos |
| 04:41 | <Hixie> | webvtt supports it, you just need a tool to write the webvtt file and a site to tie the video and the webvtt file together |
| 04:42 | <MikeSmith> | yeah, I was kidding. and there are services for doing it, have been for a long time now |
| 04:43 | <MikeSmith> | though in practice the ones I know of seem to pretty much end up being mostly used for the most batshit idiotic purposes imaginable |
| 04:43 | <MikeSmith> | e.g. Nico-Nico Douga |
| 04:45 | <MikeSmith> | http://nicovideoo.up.seesaa.net/image/niconama2.jpg http://guron.net/document/content/javascript/jspic/nicohojo3.jpg |
| 04:46 | <MikeSmith> | http://farm3.staticflickr.com/2176/2504517363_3153c87564.jpg |
| 04:47 | <Hixie> | is there a way to distinguish a pure JS {} object from an arbitrary host object? |
| 04:57 | <heycam> | Hixie, it's very unlikely a host object would return "[object Object]" from Object.prototype.toString.call(blah) |
| 04:57 | <heycam> | host objects may even be disallowed from returning that |
| 04:58 | <heycam> | > The value of the [[Class]] internal property of a host object may be any String value except one of "Arguments", "Array", "Boolean", "Date", "Error", "Function", "JSON", "Math", "Number", "Object", "RegExp", and "String". |
| 04:58 | <heycam> | so that should be a safe way to check for that |
| 05:15 | <Hixie> | heycam: k, thanks |
| 05:22 | <annevk> | MikeSmith: http://lists.w3.org/Archives/Public/www-dom/2012AprJun/0168.html looks like spam |
| 05:23 | <annevk> | MikeSmith: or potentially a legitimate question in the wrong forum with a lot of spammy links attached |
| 05:27 | <annevk> | Midsummer time |
| 05:28 | <MikeSmith> | annevk: hmm saw that message and thought it was legitimate question in the wrong forum but hadn't noticed the link spam |
| 05:28 | <MikeSmith> | pretty sure it's intentional spam |
| 05:29 | <MikeSmith> | will spam-flag it |
| 06:51 | <AryehGregor> | Ugh. I hope I never have to have anything to do with the CSSWG again. |
| 06:53 | <SimonSapin> | AryehGregor: why? |
| 06:55 | <AryehGregor> | Mostly because they think they have a right to tell me what to do, I guess. |
| 06:56 | <AryehGregor> | I guess it's a heck of a lot better than your average corporate environment, where everyone is thoroughly conditioned to do what they're told to do, however stupid. |
| 06:56 | <AryehGregor> | So coming from that perspective it's probably hard to see what the problem is. |
| 06:56 | <AryehGregor> | Unfortunately for me, I've only ever had jobs where I get to do whatever I want however I want to do it, so I don't tolerate it well. |
| 06:58 | <MikeSmith> | AryehGregor: who's trying to tell you what to do? |
| 06:59 | <MikeSmith> | oh I probably know the answer to that already |
| 06:59 | <MikeSmith> | but what are they trying to tell you to do? |
| 07:01 | <MikeSmith> | ah public-css-testsuite thread I guess |
| 07:01 | <AryehGregor> | Yes. |
| 07:01 | <AryehGregor> | There was some off-list discussion too. |
| 07:02 | <MikeSmith> | add != refs to all the things ? |
| 07:02 | <MikeSmith> | AryehGregor: yeah, I can imagine |
| 07:02 | <AryehGregor> | Well, that was the straw that broke the camel's back. |
| 07:02 | <AryehGregor> | By itself I wouldn't have minded. |
| 07:02 | <MikeSmith> | glad I'm not in your shoes there man |
| 07:02 | <MikeSmith> | AryehGregor: what else on top of that? |
| 07:02 | <AryehGregor> | The context is that I was in the process of rewriting all the tests to conform to the CSSWG's stylistic guidelines before they'd even look at them. |
| 07:02 | MikeSmith | nods |
| 07:02 | <AryehGregor> | Which has taken me a lot of hours and is totally pointless from my perspective. |
| 07:03 | <MikeSmith> | hear that |
| 07:03 | <AryehGregor> | The entire thing is set up just fine from the perspective of people who are paid by some conventional employer and told "go submit these tests and get back to me". |
| 07:03 | <AryehGregor> | Then it makes sense to force them to shoulder as much work as possible to accept their submissions. |
| 07:04 | <AryehGregor> | Not so much for me. |
| 07:04 | <MikeSmith> | yup |
| 07:04 | <MikeSmith> | they need to fix that |
| 07:04 | <MikeSmith> | well they need to fix a lot of things |
| 07:04 | <MikeSmith> | but let's start with that |
| 07:04 | <AryehGregor> | Apparently the public-css-testsuite people don't think that. |
| 07:05 | <MikeSmith> | well they don't own it |
| 07:05 | <MikeSmith> | they may think they do |
| 07:05 | <MikeSmith> | the people who own it are the people who show up and do the real work of writing tests |
| 07:05 | <AryehGregor> | TBH, the policy probably works well for them. They might lose my work, but they'll get more work out of Microsoft or Apple or Adobe or whatever, and there are more of them. |
| 07:05 | <AryehGregor> | MikeSmith, so you'd think, but apparently not. |
| 07:05 | <AryehGregor> | At least not if we want them approved. |
| 07:05 | <MikeSmith> | I will talk to plh about this |
| 07:06 | <MikeSmith> | fubar |
| 07:06 | <MikeSmith> | this situation |
| 07:06 | <MikeSmith> | AryehGregor: the problem is, the craziness spreads |
| 07:07 | <MikeSmith> | like, now we have all this metadata overhead added to testharness.js, everywhere |
| 07:47 | <odinho> | MikeSmith: Yeah, I'm not really happy about that there. Metadata overload2k. |
| 07:47 | <odinho> | What about *first* making it useful, and then do some crazy-metadata scheme to support it. |
| 07:48 | <MikeSmith> | odinho: hey that would be novel |
| 07:48 | <MikeSmith> | but no let's instead over-engineer everything up front, for everybody |
| 07:53 | <odinho> | Yea. But if I'm not supposed to be grumpy, I guess connecting tests to the different parts of the spec allows for some cool stuff to be done. |
| 07:53 | <odinho> | And I want something like that now. Maybe I should write a small tool to do it and then test how metadata could fit in there. |
| 07:53 | <jgraham> | AryehGregor: I'm pretty sure that the requirements of the CSS WG prevent Opera from submitting more tests, but I can't really prove it |
| 07:54 | <AryehGregor> | jgraham, what do you mean? |
| 07:54 | <jgraham> | Well I'm pretty sure that the bigger the delta of work between making the tests for internal use and getting them accepted into the WG is, the less likely we are to actually do the delta |
| 07:55 | <MikeSmith> | jgraham: plh needs to hear that |
| 07:55 | <MikeSmith> | really |
| 07:55 | <MikeSmith> | from somebody other than me |
| 07:55 | <MikeSmith> | directly |
| 07:55 | <jgraham> | It's not like we're sitting around looking for work to do |
| 07:56 | <MikeSmith> | indeed |
| 07:56 | <MikeSmith> | and it's also about opportunity costs |
| 07:57 | <MikeSmith> | the time you spend jumping through all their hoops you could be writing other tests |
| 07:57 | <MikeSmith> | actually you know if I took a really mercenary view of things I would encourage them to keep it up |
| 07:58 | <MikeSmith> | keep making it even more difficult for people to contribute CSS tests to them |
| 07:58 | <MikeSmith> | then all of you will instead spend the time writing tests for WebApps specs or HTML |
| 07:58 | <MikeSmith> | which would be much better for me personally |
| 07:58 | <MikeSmith> | as far as what I'm expected to help get done |
| 07:58 | <jgraham> | Heh |
| 07:59 | <jgraham> | Doesn't quite work like that either |
| 07:59 | <MikeSmith> | yeah I know |
| 07:59 | <MikeSmith> | anyway I am going to be talking with plh today about this |
| 07:59 | <MikeSmith> | but it would really help for him to hear it firsthand as well |
| 08:00 | <MikeSmith> | he gets a lot of stuff from me already |
| 08:10 | <jgraham> | Anyway, I should stress that I don't really have solid evidence and I am not someone that writes lots of CSS tests |
| 08:11 | <jgraham> | But I know even the smaller overhead that exists for WebApps / HTML tests can cause some things not to get released or to take a while to release |
| 08:12 | <jgraham> | (some of this overhead is internal) |
| 08:13 | <MikeSmith> | jgraham: sure |
| 08:16 | <AryehGregor> | The CSSWG's requirements aren't *that* bad, if you're writing the tests for them a priori. |
| 08:16 | <AryehGregor> | A lot of the problem in my case was that I had a bunch of preexisting ones from Mozilla's codebase, and then I wrote lots of others without paying attention to the CSS format guidelines (which I knew existed). |
| 08:16 | <AryehGregor> | And then I was asked to port hundreds of them before any would be considered. |
| 08:24 | <MikeSmith> | AryehGregor: but I think that is the common case, actually |
| 08:24 | <MikeSmith> | what you describe |
| 08:24 | <AryehGregor> | Yes, probably that is quite common. |
| 08:24 | <MikeSmith> | at least it is one of the common cases |
| 08:24 | <MikeSmith> | yeah |
| 08:25 | <AryehGregor> | I wouldn't object to saying that the tests have to be formatted a certain way, but if other people want that, they should do it themselves. Or accept them with the understanding that someone will reformat them later. |
| 08:25 | <AryehGregor> | If they care, maybe they can write scripts to help. |
| 08:25 | <MikeSmith> | yes |
| 08:25 | <MikeSmith> | yes to that too |
| 08:26 | <MikeSmith> | with all the coding they already do of test infrastructure stuff, they are quite capable of doing that kind of scripting |
| 08:26 | <AryehGregor> | But when I've spent a lot of time making sure I understand the spec on a technical level and figuring out in depth how implementations behave and so on, it's extremely aggravating to be told that nobody's going to look at my tests unless I spend hours mindlessly reformatting them. |
| 08:27 | <MikeSmith> | any sane person would be frustrated the same way |
| 08:27 | <AryehGregor> | The stuff that actually required knowledge of what the test was doing, like adding <meta name=assert> explaining what the test did, I had no problem doing at all. |
| 08:27 | <MikeSmith> | and will be |
| 08:31 | <Ms2ger> | MikeSmith, maybe that is the common case because it's so annoying to submit pre-existing tests... |
| 08:38 | <MikeSmith> | Ms2ger: I meant the second case AryehGregor described |
| 08:38 | <AryehGregor> | It's the common case because vendors write lots of tests for their own internal use. |
| 08:38 | <AryehGregor> | And they aren't formatted to CSSWG guidelines. |
| 08:38 | <MikeSmith> | I mean I think it's far less common to have orgs writing tests for W3C WGs a priori |
| 08:39 | <MikeSmith> | Ms2ger: what AryehGregor said |
| 08:39 | <MikeSmith> | and it's not just browser vendors |
| 08:39 | <Ms2ger> | Yeah |
| 08:39 | <MikeSmith> | Adobe is writing tests also for example |
| 08:39 | <AryehGregor> | They work on WebKit. |
| 08:39 | <MikeSmith> | yeah |
| 08:39 | <Ms2ger> | That's not the common case for tests in the CSS test suite, unfortunately |
| 08:39 | <AryehGregor> | The actual people doing it are WebKit hackers, I think. |
| 08:40 | <MikeSmith> | Ms2ger: indeed it is unfortunate if that's the case |
| 08:40 | <MikeSmith> | AryehGregor: OK |
| 08:40 | <AryehGregor> | FWIW, I wrote Range/editing/etc. tests for non-internal use and now Mozilla often uses them in lieu of their own internal tests. |
| 08:40 | <MikeSmith> | yeah that happens too I guess |
| 08:41 | <MikeSmith> | but certainly less common |
| 08:41 | <AryehGregor> | Lots of things are already covered by my editing tests, so many bugs that would normally require new tests instead just have expected fails changed to expected passes. |
| 08:41 | <AryehGregor> | That's really the ideal, IMO. |
| 08:41 | <AryehGregor> | Because then vendors are working together on the same tests. |
| 08:41 | <AryehGregor> | Which saves work and helps convergence. |
| 08:42 | <Ms2ger> | ^ |
| 08:42 | <MikeSmith> | yeah |
| 08:42 | <Ms2ger> | That's the main reason I worked on importing tests :) |
| 08:42 | MikeSmith | has to drop off for a few minutes now to re-locate to a cafe nearby and get some food. back on in bit |
| 08:43 | <AryehGregor> | Ms2ger, want to help WebKit import the tests too? :) |
| 08:44 | <Ms2ger> | I don't need to get patches landed in WebKit :) |
| 08:44 | <Ms2ger> | And if I do, I just make someone else write the test |
| 08:53 | <Ms2ger> | summary: So that assert_equals('\n', '\n') actually shows the difference. |
| 08:56 | <odinho> | Ms2ger: wat? |
| 08:56 | <odinho> | Oh, bash did a double take on me? |
| 08:56 | <Ms2ger> | A backslash got lost somewhere |
| 08:56 | <odinho> | lol |
| 08:56 | <jgraham> | In the long term everyone agrees that having common tests is good |
| 08:57 | <jgraham> | In the short term no one ever gets assigned the task "write a W3C testsuite for feature X" |
| 08:57 | <Ms2ger> | No? |
| 08:57 | <Ms2ger> | I mean, that's part of AryehGregor's job :) |
| 08:57 | <Ms2ger> | (AFAIK, haven't read the contract, obviously.) |
| 08:57 | <jgraham> | Really? Or is his job to write tests for Mozilla? |
| 09:00 | <Ms2ger> | https://twitter.com/philikon/status/214890984272691200/photo/1/large |
| 09:01 | <jgraham> | Sure, but that's specs work |
| 09:01 | <jgraham> | Anyway, maybe "no one ever" isn't quite true |
| 09:01 | <Ms2ger> | (Unrelated, I just saw it) |
| 09:01 | <jgraham> | But it isn't the common case |
| 09:04 | <jgraham> | even people who are quite involved in the standards process have said that they can't justify spending time converting their ad-hoc testing to something submittable |
| 09:12 | <MikeSmith> | right |
| 09:12 | <MikeSmith> | so we shouldn't be doing things that cause them to spend even *more* time |
| 09:13 | <MikeSmith> | unnecessary things |
| 09:14 | <MikeSmith> | Ms2ger: um as far as the comment on that image I really wonder if that's the actual effect in practice |
| 09:14 | <Ms2ger> | We can try :) |
| 09:15 | <MikeSmith> | it certainly doesn't jibe with the writing modes spec and lobbying for complex ruby and a number of other things |
| 09:16 | <MikeSmith> | which frankly seem more like personal crusades rather than efforts to represent Mozilla's best interests |
| 09:16 | <MikeSmith> | but oh well |
| 09:24 | <mpt> | fantasai has been fascinated by writing modes since before she joined Mozilla |
| 09:25 | <Ms2ger> | Oh hey, it's mpt |
| 09:27 | <mpt> | but I wouldn't call it a crusade so much as "well if browsers are going to present those languages anyway, let's at least do it the way they're supposed to be presented" |
| 09:34 | <MikeSmith> | mpt: yeah, fair enough |
| 09:34 | <MikeSmith> | I been in a foul mood today |
| 09:35 | <MikeSmith> | not looking forward to HTML WG charter lawyering |
| 10:09 | <Ms2ger> | MikeSmith, today? :) |
| 10:09 | <MikeSmith> | hah |
| 12:29 | <odinho> | Hmz... Microsofts IDB tests expect result==null when the spec says result should be undefined. |
| 12:29 | <odinho> | And here I go and "fix" the bug, only to check the spec and see that the test is wrong. Bah. |
| 12:32 | <AryehGregor> | odinho, undefined == null, so that's fine. Did you mean ===? :) |
| 12:33 | <AryehGregor> | Where does it require undefined, though? That's usually considered bad form unless it's a place where JS would normally put it. |
| 12:33 | odinho | averts eyes <o_o> |
| 12:33 | <AryehGregor> | E.g., functions should normally not return undefined unless they always do (i.e., they don't return anything). |
| 12:34 | <odinho> | I'm *totally* open to change it to null. But, well, it doesn't say that :] |
| 12:34 | <AryehGregor> | Do you have a spec link handy? |
| 12:34 | <odinho> | AryehGregor: I use a local mirror, so will find it now :] |
| 12:35 | <odinho> | http://dvcs.w3.org/hg/IndexedDB/raw-file/tip/Overview.html#widl-IDBFactory-deleteDatabase-IDBOpenDBRequest-DOMString-name |
| 12:35 | <odinho> | If the steps above are successful, the implementation must set the result of the request to undefined and fire a success event at the request. The event must implement the IDBVersionChangeEvent interface and have oldVersion set to database version and have the newVersion property set to null. |
| 12:36 | <AryehGregor> | How does one access the result? |
| 12:36 | <AryehGregor> | From JS. |
| 12:36 | <odinho> | AryehGregor: deleteDatabase("ohman").onsuccess = function(e) { assert(e.target.result === undefined) } |
| 12:37 | <AryehGregor> | What sort of object is e.target? Does it implement a WebIDL interface? |
| 12:37 | <odinho> | EventTarget? Yeaa, think that. |
| 12:37 | <AryehGregor> | Oh, blech, .result is defined as "any"? |
| 12:38 | <odinho> | readonly attribute any result; |
| 12:38 | <odinho> | ( http://dvcs.w3.org/hg/IndexedDB/raw-file/tip/Overview.html#idl-def-IDBRequest ) |
| 12:39 | <AryehGregor> | I'd say the spec should require it be set to null. |
| 12:40 | <AryehGregor> | Offhand. |
| 12:42 | <odinho> | AryehGregor: There are many places it sets itself to undefined though. AH! |
| 12:42 | <odinho> | AryehGregor: Nah, it should be undefined, because null is an actual value you can get in result. |
| 12:42 | <AryehGregor> | Oh! |
| 12:42 | <AryehGregor> | That's bad. |
| 12:42 | <AryehGregor> | Hmm. |
| 12:42 | <odinho> | That why it's always set to undefined in all the places. |
| 12:43 | <AryehGregor> | So that's why it's "any"? |
| 12:43 | <AryehGregor> | It can be any JS value, because the query can return any value? |
| 12:43 | <odinho> | AryehGregor: Yeah. |
| 12:43 | <AryehGregor> | Okay, that seems like a potentially reasonable justification. |
| 12:43 | <AryehGregor> | I don't see any better way to do it. |
| 12:43 | <odinho> | So, pestering MS is the route then. |
| 13:03 | <hsivonen> | oh look. finally a TPAC that won't require an entrance fee from Hixie. |
| 13:33 | <Ms2ger> | hsivonen, oh? |
| 14:00 | <Ms2ger> | matjas, hmm, http://www.relentless-coding.com/projects/jsdetox/samples is interesting |
| 14:01 | <matjas> | turns malware into readable code… makes you wonder what happens if you feed it testharness.js |
| 14:01 | <Ms2ger> | Zing |
| 14:01 | <matjas> | I’ll see myself out. |
| 14:02 | <Ms2ger> | I wonder how good it is at detecting side effects |
| 14:31 | <MikeSmith> | does anybody know if Windows 8 exposes to Web Apps running in IE any mechanism similar to Web Notifications? |
| 14:31 | <MikeSmith> | or if current Windows does? |
| 14:36 | <gsnedders> | MikeSmith: I believe not. |
| 14:36 | <MikeSmith> | ok |
| 14:36 | <MikeSmith> | I thought as much |
| 14:37 | <MikeSmith> | and I guess there are no third-party notification things like Growl for WIndows? |
| 14:37 | <MikeSmith> | oh nm |
| 14:37 | <MikeSmith> | because I guess that would require IE to be compiled with support for whatever that third-party thing is |
| 14:37 | <gsnedders> | Growl exists for Windows. :P |
| 14:37 | <MikeSmith> | oh |
| 14:38 | <MikeSmith> | but you have to build your app to use Growl, right? |
| 14:38 | <MikeSmith> | link it in, I mean |
| 14:38 | <gsnedders> | Yeah, indeed. |
| 14:38 | <MikeSmith> | or actually you have to even ship your app along with growl |
| 14:38 | <MikeSmith> | OK |
| 14:39 | <MikeSmith> | so unless there is some other kind of magic thing, that definitely would seem to rule out the possibility of having any kind of Notifications-like support for Web apps running in IE |
| 14:56 | <MikeSmith> | OK will be stepping away from keyboard for some hours. If anybody needs me, ping me at [tm] or dm me there or on twitter |
| 16:11 | <dglazkov> | good morning, Whatwg! |
| 16:12 | <Ms2ger> | Good day, mate |
| 16:16 | <dglazkov> | :) |
| 16:16 | <gsnedders> | So, per spec, it would seem Progress Events have .total giving Content-Length and .loaded giving the decoded length (which given deflate, for example, can be longer) |
| 19:12 | <Famic> | hi |
| 19:15 | <Famic> | /j ##twitter-bootstrap |
| 19:55 | <Hixie> | ok, time to write tests for detaching iframes |
| 19:56 | <Ms2ger> | I bet bz can tell you stories about that :) |
| 19:57 | <Hixie> | indeed, i have e-mails from him in the relevant folder already :-) |
| 19:58 | <Hixie> | hm, i thought chrome had removed support for this |
| 20:00 | <Hixie> | wtf, even gecko seems to suppor it |
| 20:01 | <Hixie> | http://www.hixie.ch/tests/adhoc/html/frames/iframes/detaching/001.html |
| 20:01 | <Hixie> | safari and opera act as i'd expect |
| 20:01 | <padenot> | │ Dashiva |
| 20:01 | <padenot> | 19:59:18 Hixie | indeed, i have e-mails from him in the relevant folder already :-) │ dcheng |
| 20:01 | <Hixie> | well, not opera, opera does some crazy stuff |
| 20:01 | <Hixie> | with a [Object object] object with one property .close() |
| 20:02 | <Ms2ger> | padenot? |
| 20:02 | <padenot> | Ms2ger: mouse problem |
| 20:02 | <Ms2ger> | I see |
| 21:48 | <Hixie> | i crack myself up |
| 21:49 | <Hixie> | there's a comment right at the top of the spec's section on "origin": |
| 21:49 | <Hixie> | <h3>Origin</h3> |
| 21:49 | <Hixie> | <!-- Hallowed are the Ori --> |
