| 00:02 | <rektide> | postmessage is a useless pile of shit. :( |
| 00:02 | <rektide> | because i cant fucking talk to anyone. |
| 00:04 | <othermaciej> | the objection situation also made it to Hacker News http://news.ycombinator.com/item?id=1120394 |
| 00:04 | <othermaciej> | (more discussion there than on reddit it seems; maybe all the cool kids left proggit) |
| 00:20 | <rektide> | chrome you anger me |
| 00:21 | <rektide> | even after window.name = "foo", from another tab doing window.open(sameUrl,"foo",null,false) tries to open a new window |
| 00:21 | <rektide> | i use this as my final contention for postMessage() being utterly and completely worthless. unless you open it, you cant talk to it. |
| 00:32 | <Philip`> | rektide: It ought to work in any situation where you already have a reference to another window, I believe |
| 00:34 | <Philip`> | There's no guarantee that window names are unique within a browser session, so you can't use them as way to acquire references |
| 00:35 | <Philip`> | I guess you have to use something like localStorage if you want some shared communication medium between otherwise-independent windows |
| 00:38 | <rektide> | Philip`: theres no guarantee window names are unique, so you cant _rely_ on them to acquire references |
| 00:39 | <rektide> | however you should be able to acquire references to other windows, according to the window.open() spec |
| 00:40 | <rektide> | and according to my own testing, and google chrome bug 8751, chrome is currently in violation of the spec |
| 00:40 | <rektide> | http://code.google.com/p/chromium/issues/detail?id=8751 |
| 00:42 | <Philip`> | I think HTML5 allows Chrome's behaviour |
| 00:43 | <Philip`> | http://www.whatwg.org/specs/web-apps/current-work/multipage/browsers.html#dom-open - "the user agent must first select a browsing context to navigate by applying the rules for choosing a browsing context given a browsing context name using the target argument as the name" |
| 00:43 | <Philip`> | http://www.whatwg.org/specs/web-apps/current-work/multipage/browsers.html#the-rules-for-choosing-a-browsing-context-given-a-browsing-context-name |
| 00:43 | <Philip`> | "If the given browsing context name is not _blank and there exists a browsing context whose name is the same as the given browsing context name, and the current browsing context is allowed to navigate that browsing context, and the user agent determines that the two browsing contexts are related enough that it is ok if they reach each other, then that browsing context must be the chosen one." |
| 00:44 | <Philip`> | The "related enough" bit allows Chrome to consider browsing contexts associated with separate browser windows to be unrelated |
| 00:50 | <Philip`> | (Oh, that bug points at SharedWorker which is possibly a reasonable way to communicate between independent windows) |
| 00:52 | <rektide> | yeah i'm catching up on the SharedWorker work now |
| 00:56 | <rektide> | where is SharedWorkers spec being worked on? |
| 01:01 | <rektide> | are SharedWorkers implemented anywhere yet ? |
| 01:01 | <rektide> | it'd be nice if i didnt have to kludge together some monstrosity that polls the session store |
| 01:04 | <Philip`> | http://dev.w3.org/html5/workers/Overview.html |
| 01:04 | <Philip`> | http://www.whatwg.org/specs/web-workers/current-work/ |
| 01:05 | Philip` | has no idea about implementations |
| 01:07 | <othermaciej> | rektide: they are implemented in WebKit nightlies |
| 01:08 | <dimich> | rektide: SharedWorkers are enabled in chrome 4 |
| 01:10 | <rektide> | well, at least i have no excuse to stop coding |
| 01:10 | <rektide> | i still think is silly postMessage can only talk to child documents |
| 01:11 | <rektide> | but i can at least keep going |
| 02:43 | <Traveler--> | hello |
| 02:43 | <Traveler--> | Anybody online in here? |
| 05:51 | <TheOutlawTorn> | Good morning. |
| 09:53 | <wycats> | Hixie: ye be aboot? |
| 09:54 | <Hixie> | hey |
| 09:55 | <wycats> | I'm working on a project to create an OSS book toolchain and I got a (long) email from a guy in the Ruby community who suggests just using HTML5 |
| 09:55 | <wycats> | was curious if you consider such a thing an actual goal of HTML5 |
| 09:56 | <Hixie> | you're asking if HTML would be an adequate authoring format for a book? |
| 09:56 | <wycats> | yes |
| 09:56 | <wycats> | HTML4 is not |
| 09:57 | <wycats> | I had started to build my own markup and this guy (rightly) pointed out that perhaps HTML5 is already suitable |
| 09:58 | <Philip`> | Why not use something that's designed for that purpose? |
| 09:58 | <Philip`> | like LaTeX or DocBook or Word or whatever |
| 09:58 | <wycats> | Philip`: LaTeX is unpleasant to work with |
| 09:58 | <wycats> | Word is not suitable for version control |
| 10:00 | <Hixie> | I was about to say that I thought HTML4 was fine for writing a book, personally |
| 10:01 | <Hixie> | Hakon and Bert wrote their CSS book in HTML |
| 10:01 | <Hixie> | Mark is writing his HTML5 book in HTML |
| 10:03 | <wycats> | Hixie: there are a number of new HTML5 elements that are particularly helpful |
| 10:04 | <wycats> | the one thing I'm missing -- perhaps you can provide guidance -- is a way to say that some <code> is actually stored elsewhere |
| 10:04 | <wycats> | so I can stick my code in a VCS and have a build process pull them in |
| 10:04 | <Hixie> | i just use <!--INCLUDE ...path...--> |
| 10:04 | <Hixie> | or some such |
| 10:04 | <wycats> | Hixie: hm |
| 10:04 | <Hixie> | for example the HTML5 spec pulls in the examples from http://whatwg.org/demos/ |
| 10:05 | <wycats> | seems non-semantic :P |
| 10:05 | <Hixie> | (not all of them, but some) |
| 10:05 | <Hixie> | "semantic"? |
| 10:05 | <wycats> | Hixie: I was just being annoying for no reason |
| 10:05 | <Hixie> | heh |
| 10:05 | <wycats> | Hixie: there's no way to express it without comments? |
| 10:05 | <wycats> | perhaps an <a>? |
| 10:06 | <Hixie> | if it's an authoring format that you preprocess before you publish, you can express it however you like |
| 10:06 | <Hixie> | because it's effectively just your proprietary format |
| 10:06 | <wycats> | yep |
| 10:06 | <wycats> | I'm my own UA ;) |
| 10:08 | <wycats> | <pre data-ref=... data-file=...> or somesuch |
| 10:09 | <wycats> | I know! I can use RDFa ;) ;) |
| 10:11 | <Hixie> | RDFa and microdata wouldn't be good for this because they don't correspond to specific parts of the document |
| 10:12 | <Hixie> | data-* wouldn't be bad |
| 10:12 | <wycats> | Hixie: yeah |
| 10:12 | <wycats> | esp. in combo with the pre tag |
| 10:12 | <wycats> | which I could use for inline stuff if I didn't want to bother with linking to a file |
| 10:17 | <Philip`> | Use <xmp> :-) |
| 10:22 | <wycats> | <xmp>? |
| 10:22 | <wycats> | omg adobe |
| 11:02 | <Philip`> | wycats: It's just like <pre> but without having to escape the content |
| 11:02 | <Philip`> | Nothing to do with Adobe |
| 11:03 | <wycats> | Philip`: It's in HTML5? |
| 11:03 | <wycats> | I don't see it |
| 11:03 | <Philip`> | It's been deprecated for decades |
| 11:03 | <Hixie> | damn, mjs left |
| 11:03 | <wycats> | apparently deprecated since 3.2 |
| 11:03 | <wycats> | :P |
| 11:03 | <Philip`> | (but still works) |
| 11:04 | <wycats> | hm... it actually seems good |
| 11:04 | <wycats> | why was it deprecated? |
| 11:05 | <Philip`> | Maybe because you can't put the string "</xmp>" inside an xmp element, and so the HTML people preferred to go with the more powerful <pre> instead |
| 11:06 | <Philip`> | It's still kind of useful for quick tests and private documents and things, though :-) |
| 11:07 | <wycats> | Philip`: having to escape <pre>s makes them much less useful |
| 11:07 | <Philip`> | and much easier to remember than XML's <![CDATA[...]]> |
| 11:07 | <wycats> | because you can't copy/paste large chunks of code |
| 11:08 | <Philip`> | I suppose <pre> is okay if you have a text editor with an "escape selected text" feature |
| 11:08 | <Philip`> | which any decent HTML editing environment ought to have |
| 11:08 | <wycats> | hehe |
| 11:08 | <wycats> | sure |
| 11:08 | Philip` | has never had a decent HTML editing environment |
| 11:09 | <wycats> | you have to unescape to edit though |
| 11:09 | <wycats> | and it's less readable in raw form |
| 11:37 | <wycats> | Hixie: I'm going to bang out a bunch more comments over the weekend -- any luck on getting the comment form to open tickets as me? |
| 11:59 | <Hixie> | i made it so that if you have an account with the specification annotation system it'll mention your e-mail address and CC you on the bug |
| 13:51 | virtuelv | uses a couple of komodo macros to escape/unescape |
| 13:53 | <virtuelv> | the other convenient thing about it is that it's based on XUL, so I have a DOM I can wrangle |
| 14:59 | <s|k> | komodo macros? |
| 16:11 | <asmodai> | mitsuhiko: So what brings you here? :P |
| 23:48 | <mitsuhiko> | asmodai: html5lib :) |