| 00:00 | <othermaciej> | Hixie: it's not unreachable (at least on Mac) |
| 00:00 | <othermaciej> | you can always cmd-click the selected item to deselect it |
| 00:03 | <Hixie> | othermaciej: wow, so you can |
| 00:03 | <Hixie> | i had no idea |
| 00:07 | othermaciej | thought it was obvious from the way cmd-click normally works in lists... |
| 00:16 | <TabAtkins> | Does it break some notion of semantics somewhere for a document to not have a location? |
| 00:17 | <othermaciej> | that question is too abstract for me to understand it |
| 00:17 | <TabAtkins> | The context is an intended-to-be-ephemeral document created by a page, with a location that is only valid until the creating page is trashed. I think you can navigate to that url fine (by the time the document is trashed and the url becomes meaningless, the navigation action is already moving). But what does document.location then return in the new doc? |
| 00:19 | <Hixie> | TabAtkins: is this a doc created by document.open(), createDocument(), XHR, window.open(), navigating to a File API URL, a Stream URL, or something else? |
| 00:19 | <TabAtkins> | Created by a FileWriter. |
| 00:20 | <TabAtkins> | Note: This isn't defined. We're asking what it *should* return, and if it's valid for it to have no location. |
| 00:20 | <Hixie> | wait, FileWriter returns a Document? |
| 00:20 | <Hixie> | you don't want to do that |
| 00:20 | <Hixie> | just have FileWriter use the same URL generation mechanism as the File API |
| 00:21 | <Hixie> | looks like the recent bugzilla update broke my automatic scripts, i wonder if the whatwg bug submission thing still works even |
| 00:21 | <Hixie> | oh christ it even broke my bookmarklets |
| 00:22 | <TabAtkins> | Hixie: That's already being used, but the url generated thusly is intended to be trashed/meaningless when the document that created it went away. |
| 00:22 | <Hixie> | bbiab, tgif |
| 00:22 | <TabAtkins> | So if you navigate the creating document to the url of the created document (assuming this makes sense), what's the location of the created document? |
| 00:25 | <othermaciej> | why would FileWriter return a Document if it's a writer, not a reader? |
| 00:54 | <JonathanNeal> | Is sitemap a valid role in html5? |
| 00:59 | <othermaciej> | JonathanNeal: is it a valid role in ARIA? if so, it's valid in HTML5 |
| 01:01 | <JonathanNeal> | Guess not http://www.w3.org/TR/2007/WD-aria-role-20071019/#roles |
| 01:02 | <JonathanNeal> | Or maybe this is more accurate / recent http://www.w3.org/TR/wai-aria/roles#role_definitions |
| 01:04 | <othermaciej> | JonathanNeal: I believe your second link is the most accurate |
| 01:04 | <othermaciej> | JonathanNeal: is the "navigation" role (or the HTML5 <nav> element) appropriate for what you are trying to do? |
| 01:05 | <JonathanNeal> | It is, but I get confused every now and again when I see an HTML5 blogpost use "breadcrumbs" or "sitemap" |
| 01:05 | <othermaciej> | as a role? |
| 01:06 | <JonathanNeal> | yes, usually next to "navigation" on the ul. |
| 01:07 | <JonathanNeal> | Which is weird anyway, because nav has "navigation" inherently. |
| 01:11 | <webben> | JonathanNeal: Don't think sitemap has ever been an ARIA role. |
| 01:11 | <JonathanNeal> | poor confused web designers out there. |
| 01:12 | <webben> | JonathanNeal: "breadcrumbs" used to be in the draft: http://www.w3.org/TR/2006/WD-aria-role-20060926/ |
| 01:14 | <JonathanNeal> | Yea ... I often try to convince folks that breadcrumbs shouldn't be a list. |
| 01:14 | <JonathanNeal> | Rather, a paragraph. |
| 01:15 | <JonathanNeal> | I think the spec has it that way. |
| 01:15 | <JonathanNeal> | And I saw it, thought it was odd, then thought about it, and thought it was neat. |
| 01:15 | <TabAtkins> | JonathanNeal: What's your reasoning? I do breadcrumbs as an <ol>. |
| 01:35 | <othermaciej> | JonathanNeal: seems like it is semantically a list |
| 03:08 | <JonathanNeal> | hi all |
| 03:10 | <JonathanNeal> | We had a big company meeting and it was really engaging, sorry for neglecting to reply to you guys, TabAtkins and othermaciej. |
| 03:10 | <othermaciej> | JonathanNeal: no worries, I wasn't saying anything important |
| 03:11 | <JonathanNeal> | I think breadcrumbs can be either. I've pushed for them to be paragraphs because I think they can work as paragraphs too. |
| 03:13 | <JonathanNeal> | I can see how they would be ol, because they're part of a hierarchical structure. |
| 03:13 | <JonathanNeal> | That's actually the reason I had us change from <ul>'s to <p>'s. |
| 03:14 | <JonathanNeal> | My initial understanding was tha tI didn't think of <ol> and <ul> as accurately portraying a breadcrumb in a heirarchical manner. |
| 03:14 | <othermaciej> | well |
| 03:14 | <othermaciej> | you can think of it two ways: |
| 03:14 | <othermaciej> | 1) a nesting hierarchy |
| 03:15 | <othermaciej> | 2) an ordered list of the user's path through the site |
| 03:15 | <othermaciej> | in a way they are logically equivalent, because a list is a degenerate tree |
| 03:16 | <JonathanNeal> | I might hear you describing two types of breadcrumbs, the hansel and gretel style where it targets your journey, or a branch style where you see are in relationship to the site hierchy (which is how I see it) |
| 03:17 | <JonathanNeal> | I like the later, as I've stated, and I felt it worked excellent with the idea and example @ http://www.whatwg.org/specs/web-apps/current-work/multipage/links.html#hierarchical-link-types |
| 03:44 | <TabAtkins_> | JonathanNeal: Yeah, no problem. Actually I use the latter style of breadcrumbs, where it shows the current path from the site root, but still believe <ol> is an appropriate tag for that. |
| 03:45 | <Hixie> | i think things are back in working order after the bugzilla upgrade |
| 03:45 | <Hixie> | but we'll see tomorrow |
| 03:46 | <shepazutoo> | Hixie: I think you pinged me last night when I was away? |
| 03:47 | <Hixie> | yeah, was going to ask you if you were aware of the plans for killing DOM Views, and if so, if you had plans for dealing with it in UIEvent |
| 03:47 | <Hixie> | i ended up just going with changing viewArg to dummyArg in the drag and drop event stuff, but i'll change it to whatever you end up deciding on |
| 03:48 | <Hixie> | gotta go right now though |
| 03:48 | <Hixie> | should be online tomorrow |
| 03:48 | <shepazutoo> | Hixie: ok, let's talk tomorrow |
| 03:48 | <Hixie> | k |
| 04:58 | <JonathanNeal> | whos up! |
| 05:28 | <zcorpan> | JonathanNeal: i accept patches (re http://krijnhoetmer.nl/irc-logs/whatwg/20100402#l-471 ) |
| 05:28 | <JonathanNeal> | reading |
| 05:28 | <JonathanNeal> | zcorpan, you're simon's html5 elements page? |
| 05:29 | <zcorpan> | yes |
| 05:29 | <JonathanNeal> | Then I will get to work immediately. |
| 05:29 | <zcorpan> | cool :) |
| 05:31 | <JonathanNeal> | Is there an even more multipage version of the html spec? |
| 05:31 | <JonathanNeal> | I want to make sure I'm not reinventing the wheel unnecessarily. |
| 05:31 | <zcorpan> | there's an index in the multipage spec |
| 05:32 | <zcorpan> | http://www.whatwg.org/specs/web-apps/current-work/multipage/section-index.html#index |
| 05:33 | <zcorpan> | i guess the table of contents also has all the elements listed |
| 05:34 | <zcorpan> | the toc is in the right order |
| 05:34 | <zcorpan> | i guess html5-elements' order is out of date |
| 05:34 | <JonathanNeal> | I just noticed that the html5 spec is written in the us language of hixie |
| 05:35 | <zcorpan> | it would be nice if i could generate html5-elements with a script that scrapes the spec |
| 05:36 | <JonathanNeal> | Well, it should be pretty easy with ajax. |
| 05:36 | <JonathanNeal> | It just means we'll need to host a local coy and fetch it via ajax. |
| 05:36 | <JonathanNeal> | Or curl it to proxy ajax. |
| 05:37 | <zcorpan> | generating it with javascript fails on the same-origin policy |
| 05:38 | <zcorpan> | could write a bookmarklet |
| 05:40 | <zcorpan> | actually a bookmarklet would not be a bad idea. would be easy to just appendChild()ing the right elements and getting innerHTML when you're done |
| 05:41 | <zcorpan> | the elements from the toc, and the attributes of each element from the index (loaded in an iframe) |
| 06:01 | <JonathanNeal> | zcorpan, i don't get, how does the bookmarklet work? |
| 06:01 | <JonathanNeal> | It runs locally so how can it get the content? |
| 06:02 | <zcorpan> | you run it when having the html5 spec loaded |
| 06:03 | <JonathanNeal> | oh ... |
| 06:03 | <JonathanNeal> | Maybe it would be better if I just copied the parts I wanted from the spec into separate files :P |
| 09:37 | zcorpan | adds wbr and device to html5-elements |
| 13:36 | gsnedders | wonders what a graph of number of songs owned per artist against number of plays per artist would look like |
| 13:37 | gsnedders | expects for him it'd be a line through the origin, more or less |
| 14:21 | <annevk> | zcorpan, can't you just prefix the current links with http://www.whatwg.org/html |
| 14:21 | <annevk> | zcorpan, the automatic redirects should take care of the rest |
| 15:29 | <JonathanNeal> | Goodmorning! |
| 15:43 | <annevk> | certainly feels like morning |
| 15:53 | <annevk> | shouldn't the spec explain how <wbr> is different from zero width whitespace? |
| 17:24 | <divya> | is it right to say that there is no stable release of a browser with HTML5 parser? |
| 17:25 | <lazni> | fx 3.6 has one, but disabled |
| 17:26 | <lazni> | by default |
| 17:26 | <divya> | lazni: true, but was wondering if there were anything else I did not know about. |
| 17:57 | <Dashiva> | Six different types of "conforming", I guess it'll be easy to adhere to at least one of them |
| 18:02 | <jgraham> | I think I am losing my mind |
| 18:02 | <Philip`> | Present tense? |
| 18:03 | <jgraham> | What happened to the bit of the websockects protocol spec where it talks about constructing the random tokens? |
| 18:03 | <jgraham> | Philip`: Add the word "further" if that helps |
| 18:04 | <Dashiva> | Whoa, Shelley is riding the deletionist wave like a pro... |
| 18:04 | <jgraham> | Why does WAHTWG not publish a version of websockts in a sane format (i.e. HTML) |
| 18:05 | <jgraham> | and without a multitude of different versions |
| 18:28 | <Dashiva> | Eating broccoli sure is fun :D |
| 19:42 | <Hixie> | jgraham: complete.html has the websockets protocol in it |
| 19:44 | gsnedders | wants a tiling window manager that actually behaves as he wants one to :P |
| 19:54 | <Hixie> | gsnedders: ion? |
| 20:03 | <Philip`> | gsnedders: Write your own |
| 20:03 | <gsnedders> | Philip`: That's my general feeling |
| 20:03 | <gsnedders> | Gah. |
| 20:03 | <Philip`> | based on xmonad |
| 20:04 | <gsnedders> | That means learning HAskell |
| 20:04 | <gsnedders> | *Haskell |
| 20:04 | <Philip`> | That makes it a doubly good option |
| 20:39 | <jgraham> | Hixie: Yeah I realised that eventually |
| 20:39 | <jgraham> | We could do with a multipage complete.html though |
| 20:40 | jgraham | likes multipage when he knows exactly what he is looking for and the section is relatively self contained |
| 20:44 | <jgraham> | gsnedders: Define "as you want one to" |
| 20:45 | jgraham | tried xmonad for a while but found that it was very inconvenient when you wanted some unusual layout |
| 20:45 | <jgraham> | (so prefers the relatively simple solution of using compiz+the grid plugin) |
| 20:46 | <jgraham> | (but would like it if the grid plugin had more tiling-window-manager like features) |
| 22:39 | <Hixie> | jgraham: i'd be up for that if someone sets up the multipage splitter in such a way that i can use it from the script |
| 22:40 | gsnedders | is somewhat scared that he can go into an IRC channel he's never been in before, and have said to him from someone he's never heard of, "Hey Geoffrey! Nice to see you here!" |
| 22:41 | <Hixie> | i know the feeling |
| 22:41 | <gsnedders> | I mean, since when was I that well known on the intartubes? |
| 22:41 | <Hixie> | i usually reply "um, hi" |
| 22:41 | <gsnedders> | Yeah, my normal response too |
| 22:42 | <gsnedders> | Obviously your influence is wearing off on me :P |
| 22:47 | <jgraham> | Philip`: Can you do whatever needs to be done with the multipage splitter? |
| 23:21 | <gsnedders> | <meta http-equiv="text/html" value="text/html;charset=utf-8"> works if I'm testing this correctly |