| 00:17 | <alystair> | paul_irish: html5-boilerplate's .htaccess should have notes about servers utilizing Google Page Speed Apache module as it may severely interfere and/or confuse developers :) |
| 00:17 | <alystair> | http://code.google.com/speed/page-speed/ |
| 00:29 | <ap> | Hixie: why does <http://dev.w3.org/html5/websockets/>; link to <http://www.whatwg.org/specs/web-socket-protocol/>; in status section? is it just a mistake? |
| 00:42 | <Hixie> | ap: yeah |
| 00:43 | <ap> | Hixie: there goes my hope that we don't need SHA-1 |
| 00:43 | <Hixie> | hm? |
| 00:44 | <ap> | Hixie: nothing really. the IETF version feel awfully complicated to me, but I haven't been following the discussion to be qualified to judge it |
| 00:46 | <Hixie> | ah yeah. me either. |
| 00:46 | <Hixie> | websockets has pretty much convinced me that the ietf doesn't know how to write web standards |
| 00:47 | <Hixie> | live and learn |
| 00:50 | <Dashiva> | Isn't using SHA-1 a bit old hat? |
| 01:17 | <Dashiva> | I'm clearly staying up too late and delirious, because I could swear I saw a change proposal about longdesc |
| 01:18 | <Hixie> | wasn't that issue just reopened? |
| 01:19 | <Dashiva> | Seems so |
| 01:19 | <Dashiva> | Not that anything seems to have changed, except for the amount of text |
| 01:21 | <Hixie> | dunno, i'll have to read their new information |
| 01:22 | <Dashiva> | "Recent research finds that obsoleting longdesc specifically breaks the web for over 150 sites in the wild that are using it to describe images." |
| 01:22 | <Hixie> | how does it break the web for them? |
| 01:23 | <TabAtkins> | It sounds like they're treating "breaks the web" as a talisman without understanding what it means. |
| 01:23 | <Hixie> | also that seems like a really broken interpretation of the term "breaks the web" |
| 01:23 | <Hixie> | yeah |
| 01:23 | <Hixie> | what tab said |
| 01:23 | <Hixie> | anyway i need to read their data before i can form an opinion |
| 01:23 | <TabAtkins> | Specifically, it doesn't mean "makes non-conformant". |
| 01:24 | <Hixie> | i was thinking more that "break the web" means to break "the web", i.e. a large percentage of the web |
| 01:24 | <Hixie> | you know, at least like 0.2% or something |
| 01:24 | <Dashiva> | I'm not entirely against the idea that non-conforming is to some degree breaking, but I suspect almost every one of these sites is already non-validating for other reasons |
| 01:25 | <Hixie> | 150 sites is such a tiny number it's ridiculous |
| 01:25 | <Hixie> | i myself host 60 sites on my one shared server :-) |
| 01:25 | <Hixie> | Dashiva: "break" in the context of "break the web" means that the user experience would change |
| 01:26 | <TabAtkins> | No, I wouldn't support "make non-conforming" being "breaking" at all. "Breaking" means *breaking* - some functionality that the site was depending on for correct treatment no longer works. |
| 01:26 | <Hixie> | Dashiva: nobody is suggesting changing the user experience for anyone |
| 01:26 | <TabAtkins> | s/treatment/user experience/, exactly. |
| 01:26 | <Dashiva> | I know that myself, but I can see normal people thinking it means they can't use it at all |
| 01:27 | <Dashiva> | Even just "deprecated" seems to spell doom and failure for a lot of people |
| 01:27 | <Hixie> | not using it in new sites is not breaking the web |
| 01:29 | <TabAtkins> | Dashiva: We shouldn't be having "normal people" writing change proposals. They should be knowledgeable enough to understand how the web works. I wouldn't expect my mom to understand the distinction, but everyone in the HTMLWG should. |
| 01:30 | <Dashiva> | I know that too, but normal people will write websites |
| 01:30 | <TabAtkins> | Sure, and they shouldn't use @longdesc. |
| 01:30 | <TabAtkins> | Which obsoleting the attribute suggests. ^_^ |
| 01:30 | <Dashiva> | Every day I see people doing silly things in their markup because of some groundless fear of deprecation or validation |
| 01:41 | <zewt> | <annevk> Pretty good email filters then <- *cough* |
| 02:34 | <boblet> | is there a short URL like http://whatwg.org/C for HTML-Living Standard? |
| 02:37 | <nimbupani> | http://www.whatwg.org/specs/web-apps/current-work/multipage/ is the html5 spec in whatwg |
| 02:39 | <boblet> | nimbupani: HTML5++ you mean :p |
| 02:39 | <nimbupani> | :))) |
| 02:39 | <boblet> | aah there tis: http://whatwg.org/html |
| 02:39 | <nimbupani> | aha |
| 02:39 | <nimbupani> | neet |
| 02:39 | <nimbupani> | thnx boblet |
| 02:40 | <boblet> | nimbupani: haven’t you heard? there ain’t no 5 no more round these here parts |
| 02:40 | <nimbupani> | oops :| |
| 02:40 | <nimbupani> | MY BAD |
| 02:42 | <boblet> | hmm, maybe I should write something about that… ;) |
| 08:06 | <Hixie> | anyone remember why we dropped inputmode from wf2? |
| 08:09 | <Hixie> | aha, r2220 |
| 08:10 | <erlehmann> | why opera cannot into cross-origin? is there something wrong with CORS or have they just not bothered? |
| 08:41 | <annevk> | erlehmann, various people think something is wrong with CORS, but I guess you could say we have just not bothered so far |
| 08:41 | <erlehmann> | annevk, please bother :) |
| 08:41 | <erlehmann> | kthx |
| 08:41 | <annevk> | erlehmann, it's on the roadmap and I plan to continue fixing the spec :) |
| 08:42 | <erlehmann> | annevk, what *is* wrong with CORS? |
| 08:43 | <annevk> | well, some security researchers believe that the web security model ( http://tools.ietf.org/html/draft-abarth-principles-of-origin ) is wrong and that by entrenching it further with CORS we are making things worse |
| 08:43 | <annevk> | specifically, that people will design solutions with CORS that will open them up to become "confused deputies" |
| 08:43 | <Hixie> | MikeSmith: yt? |
| 08:45 | <erlehmann> | annevk, i am confused. this is duck. QUACK |
| 08:46 | <Hixie> | i don't understand http://www.w3.org/Bugs/Public/show_bug.cgi?id=11893 |
| 08:46 | <Hixie> | why is the html spec the place to define how aria works in svg? |
| 08:47 | <Hixie> | or if that's not the proposal, why would we want multiple definitions of how to implement aria? |
| 08:47 | <annevk> | oh hey, Doug Schepers objects to publishing DOM Core too now |
| 08:47 | <zewt> | <zewt> <annevk> Pretty good email filters then <- *cough* |
| 08:47 | <zewt> | heh |
| 08:48 | <zewt> | (he's who I was surprised at not saying anything the other day...) |
| 08:50 | <annevk> | oh, and he's also rewriting history claiming I'm the one who's doing it |
| 08:50 | <zewt> | his mail seemed not-so-subtly political, so, uh, have fun I guess |
| 08:50 | <annevk> | I wonder if that discussion was minuted |
| 08:57 | <karlcow> | annevk: your css doesn't make it easy to see what is a quote and what is a prose. Maybe a border-left would help |
| 08:57 | <karlcow> | re: your last blog post |
| 09:00 | <annevk> | hmm |
| 09:01 | <annevk> | you have a point, but I don't really feel like addressing it now |
| 09:01 | <annevk> | so I found the minutes and Doug Schepers is wrong |
| 09:01 | <annevk> | I wonder why he didn't check them himself |
| 09:02 | <annevk> | if you want to make broad claims like someone attempting to rewrite history you ought to check your facts |
| 09:12 | <Hixie> | hey annevk |
| 09:12 | <Hixie> | dom core doesn't define that children are ordered |
| 09:13 | <Hixie> | (so e.g. the definition of "preceding sibling" is unclear) |
| 09:13 | <annevk> | that doesn't follow from finite hierarchical tree? |
| 09:14 | <Hixie> | no |
| 09:14 | <Hixie> | you can have unordered trees |
| 09:14 | <Hixie> | (e.g. nested hash tables) |
| 09:15 | <Hixie> | best to just explicitly say that an object has an ordered list of children |
| 09:15 | <Hixie> | and define things in terms of that, imho |
| 09:15 | <annevk> | so " finite hierarchical tree structures" -> " finite hierarchical ordered tree structures" |
| 09:15 | <Hixie> | i still think it'd be better to be explicit in the definition |
| 09:16 | <annevk> | or "An object A is a child of an object B if and only if B is the parent of A." -> "An object A is a child of an object B if and only if B is the parent of A. The children of an object are ordered." |
| 09:16 | <annevk> | that makes it a statement of fact; not entirely correct hmm |
| 09:16 | <Hixie> | i think you're better off introducing an explicit child list, something like "an object in a tree has a parent and a child list" and then define everything in terms of that |
| 09:17 | <Hixie> | and have that instead of "has a number of properties that other objects do not" which is a bit weak |
| 09:17 | <annevk> | hmm yeah, that could work |
| 09:17 | <annevk> | thanks |
| 09:18 | <Hixie> | np |
| 09:18 | <Hixie> | i'll probably hook into this for webvtt |
| 09:31 | <Hixie> | right, bed time |
| 09:31 | <Hixie> | nn |
| 09:36 | <annevk> | https://bitbucket.org/ms2ger/dom-core/changeset/d57a85905b1c |
| 09:42 | <zewt> | annevk: "A is called a sibling of an object B, if and only if B and A share the same parent." should that say eg. "the same non-null parent"? |
| 09:44 | <MikeSmith> | annevk: I obviously did not get the Notifications draft pub-ready yesterday, but I am now getting it set up for publication on Tuesday |
| 09:45 | <annevk> | zewt, yes |
| 09:45 | <annevk> | MikeSmith, cool, I'll update the WG |
| 09:46 | <MikeSmith> | thnks |
| 09:48 | <zewt> | should it also specifically prohibit cycles? |
| 09:48 | <zewt> | (implicit from calling it a tree, but most of that section is) |
| 09:49 | <zewt> | well, I guess the point of that section is defining the terms, which the cycle restriction isn't relevant to |
| 09:50 | <annevk> | it might be fun to define tree from first principles and that is what I wanted to do initially, but I gave up on that |
| 09:50 | <annevk> | I would need to learn a bit more for that I think :) |
| 09:52 | <karlcow> | hmm Norway has a strange effect on my sleepiness level. Lack of light? |
| 09:53 | <annevk> | you should go there in summer |
| 09:53 | <annevk> | it's crazy |
| 09:53 | <annevk> | zewt, required an actual parent now :) |
| 09:53 | <zewt> | those poor lonely root nodes |
| 09:54 | <annevk> | I have not needed "root" so far |
| 09:54 | <annevk> | there's a note in the spec about nuking it |
| 09:55 | <zewt> | seems rare enough that just saying "if parent is null" when needed is good enough |
| 09:56 | <annevk> | I'll wait for hixie to redo HTML and see what comes out of that before getting rid of it |
| 11:26 | <annevk> | http://xkcd.com/865/ nice |
| 13:01 | <MikeSmith> | annevk: Respec-tified version of Notifications draft: |
| 13:01 | <MikeSmith> | http://dev.w3.org/2006/webapi/WebNotifications/ |
| 13:02 | <MikeSmith> | give it a few seconds to load |
| 13:02 | <MikeSmith> | there are some tweaks that still need to be done to it |
| 13:02 | <MikeSmith> | but I've done as much as I'm going to on it for today |
| 14:31 | <annevk> | MikeSmith, neat |
| 14:31 | <annevk> | MikeSmith, hopefully John is up for using that |
| 15:45 | <Hixie> | guess i'd better join webperf |
| 15:49 | <zcorpan> | http://www.w3.org/TR/FileAPI/#dfn-FileException - what should be the value of FileException.prototype.code ? |
| 15:51 | <annevk> | Hixie, yeah maybe |
| 15:52 | <annevk> | Hixie, re tests, that testharness is something people want to use for a bunch of tests; and I know 2xx needs updating :) |
| 15:52 | <annevk> | Hixie, I guess there could be more parsing tests too, feel free to add :) |
| 16:08 | <Hixie> | annevk: the harness and php scare me :-) |
| 16:08 | <Hixie> | annevk: fr(om writing tests) |
| 16:13 | <Ms2ger> | Hixie, jgraham would appreciate more concrete comments ;) |
| 16:20 | <Hixie> | Ms2ger: making the external file optional would be my main suggestion |
| 16:20 | <Hixie> | Ms2ger: see my commen in webapps |
| 16:20 | <Hixie> | bbiab |
| 16:21 | <Ms2ger> | That's a lot of parent.'s |
| 23:47 | <Hixie> | benschwarz: ping |
| 23:48 | <benschwarz> | pong |
| 23:48 | <benschwarz> | Whats up? |
| 23:48 | <benschwarz> | @Hixie |
| 23:53 | <Hixie> | hey |
| 23:53 | <Hixie> | sorry, been out for a few days |
| 23:53 | alystair | nods |
| 23:53 | <Hixie> | launch looks good |
| 23:53 | <Hixie> | you should do a blog post about it on the whatwg blog, get some more traction on it |