| 10:02 | <Lachy_> | good morning everyone :-) |
| 10:04 | <annevk> | morning |
| 10:06 | <Lachy_> | any idea when Hixie will be returning? |
| 10:07 | <annevk> | returning from what? |
| 10:07 | <annevk> | (not really) |
| 10:28 | <Lachy> | well, I assume he's taking some sort of break, since I haven't seen him here for a few days. |
| 14:58 | <annevk> | https://bugzilla.mozilla.org/show_bug.cgi?id=392975 :) |
| 15:39 | gsnedders | posts a really short comment |
| 20:19 | <kig> | hahah, found my js port of the C impl in real-time fluid dynamics for games |
| 20:19 | <kig> | it's slooow |
| 20:20 | Philip` | is not surprised |
| 20:20 | <kig> | http://glimr.rubyforge.org/cake/tests/fluid.html and fluid_satan.html |
| 20:21 | <kig> | http://www.dgp.toronto.edu/people/stam/reality/Research/pdf/GDC03.pdf |
| 20:22 | <Philip`> | I ported a simple 3D polygon renderer from a toy language that compiled into JVM bytecode, to JS; it was orders of magnitude slower :-( |
| 20:30 | <gsnedders> | JS, |
| 20:30 | <gsnedders> | slow? |
| 20:32 | <Philip`> | Can you do fluid dynamics with GPU shaders? |
| 20:33 | <gsnedders> | In JS? No. But that's not what you're meaning. |
| 20:34 | <Philip`> | Running JS on a GPU would be slightly insane, I think |
| 20:39 | <kig> | blackout :< |
| 20:42 | Philip` | 's home server indicates there has not been a power outage for 267 days and 14 hours, which is quite surprisingly reliable |
| 20:43 | <gsnedders> | I think when I feared there had been a powercut elsewhere in town I think there was just thick fog, so I couldn't see lights anywhere else |
| 20:47 | <Philip`> | Hmm, it's kind of annoying that it costs at least $50 to join the more interesting X3D mailing lists |
| 20:52 | <Philip`> | It also seems a bit annoying when someone responds to questions with 'for more detailed answers, buy my book' (~£30 on Amazon Marketplace), though they gave fairly detailed answers on the (free) X3D list too so I won't complain :-) |
| 21:39 | <kig> | webkit's CSS -webkit-transform is awesome |
| 22:34 | <Philip`> | How many web-related specifications try to define graceful error recovery to a reasonable extent, rather than being fatal or leaving it undefined? (HTML5 and CSS are the ones that seem most obvious to me) |
| 22:37 | <Lachy_> | Philip`, XBL, XHR, Selectors API and presumably any other spec worked on by Hixie, Anne, myself and others who think like us |
| 22:40 | <Philip`> | Hmm, I can't see anything in Selectors API which authors must not do and that UAs must handle in a certain way |
| 22:47 | <Hixie> | Philip`: CSS was the first spec to really do it |
| 22:47 | <Hixie> | though that was just for forward compat, not for bug compat and interop |
| 22:47 | <Hixie> | so it didn't define all cases |
| 22:47 | <Hixie> | we tried to fix that in 2.1 |
| 22:55 | <gsnedders> | Hixie: 2.1 doesn't seem to have complete error handling when I last looked, though |
| 22:55 | <Hixie> | what got overlooked? |
| 22:55 | gsnedders | can't remember |
| 22:55 | <gsnedders> | I don't have the time to look right now |
| 22:55 | gsnedders | is half asleep |
| 23:37 | <Hixie> | it's very sad that despite not actually targetting IE at all, IE7 fails every test in acid3 so far |
| 23:38 | <Philip`> | Does it run the test framework correctly? |
| 23:38 | <Philip`> | (That was a significant problem for me with canvas tests in Safari and Firefox...) |
| 23:38 | <Hixie> | oh it doesn't fail _every_ test |
| 23:39 | <Hixie> | it passes 80 and 83 |
| 23:39 | <Hixie> | wait 83 isn't done yet |
| 23:39 | <Hixie> | so it passes exactly one test |
| 23:39 | <Hixie> | i wonder who fails it |
| 23:40 | <Hixie> | seems nobody fails 80 |
| 23:40 | <Hixie> | maybe i should change it |
| 23:40 | <gsnedders> | :D |
| 23:41 | <gsnedders> | the fun of acid tests… "wait, nobody fails this. better change this then…" |
| 23:42 | <Hixie> | oh hold on |
| 23:42 | <Hixie> | IE7 _does_ fail 80 |
| 23:42 | <Hixie> | it just fails it in a way that does something odd to the results |
| 23:45 | <takkaria> | Hixie: have the ie8 team talked to you at all about more acid tests, or are you just doing acid3 because its time has come? |
| 23:47 | <Hixie> | neither |
| 23:48 | <Hixie> | i have asked the IE team to suggest tests, but they have refused to do so |
| 23:48 | <Hixie> | and i've been working on acid3 for at least 6 months now, on and off |
| 23:49 | <G0k> | what kind of stuff does acid3 test? |
| 23:50 | <Hixie> | dom |
| 23:50 | <G0k> | does anyone come close to passing it? |
| 23:50 | <Hixie> | any suggestions are welcome, and should ideally come in the form of a 1-20 line JS function that returns true or false |