00:07
<Von_Davidicus>
If a browser supports SVG, is there a 100% chance (or at least close) that it will support XHTML + SVG?
00:07
<TabAtkins>
No.
00:12
<Von_Davidicus>
What is the likelihood, then?
00:13
<TabAtkins>
It's high, but that's not a useful number. You need to know the specific versions of browsers that didn't support inline SVG.
00:22
<Von_Davidicus>
Okay. *Is working on a way to let people know if their browser will support SVG+XHTML+CSS3*
01:27
<MikeSmith>
jzaefferer: you around?
01:34
<MikeSmith>
TabAtkins: Syntax component added
01:36
<MikeSmith>
annevk: http://dvcs.w3.org/hg/encoding/rev/94f075e5eabb seems to be working fine now
01:36
<MikeSmith>
I don't know what the problem was
03:39
<Von_Davidicus>
Managed to scare a computer tech today. I've been saying over the past couple of weeks that I needed a computer that was over 15 years old repaired. And today, I brought it in.
06:14
<zcorpan>
Hixie: text-wrap is supported in browsers (at least 'normal' and 'word-wrap')
06:16
<zcorpan>
though i think making white-space a shorthand for text-wrap might not be web-compatible (it'd make text-wrap reset if you do text-wrap:word-wrap; white-space:normal;)
06:17
<MikeSmith>
zcorpan: https://bitbucket.org/validator/validator/changeset/7dac5fb16df7
06:18
<MikeSmith>
hmm, isn't http://www.whatwg.org/specs/web-apps/current-work/multipage/#text-type-text-state-and-search-state-type-search supposed to redirect?
06:19
<MikeSmith>
dev.w3.org/html5/spec/#text-type-text-state-and-search-state-type-search does
06:19
<MikeSmith>
oh
06:20
<MikeSmith>
I guess that's because I added code on the w3c side to make it work
06:25
<zcorpan>
MikeSmith: LGTM
06:26
<MikeSmith>
thanks
06:54
<TomPenzer>
I was thinking about the problem of support for double-res images for devices like Apple's Retina display, and I was wondering what ppl think of adding support for a <meta img-x2-res="_x2" /> tag, which would alert the browser of the likely presence of 'filename_x2.ext' when it sees img tags, and it would fall back with a second request for 'filename.ext' in case it's not found. This seems
06:54
<TomPenzer>
relatively easy to implement (especially for web developers), and degrades gracefully on older browsers.
06:56
<Von_Davidicus>
what does "img-x2-res" stand for?
06:56
<TomPenzer>
img times 2 resolution
06:57
<TomPenzer>
or double-resolution
06:59
<Von_Davidicus>
Hm.
07:00
<TomPenzer>
as a web developer, i would be eternally grateful! : P
07:02
<TomPenzer>
i mean, besides everything these people have already done to make that even possible
07:05
<annevk>
Hixie: yes, I do want to define encoding sniffing
07:05
<annevk>
Hixie: presumably by defining what Gecko has, unless someone has a better idea
07:12
<TomPenzer>
a more fleshed out version of the idea is to have the meta tag include a property for 'assume-present', as such: <meta img-x2-res="_x2" assume-present="false" />, and then for any img tags for which double-res assets are available, you could add a property to the img tag as such: <img x2-res="true" />. Or, you could leave the 'assume-present' property off, as 'true' is default, and put <img
07:12
<TomPenzer>
x2-res="false" /> on any images for which double-res assets are unavailable. This would avoid a second request when the first fails.
07:13
<rniwa>
annevk: hi
07:13
<Von_Davidicus>
How about x2-res work like sort of like a boolean attribute? If it's not there, it's *assumed* to be false?
07:14
Von_Davidicus
rereads your post.
07:14
<TomPenzer>
yeah, if the meta tag isn't present, it's assumed to be false. The ="_x2" is to allow different developers to choose a double resolution filename key that works for their naming conventions
07:14
<Von_Davidicus>
Oh, okay.
07:15
<rniwa>
annevk: we might need to re-visit the decision to make NodeList inherit from Array
07:15
<Von_Davidicus>
I have a question of my own: Is it acceptable to include a test in a webpage that will determine whether or not a browser will properly display something? What I have after a hyperlink is an <object> element that displays an XHTML + SVG page that shows a checkmark, coloured green via HSL notation. If the viewer sees that the checkmark is black, or doesn't show up at all, then the viewer knows that the browser won't properly support the page on
07:15
<rniwa>
annevk: https://bugs.webkit.org/show_bug.cgi?id=81573#c36 indicates that making this change isn't compatible with the Web :(
07:18
<annevk>
sad
07:19
<annevk>
just file a bug I guess if it doesn't work out and we'll revert the spec change
07:20
<rniwa>
annevk: yeah... not a happy ending :(
07:20
<annevk>
Von_Davidicus: acceptable measured how?
07:20
<rniwa>
annevk: i'll double check with arv before filing a bug
07:20
<annevk>
rniwa: not entirely unexpected I guess that changing such a core API breaks the web
07:20
<rniwa>
annevk: right.
07:21
<annevk>
I'm curious what removing Attr nodes will do
07:21
<Von_Davidicus>
As in "considered a good idea by the webdeveloping community at large"
07:22
<annevk>
given that you can get on the final page directly, worrying about it is probably not worth it
07:23
<Von_Davidicus>
Hm?
07:27
<rniwa>
annevk: yeah, that'll be an interesting adventure.
07:44
<annevk>
oh cool
07:44
<annevk>
WebKit has a shared encoding sniffing library
07:44
<annevk>
http://trac.webkit.org/browser/trunk/Source/WebCore/loader/TextResourceDecoder.cpp
07:44
<annevk>
is that really all WebKit has?
07:44
<TomPenzer>
worst idea ever: -webkit-image-set( url(image.png) 1x, url(image⊙2p) 2x )
07:46
<annevk>
TomPenzer: why?
07:46
<annevk>
zcorpan: https://bugs.webkit.org/show_bug.cgi?id=18270 says that "x-user-defined" must mean windows-1252
07:46
<TomPenzer>
because all the 2x images on my site are going to have the same "_2x" or "@2x" filename key, and i shouldn't have to do all this busy work
07:48
<annevk>
zcorpan: also https://bugs.webkit.org/show_bug.cgi?id=15555 ugh
08:12
<annevk>
http://mxr.mozilla.org/mozilla-central/source/extensions/universalchardet/src/base/JISFreq.tab#338 so why include it?
08:12
<annevk>
I wonder how much of the universalchardet code is still relevant
08:13
<annevk>
e.g. there's code for iso-2022-cn in there, which IE does not even support
08:16
<annevk>
and Thai was not detected before 2009? https://bugzilla.mozilla.org/show_bug.cgi?id=488426
08:16
<annevk>
I guess I should start keeping notes somewhere
08:24
<Von_Davidicus>
If you wish to keep notes, I can suggest a few good music programs.
08:26
<annevk>
not sure if serious... or joking
08:27
<Von_Davidicus>
I was being ridiculous. :)
08:27
<Von_Davidicus>
I like to have a little pun with English now and then.
08:33
<annevk>
hsivonen: why did you limit utf-16 sniffing to HTML?
08:40
<annevk>
notes are tracked here http://wiki.whatwg.org/wiki/Encoding#Sniffing
10:34
<annevk>
what would also be interesting is Mozilla Telemetry data on universalchardet
10:35
<annevk>
but probably only if it's included in a non-beta release...
11:19
<annevk>
oh sweet
11:19
<annevk>
Fullscreen advocacy is already underway and only caters to WebKit/Gecko proprietary extensions
11:20
Ms2ger
sighs
11:20
<annevk>
thank your colleague http://www.slideshare.net/robnyman/javascript-apis-the-web-is-the-platform-mdn-hack-day-buenos-aires :p
11:25
<Ms2ger>
Boo
11:27
<annevk>
https://bugzilla.mozilla.org/show_bug.cgi?id=746437 "I can't think of a good reason to not implement it as-is."
11:27
<annevk>
that's good at least
13:40
<ruby_on_tails>
can anyone tell me why this is not working http://jsfiddle.net/uBXr2/1/ ?
13:43
<smaug____>
ruby_on_tails: onLoad
13:43
<smaug____>
it is onload
13:44
<ruby_on_tails>
oh
13:44
<ruby_on_tails>
it works now
13:44
<ruby_on_tails>
thanks
13:52
<hsivonen>
annevk: I wasn't needed for anything else and it was late in the release cycle
13:53
<hsivonen>
annevk: the whole thing might have been an overreaction from a single data point
14:10
<annevk>
hsivonen: Opera has something similar, but only checks the first four bytes
14:11
<ruby_on_tails>
smaug____: any idea why this doesnt grayscale the image ? http://jsfiddle.net/32f6h/
14:14
<ruby_on_tails>
weird, tried 3 codes from 3 different tutorial, none of them works, not sure whats wrong
14:16
<Philip`>
ruby_on_tails: You're not waiting for the image to finish loading, before you do the getImageData/putImageData code
14:17
<ruby_on_tails>
but there is the onload thing right ?
14:17
<Philip`>
so you're greyscaling the empty canvas and then (once the image has loaded, later) drawing the image
14:17
<ruby_on_tails>
i am trying http://spyrestudios.com/html5-canvas-image-effects-black-white/ and http://beej.us/blog/data/html5s-canvas-2-pixel/
14:17
<Philip`>
But it still won't work the other way around because if you drawImage an image from a different domain, then you're not allowed to call getImageData
14:18
<ruby_on_tails>
oh
14:18
<ruby_on_tails>
hmm
14:18
<Philip`>
to prevent people stealing image data from non-public sites
14:18
<ruby_on_tails>
let me try a local image then
14:18
<ruby_on_tails>
doesnt work on even a local image
14:19
<Philip`>
If you move the getImageData...putImageData code into the onload function, just after the drawImage, it should be closer to working
14:21
<ruby_on_tails>
bleh its my first day, i am not kind of getting it
14:24
<Philip`>
jsFiddle seems to already wrap your stuff in an onload thing, and runs it in some not-immediately-obvious domain
14:25
<Philip`>
http://software.hixie.ch/utilities/js/live-dom-viewer/?saved=1485 is a less confusing environment, and seems to work
14:29
<ruby_on_tails>
oh man
14:29
<ruby_on_tails>
i had to put the greyscaling code inside that onload
14:30
<ruby_on_tails>
it works now
14:30
<ruby_on_tails>
thanks Philip`
19:27
<shepazu>
once I've used pushState() to shove parameters up into the URI, what event can I use instead of 'load' to get older param states as the user walks back through the history?
19:38
<zewt>
popstate?
19:40
<shepazu>
zewt: thanks, tried it, didn't work, nor did hashchange… wonder if I'm using it wrong, or if maybe it doesn't work in SVG for some reason?
19:45
<shepazu>
ah, I think I was using it wrong
19:45
<shepazu>
:)
19:50
<shepazu>
oh, popstate fires before the URI is changed… odd
19:51
<shepazu>
inconvenient if you're using the URI to store state in parameters
19:52
<zewt>
uri where?
19:52
<zewt>
History.state is changed before popstate
19:54
<zewt>
(window.history.state)
19:55
<shepazu>
zewt: thanks… I was using location.href
19:56
<zewt>
(also popstate's event.state)
19:57
<zewt>
i'd expect that to be set before the events, too (don't know off-hand though)
19:58
<zewt>
http://www.whatwg.org/specs/web-apps/current-work/multipage/history.html#history-traversal not sure which step that is, but i'd guess 5
19:58
<shepazu>
zewt: I'd have expected that too… SVG in FF doesn't seem to happen, though
19:58
<zewt>
(no clue about svg)
21:47
<GPHemsley>
annevk: Is the goal of the encoding standard to intercept any non-UTF-8 encoding and output it as UTF-8? If not, would that be a good idea? (I'm not great with encodings.)
21:48
<gsnedders>
GPHemsley: given a definition of a decoder and a UTF-8 encoder you can do that
21:48
<GPHemsley>
gsnedders: Would that be a good thing to do? I'm just wondering
21:48
<gsnedders>
GPHemsley: The more important thing to define is how to get a sequence of Unicode codepoints from an encoding and how to get them to the encoding
21:49
<gsnedders>
GPHemsley: If you want to change the encoding of something, that's what you do. And the spec defines enough to do that. (As it defines all the decoders, and UTF-8 encoding is well-defined.)
21:49
<GPHemsley>
well presumably the user doesn't particularly care what the encoding is, as long as it looks like it's supposed to, right?
21:50
<GPHemsley>
and the encoding would only become relevant for copy and paste and the like
21:50
<GPHemsley>
(if we're talking only about browsers here)
21:51
<gsnedders>
Basically, all browsers have an internal Unicode representaiton, which is irrelevant what it is. (I believe everyone uses UTF-16.)
21:51
<gsnedders>
The important thing is how to decode the byte-stream, and how to encode to a byte-stream.
21:51
<gsnedders>
Then the internal representation is an impl detail which isn't black-box observable.
21:56
<GPHemsley>
gsnedders: What decides the byte-stream that the user actually sees, though?
21:57
<jgraham>
9doesn't gecko use not-utf-16 sometimes?)
21:57
<jgraham>
s/9/(/
22:03
<gsnedders>
GPHemsley: Typically the browser hands off the string as UTF-16 to the platform layer.
22:04
<GPHemsley>
ah
22:05
<gsnedders>
What happens in the platform layer I don't really know, I treat it as a black-box :)