01:08
<MikeSmith>
is there an online tool for checking that background colors conform to WCAG recommendations?
01:09
<Rik`>
MikeSmith: you mean for contrast ?
01:09
<MikeSmith>
Rik`: yeah
01:09
<Rik`>
http://snook.ca/technical/colour_contrast/colour.html
01:10
<Rik`>
this one gives you the HP and WCAG2 rules
01:10
<MikeSmith>
Rik`: cool, thanks
02:50
<nessy>
Hixie: you finished the WebSRT section!?
02:52
<nessy>
now I am curious how to do styling, such as background color and general font settings etc
02:55
<MikeSmith>
yoshiaki: if the dudes you're meeting with have any specific questions I can help with, lemme know
02:56
<MikeSmith>
not trying to be rude, but don't think I need to be there for the whole meeting
03:06
<MikeSmith>
http://code.google.com/p/android/issues/detail?id=1376#c4
03:06
<MikeSmith>
<DmitryBaranovsk> “We have chosen intentionally to not support SVG in the browser to save space; full SVG adds over 1M to the webkit image size.” No comments.
03:08
<boblet>
MikeSmith: one of the most bone-headed decisions they made
03:09
<MikeSmith>
a disappointing one, at the very least
03:10
<MikeSmith>
for better or worse, SVG has become a de facto piece of the Web platform
03:11
<boblet>
given current high-dpi mobile developments, not having SVG is downright annoying
03:12
<MikeSmith>
to see a browser team unilaterally choosing to omit key pieces of the Web platform from particular OSes/platforms is really disappointing
03:14
<boblet>
especially one that would have saved me a lot of work :/
03:15
<aho>
ff4 will finally support svg as img's src or as background-image
03:16
<aho>
opera did support that since ages and chrome/chromium also supports it for a while now
03:16
<MikeSmith>
boblet: I could somewhat understand disabling some things on non-smartphone devices that don't have adequate resources
03:17
<aho>
and then i can finally get rid of those stupid object tags
03:17
<MikeSmith>
boblet: but I wonder which devices, exactly, are the ones that they are shipping Android on that don't have adequate resources
03:17
<MikeSmith>
boblet: certainly I would think the HTC devices have adequate resources for it
03:18
<MikeSmith>
here's hoping for ChromeOS mobile
03:19
<Hixie>
nessy: still have to do the rendering rules for it, and the DOM API, and a few minor other things
03:20
<MikeSmith>
I can imagine the argument being made that if a user has a device that doesn't have enough resources for handling SVG, then it's unlikely the user trying to browse the Web on that device is going to have a very adequate user experience anyway
03:20
<nessy>
Hixie: have you thought about adding styling by reusing CSS?
03:20
nessy
doesn't remember if we had that discussion
03:20
<Hixie>
nessy: yeah it'll be CSS-based
03:21
<nessy>
through a link at the top or something? curious to see an example
03:21
<MikeSmith>
boblet: because such a device would likely like have adequate resources for handling AJAXy sites that make much use of Javascript
03:22
<boblet>
MikeSmith: like I said, boneheaded. bah
03:22
<boblet>
in the absence of foolip can anyone tell me how I’d add microdatajs to a page and generate “download this vcard/ical” links?
03:25
<MikeSmith>
boblet: yeah, blech.. like God does in Revelations 3:16 "So then because thou art lukewarm, and neither cold nor hot, I will spue thee out of my mouth"
03:25
<MikeSmith>
spue spue
03:26
<boblet>
MikeSmith: aah, i needed spue.js huh ;)
03:26
<boblet>
I see your spue and raise you bah humbug
03:26
<MikeSmith>
heh
03:27
<boblet>
for a nested microdata item, does the parent property have the value of all contained text, or only act as a pointer for the nested name-value pair?
03:27
<MikeSmith>
Hixie: fwiw, I added per-element "Microdata property value" subsections to my markup ref
03:27
<MikeSmith>
e.g., http://dev.w3.org/html5/markup/object.html#object-microdata
03:27
<Hixie>
i saw
03:27
<Hixie>
looks good
03:28
<Hixie>
dunno how you'll explain it to people who think microdata isn't in html :-)
03:28
<Hixie>
nessy: using the style sheets in the HTML file
03:28
<boblet>
e.g. for this code sample:
03:28
<boblet>
<p itemscope>The <span itemprop="band">Salter Cane</span> drummer is <span itemprop="members" itemscope><span itemprop="name">Jamie Freeman</span>.</span></p>
03:28
<nessy>
so no reference in the websrt file itself?
03:29
<Hixie>
nessy: if it takes off we can add something to the srt file, dunno if it's really necessary though
03:29
<boblet>
is members equal to Jamie Freeman, the property name with the value Jamie Freeman, or both?
03:29
<Hixie>
nessy: i'd imagine most of the styling will be for web authors trying to make their subs fit their site
03:29
<Hixie>
nessy: most of the time, subtitles on dvds and the like aren't styled at all
03:29
<nessy>
Hixie: makes for little reusability, I guess
03:30
<Hixie>
styles aren't often reused anyway
03:30
<nessy>
Hixie: yeah, that's the current state of tech - I was hoping we can get something more powerful for the web ;)
03:30
<Hixie>
well for the web we'll have css :-)
03:30
<nessy>
Hixie: and thus also for the media players
03:30
<nessy>
yeah :)
03:31
<nessy>
I'm thinking css features will increasingly be able to be rendered in media players
03:31
<nessy>
that's part of "taking off" I guess
03:42
<nessy>
Hixie: did you describe the full spec as at http://damowmow.com/temp/srtspec ?
03:43
<nessy>
ah, don't worry - I was confused - I wondered if you missed the D, L, T, S etc spec - but it's there
03:53
<yoshiaki>
MikeSmith, thank you for your kind!
03:53
<MikeSmith>
hai
03:56
<MikeSmith>
paul_irish: html5rocks.com could maybe use a feed
03:57
<MikeSmith>
for teh "Latest additions" part
04:14
<MikeSmith>
somebody may want to point out to Ben Schwarz on the whatwg mailing list that the value of the rel attribute is a space-separated list of values
04:14
<MikeSmith>
so his <a rel="prefetch next">Next page</a> example is valid
04:14
<MikeSmith>
http://dev.w3.org/html5/spec/links.html#attr-hyperlink-rel
05:25
<MikeSmith>
hmm, maybe also point out that "prefetch" is not allowed link type for <a> anyway...
06:28
<MikeSmith>
there's no way to selectively apply character formatting to individual words in a string of CSS generated text, right?
06:28
<MikeSmith>
I mean, in the value of a CSS "content: " property
06:29
<MikeSmith>
boblet: ↑ ?
06:29
<MikeSmith>
I want to prepend the text "The following example is invalid."
06:29
<MikeSmith>
and I want just the word "invalid" to be bold
07:44
Slaanesh
ponders MikeSmith's question
07:49
<Slaanesh>
I think the answer was that there's no ::first-word
07:49
<Slaanesh>
And no other selectors on the word level either
08:15
<FastJack>
heh, and no nth-word? ;)
09:30
<boblet>
mikesmith: you want css3-content with p:before {content:"invalid";font-weight:bold;} p:before(2) {content:"This is ";}. unfortunately browsers do not love you.
09:32
<asmodai>
hsivonen: Is that 3.6.7 a beta or not?
09:32
<boblet>
mikesmith: you could add 2nd :before on a nested or parent element if there’s a convenient one, otherwise the super-fugly hack would be p::after {content:"This is ";float:left;margin-right:.3em;}
09:35
<boblet>
mikesmith: check chat log — just replied
09:35
<MikeSmith>
k
09:56
<boblet>
bbl
09:56
<boblet>
oh anyone know when foolip will be around?
09:57
<boblet>
(guessing he’s off enjoying RL or something atm)
10:00
<nielsle>
If browsers get full screen html5 capability, how easy will it be to create a page that looks like a login screen and steal peoples passwords?
10:02
<kennyluck>
lol
10:03
<annevk>
nielsle, there will be keyboard restrictions
10:07
<nielsle>
So in order to make a fully canvas based solution then I need to write some javascript that reads which key people are pressing at a given type, and write the corresponding "*" in the correct area of the canvas.
10:08
<nielsle>
type -> time
10:08
<annevk>
well you cannot, because there are keyboard restrictions
10:09
<nessy>
like going into TV remote control mode? ;)
10:15
<MikeSmith>
nielsle: there's recently been on a long discussion on the whatwg mailing list about how to make full-screen video secure
10:15
<MikeSmith>
dunno if you saw that yet or not
10:17
<nielsle>
I am googling for it :)
10:24
<nielsle>
Ah, here it is. Thank you for the answers. I feel safer on the internet now. http://thread.gmane.org/gmane.org.w3c.whatwg.discuss/24221/focus=24271
11:12
<Philip`>
jgraham: http://news.bbc.co.uk/1/hi/entertainment_and_arts/10507286.stm may be relevant to your interests
11:13
<Philip`>
(if you haven't seen it already)
12:37
<gsnedders>
jgraham: We have Jeppe to replace you, FYI.
12:37
<gsnedders>
(for fika)
12:58
<remysharp>
I might be missing something here, but any ideas every browser renders a canvas 1px width line as 2px? example: http://jsbin.com/apeti3/3
12:58
<remysharp>
(sorry - I said every - I've not checked IE9pre3 yet)
13:00
<Peter`>
Looks like the browser is trying to render it between the two pixels and thus applies anti-aliasing
13:03
<annevk>
yeah, iirc lines are drawn from the middle point
13:03
<Peter`>
If you move the line to 10.5x10.5 instead of 10x10 it's perfectly sharp
13:03
<annevk>
or you could use rectangles, though that may be slower
13:04
<remysharp>
Peter`: blimey - who knew!
13:04
<remysharp>
cheers for the tip
13:06
<Peter`>
No problem :)
13:07
<annevk>
the spec prolly ought to define it
13:11
<Peter`>
11 replies on Twitter, two here, hmmmm
13:12
<remysharp>
Peter`: I'm never sure how quick replies come in on #whatwg :-)
13:12
<remysharp>
it makes for a good backup :)
13:14
<Peter`>
Indeed :-)
13:26
<boblet>
remysharp: IRC is good for those of us without thousands of groupies :P
13:28
<remysharp>
boblet: groupies help ;-)
13:29
<boblet>
remysharp: lucky for some — none of the bots and spammers following me help much :)
13:33
<annevk>
remysharp, http://lists.w3.org/Archives/Public/www-style/2010Jul/0040.html
13:34
<remysharp>
annevk: not mine I'm afraid to say (if that was your intent)
13:35
<annevk>
oh, I thought you were behind html5rocks
13:35
<remysharp>
annevk: nah, html5demos - hence the ugly grey :)
13:35
<boblet>
annevk: that’s Chrome dev team, paul_irish is one contact
13:35
<annevk>
ah okay
13:35
<remysharp>
Paul Irisi might be a starting point
13:35
<annevk>
it doesn't really matter
13:37
<boblet>
annevk: it’s all in svn and you can add patches to comments in the Google Code repo
13:38
<boblet>
and file bugs. there was already a bug for transition examples being webkit only, but I already submitted patches for most of them
13:38
<annevk>
good good; it seems we need to change the specification and Opera to match WebKit
13:40
<boblet>
annevk: http://code.google.com/p/html5-slides/issues/detail?id=20&can=1&q=css
13:40
<boblet>
your issue is different tho so file a new issue
13:44
<Philip`>
remysharp: http://diveintohtml5.org/canvas.html#pixel-madness
13:44
<remysharp>
Philip`: damn that man, I didn't cover any of that madness in Bruce and my book!
13:44
<remysharp>
well - at least it's a bit of proof that we didn't read his book first :D
13:45
<boblet>
tehehe
13:52
<Philip`>
remysharp: If I remember correctly, he didn't cover it either, until I pointed out why the lines in his examples were all fuzzy :-)
13:52
<boblet>
the advantage of writing in the open :)
13:52
<boblet>
(not that I am)
13:52
<remysharp>
Philip`: absolutely the advantage of an online book. Bruce and I wanted to something similar, but our publishers weren't game :(
13:54
<annevk>
can't you just write and then find a publisher later?
13:54
Philip`
would be happy to review offline books too, in exchange for vast quantities of money :-p
13:55
<boblet>
remysharp: I wanted to do the book as a website then generate the print copy via print styles. naive fool that I was
13:55
<boblet>
(that’s why oli.jp is quite reminiscent of print)
13:56
<Philip`>
Would publishers prefer you to do the book as a website then copy-and-paste into Word before they edit and print it?
13:56
<boblet>
Philip`: would you like to read about microdata? ;)
13:57
<Philip`>
boblet: Unfortunately I know very little about that
13:57
zcorpan_
is happy to review html5 books
13:57
<boblet>
Philip`: publishers prefer you use their shitty-ass MS Word templates right from the start
13:58
boblet
was rather depressed to discover his website had better typography than a professional publisher did
14:01
<boblet>
I actually wanted to have an online version of the book be part of the purchase price
14:03
<boblet>
publishers are just as prehistoric as record labels and movie studios sadly (well apart from a couple of indy exceptions which already have HTML5 books coming out)
14:03
<Philip`>
Do people still make books with CDs inside the back cover?
14:03
<Philip`>
You could put the web site content onto that
14:04
<Philip`>
(That would be entirely missing the point of the web, but it'd be better than nothing)
14:06
<boblet>
Philip`: nah they have code snippet sites. but it’d be much better to have a website that you could then store locally, copy&paste from, access on a bunch of devices etc
14:06
<boblet>
(as in a website with the full content of the book)
14:07
<annevk>
so why seek out a publisher that prevents that?
14:07
annevk
still doesn't quite get it
14:07
<annevk>
is there more money to be made or so if you give in freedom?
14:28
<boblet>
annevk: in my case I’m not famous, so it’s selling my soul for the cred that comes with being published
14:28
<boblet>
sad but true
14:29
<boblet>
I suspect the current model will be in crisis in a couple of years as Five Simple Steps and A Book Apart start to steal all the good authors
14:30
<boblet>
and other indy publishers, self-publishing etc
14:30
<boblet>
so re: microdata “Properties can also have values that are URLs. This is achieved using the a element and its href attribute, the img element and its src attribute, or other elements that link to or embed external resources”
14:30
<boblet>
why doesn’t mpilgrim have script in his list?
14:31
<boblet>
http://diveintohtml5.org/extensibility.html#property-values
14:39
<zcorpan_>
because script is not listed in http://www.whatwg.org/specs/web-apps/current-work/complete/microdata.html#values
14:40
<zcorpan_>
"The URL property elements are the a, area, audio, embed, iframe, img, link, object, source, and video elements."
14:40
<boblet>
zcorpan_: aah, I didn’t check the encoding part. d’oh
14:40
<boblet>
thanks
14:49
<annevk>
boblet, there's just as much cred for publishing online, no?
14:50
<annevk>
boblet, if not more as it makes it much easier to see if your stuff is actually legit :)
14:51
<boblet>
annevk: I think you need to have some level of network first. there’s also the … monetisation issue
14:52
<boblet>
(not that publishing traditionally makes any money unless you have a bazillion reprints)
14:52
<boblet>
that’s what I have been doing too with my weblog and html5doctor :)
14:53
boblet
wonders if his lack of network is due to his stuff not being legit
15:01
<boblet>
(psst that was everyone’s chance to say “no, no! your stuff is great!” btw :P )
15:05
<Dashiva^2>
I think they stumbled on your use of 'legit'
15:07
<akamike>
Dr boblet your incisions are of the highest quality and i have never seen a man attempt 3 double bypasses at the same time
15:07
<akamike>
and succeed...
15:07
<annevk>
boblet, a book is different from a couple of blog posts, but maybe indeed you are not legit :p
15:07
<Workshiva>
As dual-classed ranger/ranger I can dual wield my dual wield
15:10
<boblet>
akamike: you are a scholar and a gentleman
15:10
<boblet>
(the rest of you rabble on the other hand… :P )
15:10
<boblet>
hehe
15:11
<boblet>
here’s hoping I climb the mountain, indeed
15:11
<boblet>
bbl
15:16
<zcorpan_>
Philip`: i get a fraud warning in opera for philip.html5.org :)
15:18
<annevk>
wow
15:18
<annevk>
quite the bug
15:19
<Peter`>
remysharp: I think your Tron mmo uses an outdated version of WebSocket, it's missing the 'sec-websocket-origin' header
15:19
<remysharp>
Peter`: yeah, it doesn't work on Chrome 6
15:19
<remysharp>
:(
15:20
<Peter`>
Works fine on Firefox though
15:21
<remysharp>
yeah - flash shim ftw :)
15:21
<remysharp>
not mine - obviously someone way smarter :)
15:27
<Philip`>
zcorpan_: It'd be nice if the warning gave some details that made it possible to debug
15:27
<Philip`>
e.g. to determine if it's a false positive or not
18:05
<llrcombs>
hey, what's the rule for right-clicking on <object>s and <embed>s?
18:36
karlushi
wonders if webkit became stricter for cookies
18:36
<karlushi>
encoding
18:38
<Workshiva>
Dear lazy IRC, can I customize the time zone used for \t in a bash prompt?
18:41
<karlushi>
Workshiva, do you want the local time, or the TZ info?
18:41
<Philip`>
Workshiva: Do you want to change it independently of other applications?
18:42
<Workshiva>
I'd be fine with changing it in all applications, but I can't change the system timezone
18:43
<Philip`>
You could put TZ=Whatever in .bash_profile
18:44
<Philip`>
(where Whatever is something from /usr/share/zoneinfo)
18:45
<karlushi>
do not use \t and use date something like?: export PS1="[\$(date +%k:%M:%S)]> "
18:45
<Workshiva>
TZ seems to work
18:46
<Workshiva>
Thanks lazy IRC!
18:46
Philip`
thanks http://www.gentoo.org/doc/en/guide-localization.xml
19:17
karlushi
is discovering the webkit difference of handling cookies with commas http://cookietest.livelyconsulting.com/
19:32
karlushi
wonders if webkit has started to implement https://datatracker.ietf.org/doc/draft-ietf-httpstate-cookie/
19:37
<llrcombs>
hey, what's the rule for right-clicking on <object>s and <embed>s?
19:44
<annevk>
karlushi, I suspect so; given how abarth usually goes about things
19:50
<llrcombs>
assuming it's an empty <object> (say, <object id="nothing"></object>)
19:51
<llrcombs>
should the UA display its own context-menu
19:51
<llrcombs>
or should it ignore it (since no plugin is loaded)
19:56
<karlushi>
annevk, thanks
20:01
<karlushi>
http://assets.en.oreilly.com/1/event/44/TCP%20and%20the%20Lower%20Bound%20of%20Web%20Performance%20Presentation.pdf
20:02
<annevk>
llrcombs, it's up to the UA
20:02
<annevk>
llrcombs, though if the element is not displayed at all there's prolly no context menu for it :)
20:12
<llrcombs>
no, assuming it's displayed, but there's nothing loaded in it
20:17
<annevk>
nothing much really
20:17
<annevk>
semantically in that case it's like <span>
20:17
<annevk>
so you could do whatever you do for <span> I suppose
20:22
<llrcombs>
so, then, WebKit's doing it wrong
20:22
<llrcombs>
good!
20:22
<llrcombs>
good
20:22
<llrcombs>
I've already got a bug report in for that
20:23
<llrcombs>
just wanted to confirm that what I thought was the right way was, in fact, the right way
20:23
<llrcombs>
except, wouldn't it be a block-level element (e.g. a <div>), not a <span>?
20:23
<AryehGregor>
I'm pretty sure the spec doesn't say anything here.
20:24
<AryehGregor>
Behavior is undefined, as with most UI-related things.
20:24
<AryehGregor>
The spec doesn't assume that there's any such thing as a context menu on your platform (and on some platforms, indeed, there's not).
20:25
<llrcombs>
heh
20:25
<llrcombs>
true
20:26
<llrcombs>
and the spec isn't exactly crystal-clear on <embed>s and <object>s either
20:28
<AryehGregor>
Because their functionality is really "give control to some non-HTML program, quite possibly a binary blob controlled by an entirely different party from the browser implementer".
20:29
<llrcombs>
exactly
20:29
<llrcombs>
but the <object> tag was always a bad notation, imho
21:03
<annevk>
llrcombs, nah, <object> like <img> is inline during fallback
21:03
<annevk>
(not that it should matter for what kind of context menu you would get)
21:28
<AryehGregor>
Where is it specced how big iframes should be if they have no height or width?
21:28
AryehGregor
checks the rendering section
21:30
AryehGregor
doesn't see it
21:41
<zcorpan_>
i think it's http://www.whatwg.org/specs/web-apps/current-work/complete/rendering.html#embedded-content-2 and http://www.w3.org/TR/CSS2/visudet.html#inline-replaced-width
21:43
<zcorpan_>
or maybe it's not defined properly in html5
21:43
<AryehGregor>
It doesn't say what the intrinsic height and width are supposed to be, though?
21:44
<AryehGregor>
(also, I'm not seeing from CSS what the height is if there's no intrinsic height, width, or ratio)
21:47
<zcorpan_>
it says 150px height for block/inline-block/float, but iframe is inline in opera/webkit/firefox
21:47
<annevk>
CSS 2.1 defines that
21:47
<annevk>
ultimate fallback is 150/300
21:47
<annevk>
for replaced content with no intrinsic height/width
21:48
<AryehGregor>
Where does it say 150px height here? http://www.w3.org/TR/CSS2/visudet.html#inline-replaced-width
21:49
<zcorpan_>
an iframe pointing to an svg with intrinsic width/height still uses 300x150
21:49
<AryehGregor>
Yes, an iframe without seamless has to be defined to have no intrinsic width or height or ratio.
21:49
<AryehGregor>
It doesn't seem like that is defined anywhere.
21:49
<AryehGregor>
I'll file a bug, I guess.
21:50
<annevk>
AryehGregor, that's width
21:50
<annevk>
AryehGregor, it says 300 there
21:51
<AryehGregor>
Oh, oops.
21:51
<annevk>
AryehGregor, height is defined here http://www.w3.org/TR/CSS2/visudet.html#inline-replaced-height where it says 150
21:51
<AryehGregor>
I see, it has separate sections for height and width.
21:51
<annevk>
and has had for a decade or so :)
21:53
<zcorpan_>
oh that section covered inline as well. i obviously skipped the first few words in the heading
21:53
<zcorpan_>
anyway, those are still not accurate for iframe since browsers always use 300x150 as default
21:54
<annevk>
oh, you're right
21:54
<annevk>
I think I pointed that out once
21:54
<annevk>
that HTML5 should have default style rules for it
21:54
<annevk>
I wonder if I ever filed a bug on that or emailed...
21:55
<zcorpan_>
if it was before whatwg last call, it fell between the cracks
21:55
AryehGregor
just filed http://www.w3.org/Bugs/Public/show_bug.cgi?id=10091
21:57
<annevk>
meh
21:58
<annevk>
iframe { height:150px; width:300px } is needed
21:58
<annevk>
in the rendering section
21:58
<annevk>
presumably overridden when seamless is specified and the URL is same-origin
21:59
<AryehGregor>
Wouldn't that allow authors to override it with iframe { height: auto; width: auto }?
22:01
<annevk>
zcorpan_, hmm, actually, the problem seems more that iframe creates a "viewport" of its own and that viewport has no intrinsic height/width and therefore the CSS rules apply
22:02
<annevk>
zcorpan_, so the embedded content is not SVG, it's a viewport with SVG on it (something in that direction)
22:02
<llrcombs>
so seamless would do the same thing as iFrames do on iPhone (size themselves to fit their content)?
22:02
<AryehGregor>
Do iframes really do that on the iPhone?
22:03
<llrcombs>
yep
22:03
llrcombs
is annoyed at LT (local high school)
22:03
<llrcombs>
they have an HTML course
22:03
<llrcombs>
but they have a rule
22:03
<zcorpan_>
also cross-origin?
22:03
<llrcombs>
yes, iirc
22:03
<llrcombs>
go try it!
22:03
<zcorpan_>
i don't have an iphone
22:04
<llrcombs>
that rule is: all tags must be all-caps
22:04
<llrcombs>
.....
22:04
llrcombs
pulls up the iSim
22:04
<llrcombs>
I'm pretty sure that's what MobileSafari DOES though
22:05
llrcombs
tests on his virtual iPhone 4
22:07
<llrcombs>
yep
22:08
<llrcombs>
given style="width: 100; height: 100;", it resizes the width to the minimum height to display correctly
22:08
<llrcombs>
and the height to the height of the page
22:08
<llrcombs>
(correctly = without a scrollbar)
22:09
<AryehGregor>
Even cross-origin?
22:09
<AryehGregor>
That sounds like a security vulnerability.
22:09
<llrcombs>
?
22:09
<llrcombs>
all it adjusts is the width+height
22:09
<llrcombs>
I was testing from a data: url, but I remember the same thing happening in W3Schools' TryIt editor
22:10
<AryehGregor>
Width+height leaks info.
22:10
<AryehGregor>
Whether you're logged into the site, for instance.
22:10
<llrcombs>
it's as though the "seamless" attribute is applied to every iframe
22:10
<llrcombs>
ahh, no, it's not actually specified
22:10
<AryehGregor>
And possibly lots of other fun stuff.
22:10
<llrcombs>
not in CSS
22:10
<llrcombs>
it's just as though you'd applied seamless
22:10
<llrcombs>
no vernurability beyond what would be provided by actually adding the seamless attribute
22:12
<AryehGregor>
seamless doesn't work cross-origin.
22:12
<AryehGregor>
For exactly this reason.
22:12
<AryehGregor>
Because it would be a security vulnerability if it did. :)
22:12
<llrcombs>
and...
22:12
<llrcombs>
Google put the sidebar back
22:12
<llrcombs>
just did it
22:12
<llrcombs>
I googled "iPhone iFrame"
22:12
<llrcombs>
then added "scroll" and voila, the sidebar's back
22:13
<llrcombs>
they always make their changes right when I'm searching stuff
22:13
<llrcombs>
except logos
22:13
<zcorpan_>
a<!--test-->b in VLC shows "a>b"
22:14
<zcorpan_>
a<01:00,000>b also shows "a>b"
22:14
<llrcombs>
#VLC
22:15
<llrcombs>
erm
22:15
<llrcombs>
nvm
22:15
<llrcombs>
ahh
22:15
<llrcombs>
must be one on another server
22:15
<llrcombs>
file a bug report @ their site
22:15
<zcorpan_>
the purpose of my testing is for websrt
22:15
<llrcombs>
ooh boy websrt
22:16
llrcombs
can't wait for WK to add support for that
22:17
<zcorpan_>
it seems to ignore an unclosed <i> altogether and drops the whole cue if i use <I>
22:17
<Slaanesh>
zcorpan_: VLC is legendary for failing to parse subtitles properly
22:17
<zcorpan_>
Slaanesh: you know any specific examples?
22:22
<erlehmann>
zcorpan_, i once had a subtitle file that was tuned to fail on bad VLC parsing
22:23
<zcorpan_>
erlehmann: you don't have it anymore or can find it again?
22:23
<erlehmann>
while other players coped well with it, VLC only had a continuous stream of YOUR MEDIA PLAYER SUCKS or something like that
22:23
<zcorpan_>
lol
22:23
<Slaanesh>
Yeah, a few groups would make subs like that
22:23
<erlehmann>
i can look for it, but i do not think i have it anymore.
22:23
<erlehmann>
wait.
22:23
<zcorpan_>
media player equivalent of browser sniffing
22:24
<erlehmann>
_██_
22:24
<erlehmann>
(°ᴗƪ)
22:24
<erlehmann>
hilarious insight, chap
22:25
zcorpan_
successfully used that smiley in srt in vlc
22:25
<Slaanesh>
zcorpan_: Not really, you can't sniff in subtitles
22:26
<Slaanesh>
More like media player equivalent of <blink>
22:26
<zcorpan_>
Slaanesh: you can expose bugs to block some media players
22:27
<Slaanesh>
Yeah, but that's just feature detection