03:57
<patx>
<video src="file:///home/patx/out.ogv"></video> Thats does not seem to work
03:57
<patx>
why is this?
04:08
<jwalden>
patx: what browser, and is the file containing that directly in /, /home/, or /home/patx/?
04:08
<patx>
/home/patx
04:08
<patx>
chrome
04:09
<jwalden>
hm
04:09
<jwalden>
I don't know chrome's local-file security model, but I'm pretty sure it should permit that
04:12
<patx>
doesnt work on firefox either
04:12
<jwalden>
are you sure the file isn't malformed?
04:12
<jwalden>
I *know* Firefox's security model is fine with a local HTML file accessing a sibling file
04:13
<jwalden>
what does file properties in nautilus say is the type of the file?
04:23
<patx>
now its working.
04:23
<patx>
hmm
07:15
<patx>
So I am having a little problem... http://patx.me/paste/6183.html
07:19
<Hixie>
patx: we'll prolly need to see the code to be able to help
07:19
<patx>
Hixie, the python code or the html5?
07:20
<Hixie>
the code the browsers get
07:20
<Hixie>
so html
07:21
<hsivonen>
patx: when on the local file system, the html file and the video file need to be in the same directory
07:21
<hsivonen>
patx: when on server, .ogv needs to be mapped to video/ogg
07:22
<patx>
oh ok
07:22
<patx>
hsivonen, that explains it!
07:22
<patx>
thanks a bunch
07:22
<patx>
theyre in diff directorys
07:22
<Hixie>
ah yeah that'll do it
07:23
<Hixie>
that's to protect you from downloading an HTML file that then grabs your system passwords or something
07:23
<patx>
oh ok!
07:23
<patx>
thanks
07:25
<patx>
Hixie - hsivonen : <video controls="controls" preload="metadata" src="file:///home/patx/Desktop/5496.ogv" width="800"></video>
07:25
<patx>
so i cant do that?
07:26
<Hixie>
generally speaking whether file:// works at all and how it works is really up to the browser -- but in most browsers, it'll only work if the video file is in the same directory as the HTML file, when you use file://
07:26
<Hixie>
easiest solution is to just say src="5496.ogv" instead of saying file://... etc.
07:26
<Hixie>
and just make sure they're in the same directory
07:40
<patx>
ok thanks
11:22
<jgraham>
Is there some special reason that web workers disallows workers in data: uris?
11:30
<Dashiva>
Seems to be r3857
11:31
<Dashiva>
Origin reasons, apparently
11:43
<jgraham>
Dashiva: Yeah, but what does banning data uris gain us?
11:44
gsnedders
wonders if jgraham ever stops working on HTML5-related stuff
11:44
<Dashiva>
Isn't it just that since data: URIs have unique origins, they just don't work? Disclaimer: I don't really know why.
11:45
<gsnedders>
Dashiva: pong, seeming you pinged me last week while I was away
11:47
<Dashiva>
I think the TTL expired long ago
11:48
<Dashiva>
I can't seem to find the original request, are you sure it was me?
11:48
<gsnedders>
Dashiva: Yes
12:22
<jgraham>
gsnedders: I really don't do that much useful stuff. I would really like to get more done
12:22
<jgraham>
OTOH, it is fun
12:22
<jgraham>
More fun than whining about it anyway
12:39
<annevk>
I fixed the issue on html5.org
12:39
<annevk>
not in a very pretty way
12:53
<hsivonen>
annevk: what was the root cause?
12:54
<annevk>
before when a diff was empty it would return an empty string
12:55
<hsivonen>
oh. an empty diff
12:55
<annevk>
apparently SVN changed and now something is at the top at all times leading my script to assume there was a diff and not finding a number and therefore throwing an error
12:56
<annevk>
so now I changed to a try / except where I try to find the number given a diff and if that fails I assume there was no diff
13:12
<hsivonen>
aargh. https://answers.edge.launchpad.net/launchpad/+faq/1024 is *so* annoying. I wish the Origin header had full deployment already
14:47
<annevk>
oh hey, Chrome nightlies support SVG referenced from the 'content' property
14:47
<annevk>
so they finally render my blog
15:46
<patx>
Hixie, hsivonen - i was talking to u last night... i put the vid in the same dir as the html... however i get the same result...
19:53
<jgraham>
Philip`: You on public-html-testsuite? We should import your canvas tests into the official testsuite
19:53
<jgraham>
So you can do it if you like, or I can do it if it is more convenient
19:57
<Philip`>
jgraham: I'm on the list
19:57
<Philip`>
(though I've been away for a week so I've missed anything that might have been said on it or elsewhere)
19:57
<Philip`>
I prefer solutions that don't require me to do any work
19:59
<Philip`>
though I also like solutions that let me update the tests by editing the source files instead of manually touching the HTML files, and I don't know how that'll interact with any importing of the tests
20:00
<jgraham>
Philip`: I think that is OK. I think we can say that the source is the canoncial version and everything else is just generated
20:00
<jgraham>
I hope so because we will have to do the same with the html5lib tests