00:58
<Cheery>
in constructing response the websocket spec says: "Send two bytes 0x0D 0x0A (ASCII CRLF)."
00:58
<Cheery>
but I don't see how this makes any sense
00:58
<Cheery>
shouldn't it be four bytes?
01:00
<Cheery>
or wait.. yeah it's correct
01:00
<Cheery>
each fields ends to CRLF, so of course I only need to emit one when I end the header.
01:28
<Cheery>
good. now I have everything needed for my first websocket app.
01:28
<Cheery>
except a working ubuntu-provided chromium. >:E
01:46
<MikeSmith>
hsivonen: http://lint.brihten.com/html/
01:48
<MikeSmith>
"Making sure your code looks good" .. options for "never omit closing tags", "quote attributes", etc.
02:08
<MikeSmith>
hmm, http://lint.brihten.com/html/ doesn't actually seems to work.. getting message about "Cannot continue: HTML is not valid (Line 5: Illegal closing tag </head> for <a.length;){var>)"
02:09
<MikeSmith>
ah geez
02:09
<MikeSmith>
"Cannot continue: HTML is not valid (Line 159: Illegal closing tag </script> for <").replace(/>/g,>)"
02:09
<MikeSmith>
seems to indicate it's trying to do some kind of regular-expression replacement
10:22
<Workshiva>
MikeSmith: Or it just doesn't tokenize script contents properly
11:09
<hsivonen>
MikeSmith: what parser is the lint based on?
13:53
<Workshiva>
I never realized how convenient Chrome is. Not as a browser, but as a media player.
13:53
<Workshiva>
mp3, h264, aac, it's got all the stuff missing by default :)
14:11
<gsnedders>
jgraham: Someone left their glasses at my flat. I think they're yours.
14:45
<bossss>
I'm using google chrome v5, and I'm experiencing laggy html5 video? Is there a browser that can better render it? or a fix for these kinds of problems?
14:47
<hsivonen>
bossss: Firefox trunk
14:47
<Rik`>
(on windows ?)
14:47
<bossss>
thanks and yeah on windows
14:48
<Rik`>
actually, I was asking hsivonen
14:49
<hsivonen>
Rik`: I haven't really tested the D3D back end
14:49
<hsivonen>
(for full screen, that is)
14:50
<hsivonen>
Rik`: I said trunk, because I wasn't sure if beta 1 still had the media cache bug that made video suck after x megabytes
14:50
<hsivonen>
where x was 50, IIRC
14:51
<hsivonen>
hmm. or maybe less
14:51
<hsivonen>
anyway
14:51
<Rik`>
I haven't followed that bug
14:51
<hsivonen>
oops. whoa. the bug is still on trunk, too
14:51
<kbrosnan>
webm? or ogg?
14:51
<hsivonen>
kbrosnan: both
14:55
<hsivonen>
anyway, video on Firefox trunk or beta 1 is pretty good if you don't hit the cace limit that breaks things
14:58
<bossss>
well i installed minefield and i'll see how it goes
14:59
<jgraham>
gsnedders: They are
14:59
<jgraham>
Really other people would have noticed
14:59
<jgraham>
By e.g. not being able to see
15:00
<jgraham>
But since I can basically see anyway, it is n easy and common mistake for me to make
15:01
<gsnedders>
jgraham: You're not the only one to have made the mistake. :) I guess giving them to you tomorrow is enough?
15:01
<jgraham>
Yeah
15:01
<bossss>
Didn't work >.< now I get a big X on the video
15:02
<jgraham>
gsnedders: Which day are you leaving again? The 29th?
15:03
<gsnedders>
jgraham: yeah
15:03
<jgraham>
So what are you doing on the 28th?
15:03
<jgraham>
We having a party?
15:04
<hsivonen>
gsnedders: are you leaving Opera? just leaving Sweden?
15:05
<hsivonen>
bossss: is the video in a patent-unencumbered format and served with the right Content-Type HTTP header?
15:05
<bossss>
these are the videos http://www.eseanews.com/index.php?s=downloads
15:07
<hsivonen>
bossss: you ext=webm urls redirect to .mp4
15:08
<Rik`>
and there's also a redirection : location: http://esea.net/premium
15:08
<Rik`>
(at least using curl -I)
15:08
<hsivonen>
s/you/your/
15:09
<hsivonen>
bossss: and the .mp4 URL is 404
15:12
<bossss>
thanks for the help - I'm just reading on a related site that a couple videos are corrupted, so it may not be my system
15:13
<bossss>
minefield gives me the X on the videos, chrome plays some of them
16:27
<micheil>
Rik`: has that article you mentioned on hacks.mozilla.org been published yet?
16:31
<micheil>
Rik`: also, the current patch for websockets in bugzilla is a bit wrong, there's 4 readyStates
16:31
<micheil>
connecting, open, closing, closed
16:35
<Ms2ger>
micheil, there was a followup for that landed at the same time, IIRC
16:36
<micheil>
okay
16:36
<micheil>
I can only see the one patch
16:37
<micheil>
then again, no browsers' perfect
16:37
<micheil>
in Safari 5, if you create a connection, then close it, you'll get an error saying that there was an extra character sent
16:38
<micheil>
it seems to be the last \r\n on the response
16:38
<micheil>
if you leave it off, then google chrome won't connect at all.
16:38
<titacgs>
speaking of websockets... is there examples on how to implement this?
16:39
<micheil>
titacgs: serverside?
16:39
<micheil>
or?
16:39
<micheil>
or clientside?
16:40
<micheil>
(I have working examples of both, both written in javascript with help from node.js)
16:40
<titacgs>
both... I read the websocket API but I still don't get how ti get things done
16:41
<titacgs>
micheil, could you share them?
16:41
<micheil>
sure: http://github.com/miksago/node-websocket-server/blob/development/lib
16:41
<micheil>
and http://github.com/pgriess/node-websocket-client/blob/master/lib/websocket.js
16:42
<titacgs>
thanks! I'll give them a look right now!
16:42
<micheil>
I'm personally not a fan of peter's (pgriess) coding style, but, eh'
19:29
<AryehGregor>
I didn't realize that <video> was only proposed three years ago. I thought it was part of HTML5 since the early days.
19:30
<AryehGregor>
That makes it cooler that it's actually taking off.
20:18
<Rik`>
micheil: no, I've done some other testing and I aslo need to figure out on which server to put a demo
20:21
<Rik`>
so that it can at least handle the traffic
20:30
<AryehGregor>
So it looks like as of April, three of the six hit counters here deemed IE to be a minority browser: http://en.wikipedia.org/wiki/Usage_share_of_web_browsers
20:56
<micheil>
Rik`: okay
20:56
<Rik`>
micheil: but don't worry, I'll ping you ;)
21:41
<micheil>
Rik`: okay, cool, thanks :)
22:10
<micheil>
Rik`: do you know if firefox uses a different version of websockets to draft76?
22:10
<Rik`>
it should support draft76, why ?
22:10
<micheil>
I'm being told by someone that it's actually a different version, just wanting to check the information
22:10
<Rik`>
I've seen some commits here and there around that, so it might not be perfect
22:11
<micheil>
because if I can pre-empt the release of firefox 4 with websockets, then I will.
22:14
<Rik`>
micheil: the master bug is https://bugzilla.mozilla.org/show_bug.cgi?id=websocket
22:15
<Rik`>
micheil: "pre-empt" ? I don't understand this word
22:16
<Rik`>
micheil: for the 3 or 4 demos I've made, your server implementation works
22:27
<micheil>
Rik`: pre-empt, as in, make sure node-websocket-server is automatically compatible with firefox's implementation, before it's public
22:30
<Rik`>
I don't know why it wouldn't be draft76
22:30
<micheil>
niether
22:30
<Rik`>
iirc, Mozilla pushed for this revision so…
22:31
<Rik`>
as I told you, there was the Big endian problem with a previous version of your server but it's ok now
22:55
<micheil>
yeah, alright