00:06
<midgard>
Hixie I like that way to change html to an unversioned markup language
01:48
<jamesr>
does silvia pfeiffer hang out here?
01:48
<nessy>
yes, hi!
01:49
<jamesr>
nessy: hi!
01:49
<nessy>
wassup? :)
01:49
<jamesr>
nessy: i think a source of confusion in the volume thread is that your TV does not really increase the volume past what the video says. it attenuates the volume down (from some maximum that it supports)
01:50
<jamesr>
in fact more sophisticated systems always express volume controls in terms of negative dB
01:50
<nessy>
are you trying to tell me that the video that comes over the air is encoded as loud as the maximum setting of my TV?
01:50
<jamesr>
think about it this way
01:51
<jamesr>
if you put the volume on your TV at 100% that's the same as setting it to -0dB
01:51
<nessy>
sure, if we go with dB - but the video and audio elements don't go with dB
01:51
<jamesr>
well it's the same thing
01:51
<Hixie>
volume=1.0 means "play at maximum volume"
01:51
<jamesr>
it's a % of the max
01:52
<jamesr>
1.0 or 100% = -0dB
01:52
<jamesr>
0.5 or 50% = whatever that is in dB
01:52
<jamesr>
i dunno what that is, bad at logs
01:52
<nessy>
how come then that most videos on the web are so quiet that they all the max settings on my computer cannot make them audible when my TV is easily able to go louder than my vacuum cleaner
01:52
<jamesr>
i think that's around -60dB
01:53
<Hixie>
nessy: your speakers are turned down :-)
01:53
<jamesr>
nah that can't be right. i dunno what 50% would be
01:53
<nessy>
Hixie: now, everything is cranked all the way up
01:53
<Hixie>
jamesr: wouldn't 0 be -infinity?
01:53
<Hixie>
nessy: how loud are system sounds?
01:54
<nessy>
I'm on a MacBook
01:54
<jamesr>
Hixie: the mapping isn't really mathematical because perceived loudness doesn't map linearly to energy output
01:54
<jamesr>
or it's not a strict interpretation
01:54
<Hixie>
jamesr: ah ok
01:54
<Hixie>
nessy: right so does the system beep beep loudly?
01:54
<Hixie>
nessy: do is that quiet too?
01:54
<Hixie>
s/do/or/
01:54
<nessy>
yes, very much so
01:55
<nessy>
I get a little shock every time :)
01:55
<jamesr>
what that means is that the browser (or flash) is attenuating the audio down more than system sounds
01:55
<Hixie>
and when you hit the red button on http://instantrimshot.com/ does it play as loudly?
01:55
<nessy>
yeah, that's fairly loud
01:56
<Hixie>
what videos play quietly for you?
01:56
<nessy>
it's not all of them obviously - but a lot are encoded too quietly
01:56
<nessy>
that's not something that happens on TV, where 0dB seems fairly standardised
01:57
<jamesr>
my TV at 0dB is unbearably loud
01:57
<Hixie>
welcome to the web, where anyone and their dog can publish content
01:57
<nessy>
what I am saying is that a volume setting of 1 isn't 0dB
01:57
<Hixie>
including people who know jack about audio engineering
01:58
<jamesr>
Hixie: we should make it go to 11
01:58
<jamesr>
that'd be one louder
01:58
<jamesr>
problem solved!
02:00
<nessy>
the problem is that 0dB on TV refers to a standard reference level, but our 0-1 volume range doesn't
02:00
<jamesr>
i don't think that's true and i'm not sure how else to explain it
02:01
<Hixie>
0 means muted. 1 means do not attenuate.
02:01
<jamesr>
do not attentuate is by definition the same thing as attenuate by 0dB
02:01
<Hixie>
exactly
02:01
<nessy>
that means 0dB in reference to the SPL at which the video or audio file were encoded
02:02
<nessy>
it doesn't mean in reference to a SPL of human perception, e.g. where 0dB would equate to 20 Pa
02:03
<nessy>
think of the Web as a radio station where every song is played back at a different level - so every Web page you go to with a video will play back at a different level
02:04
<nessy>
it would make a lot more sense to define the 0-1 range as a SPL range
02:04
<Hixie>
we're not saying 0dB volume, we're saying 0dB relative attentuation
02:04
<Hixie>
it's further attenuated (or amplified) by your system volume control(s), amplifier(s), speaker(s), etc
02:04
<nessy>
and I am saying relative attenuation is the wrong thing to optimise on
02:05
<nessy>
assuming all other things are fixed - volume controls, amplifiers, speakers - I would expect any video set at 0.5 volume to play back at the same perceived loudness
02:05
<nessy>
that's what my radio station does (unless they screw up)
02:06
<Hixie>
oh you're saying you want the browser to do volume normalisation?
02:06
<nessy>
exactly :)
02:06
<jamesr>
that's orthogonal to the 'volume' attribute
02:06
<Hixie>
oh, well
02:06
<Hixie>
that's a whole different ball of wax
02:06
<nessy>
but volume 0-1 is what the user / web developer sees
02:06
<Hixie>
has nothing to do with the .volume IDL attribute
02:07
<jamesr>
CDs can be encoded at different volume levels as well which results in the exact same issue
02:07
<jamesr>
in fact some labels have historically tried to jack up the encoded volume to make their albums sound louder or cooler. which hurts the dynamic range and quality, etc
02:07
<nessy>
indeed, jamesr, the Web is more like a CD collection than a radio station - and I don't think that makes any sense at all
02:08
<Hixie>
can you even do volume normalisation on streams?
02:08
<Hixie>
i thought you needed the whole track to do it
02:08
<jamesr>
Hixie: you can't do it very well
02:08
<Hixie>
how would you do it in real time?
02:08
<jamesr>
itunes volume normalization scans the whole track and figures out the dynamic range
02:09
<jamesr>
it kind of sucks too
02:09
<jamesr>
because i have CDs with quiet and loud tracks. itunes would make the quiet ones louder
02:09
<nessy>
you can do a heuristic with the distribution of volume ranges
02:09
<jamesr>
i don't want The Great gig in The Sky to be as loud as Money
02:09
<nessy>
fair enough :)
02:10
<nessy>
but something should be possible - it really is a horrendous situation to not be able to turn some audio tracks even up into the audible range
02:11
<jamesr>
but you wouldn't expose that to the author. the author (presumably) controls the encoding process
02:11
<nessy>
when I stick in external speakers, it's a lot better - so if they can increase the volume, so should we - and often people don't care about the loss in quality if it's something important they want to catch
02:12
<nessy>
jamesr, the web page author is most of the time not the same person that creates the content
02:12
<nessy>
just like a radio station isn't most of the time the one who creates the content
02:13
<jamesr>
yeah, but radio stations control the volume of what they broadcast at transmission time
02:13
<nessy>
exactly - so why can't the browser do that for us
02:14
<Hixie>
the browser is a radio, not a transmission tower
02:14
<Hixie>
the server is the tower
02:14
<Hixie>
anyway, gotta go. bbl!
02:14
<nessy>
I would have thought that's the purpose of the volume attribute to normalise to a certain SPL range and then the original setting can be the one at which the volume bar starts - but I'd be happy with an extra attribute for normalization
02:15
<nessy>
Hixie: I am aware that the browser is the radio, that's where the comparison falls down - it doesn't change the user perception issue though :)
02:17
<jamesr>
well, i guess you could propose that as a separate feature
02:19
<nessy>
on high-tech receivers your volume know also goes above 0dB
02:21
<nessy>
anyway - seems like this is going back to the list :)
05:53
<MikeSmith>
nimbupani: I think Deepak Chopra should have mated with Mother Theresa when he had the chance
05:53
<MikeSmith>
big opportunity missed there
05:53
<nimbupani>
ROFLLLL
05:54
<nimbupani>
i do not even want to think of what calamity would have been unleashed
05:54
<MikeSmith>
heh
05:54
<MikeSmith>
well, it would have given him some serious street cred in the self-help guru community
05:55
<nimbupani>
lolzzzzz along with "celebrate the poor" celebrities.
05:55
<nimbupani>
i sorta experimented with Gandhi quotes https://twitter.com/standardsgandhi
05:55
<nimbupani>
:)
05:56
<nimbupani>
it got harder to make his quotes fit webdev :P
05:57
MikeSmith
reads
05:59
<MikeSmith>
I like https://twitter.com/standardsgandhi/status/14068177858
05:59
<MikeSmith>
that's a keeper
05:59
<nimbupani>
LOL :D
06:00
<nimbupani>
i know there are tonnes of people who would beat me up for that :P
06:01
<MikeSmith>
finding a way to simultaneously mock Ghandi and XHTML in less than 140 characters is worthy goal
06:02
<nimbupani>
:D
06:02
<MikeSmith>
my favorite Ghandi quote is the one where he asks, "Is there anything funnier in the world than depictions of familiar cartoon characters have sex with each other?"
06:02
<MikeSmith>
'cause that's a profound one
06:02
<nimbupani>
I bet he never said it :P
06:03
<nimbupani>
profanely profound.
15:02
<MikeSmith>
you hipsters in Sweden are lucky to have some of the best record labels in the world there
15:07
<MikeSmith>
http://www.hybrism.com/about/
15:07
<MikeSmith>
"…situated in Sweden; in the royal capital of Stockholm; and in the not so royal city of Malmö."
15:45
<Dashiva>
Heh, I'm reading the link registry saga
15:46
<Dashiva>
Whose idea was it to suggest the HTML use a global link registry and then reject any link relation that was specific to HTML?
15:47
<MikeSmith>
Dashiva: some might say it as maybe the idea of those who were acting in bad faith from the beginning
15:47
<MikeSmith>
without naming any names
15:48
<MikeSmith>
of either the people who might say that, or the people they might say it about
15:48
<MikeSmith>
anyway, if the link registry is not going to be useful in the way it was supposed to be, everybody will just do an end run around it
15:49
<MikeSmith>
and spec out mechanisms that violate the link-relations RFC
15:49
<MikeSmith>
and it will be hard to blame them for doing so
15:54
<Dashiva>
Yeah, hopefully there won't be too much wasted effort as a result
16:02
<Dashiva>
I'm somewhat amused by the @longdesc comic having an obviously invalid @alt
16:11
<MikeSmith>
the continuing lack of explicit depictions of sex between cartoon characters in those comics is depressing
16:15
<Dashiva>
The crossdressing wasn't enough for you?
16:15
<Dashiva>
Haha, this is great
16:16
<Dashiva>
Julian demands large changes to existing specs to register link relations, but we can't fix the registry because the RFC defining it is almost final
17:05
<Dashiva>
Julian: The pingback specification has been final for almost a decade. Didn't stop you from requiring that it was rewritten as a RFC.
17:07
<Dashiva>
You seem to miss the point about HTML-specific link registrations. If everything can't fit in the IANA registry, then we need a second registry to handle the rest, and then we might as well use the second registry for everything and save massive overhead.
17:38
<JulianReschke>
I found the earlier exchanges about the link relations stuff not helpful. Apparently it's not trivial to understand what was and what is going on.
17:38
<JulianReschke>
rather t
17:38
<JulianReschke>
Rather than talking 2 each other, I'd recommend that you include those into the conversation who actually were part of the discussion.
17:39
<JulianReschke>
The best place to do so would be the HTML WG mailing list.
17:39
<JulianReschke>
That being said: if there's a question regarding the link relation stuff, I'm listening.
18:40
<Speeder>
what is this channel?
18:41
<Speeder>
iti s related to HTML? (I am seeking someone, or a channel to seek someone, to set-up a video player that supports expanding itself to native resolution, without javascript or flash hackery...)
18:59
<Moo^_^>
Speeder: whatwg is working group working on html5 spec
18:59
<Moo^_^>
Speeder: there won't be such a video player so you can stop searching
18:59
<Speeder>
that is not w3c ?
18:59
<Moo^_^>
Speeder: see topic
18:59
<Speeder>
Moo^_^ and WITH javascript, posible?
19:00
<Moo^_^>
Speeder: only for very latest browsers support html5 <video>
19:00
<Speeder>
Moo^_^: I know that
19:00
<Moo^_^>
Speeder: let's say "it will be better in the future if you don't want to use flash"
19:00
<Moo^_^>
Speeder: if you want to support all browsres
19:00
<Speeder>
Moo^_^ I will add a fallback to youtube :P
19:00
<Moo^_^>
the tube is yours
19:01
<Moo^_^>
:)
19:06
<JulianReschke>
Speeder: WHATWG = W3C
19:06
<JulianReschke>
Speeder: if you want the W3C HTML WG, the channel is irc://irc.w3.org:6665/html-wg
19:10
<Hixie>
man, don't let the w3c hear you say that, they'll have an aneurism :-)
19:17
<JulianReschke>
!=
19:17
<JulianReschke>
oops
19:18
<MikeSmith>
"sh: =: command not found"
19:19
<MikeSmith>
looks like the "=" command is missing on my system
19:19
<MikeSmith>
I guess I should fix that
19:19
<MikeSmith>
I think I'm missing the "." command
19:19
<MikeSmith>
as well
19:20
<Dashiva>
MikeSmith: Did you put whitespace around it?
19:20
<MikeSmith>
I put a return after it
19:20
<JulianReschke>
what, no dot command?
19:20
<MikeSmith>
heh
19:21
<MikeSmith>
I got a deficient shell
19:21
<JulianReschke>
command.com?
19:21
JulianReschke
ducks
19:21
<Dashiva>
The first error sounds like vanilla sh
19:21
<Dashiva>
But surely that's extinct by now
19:23
<MikeSmith>
。should do the same thing as .
19:23
<MikeSmith>
that was the problem
19:23
<MikeSmith>
stupid shell
19:23
<MikeSmith>
I prefer to type my dot as a reverve-video do
19:23
<MikeSmith>
dot
19:23
<MikeSmith>
and the shell should just understand that
19:23
<MikeSmith>
it should learn all my personal idiosyncracies
19:23
<Dashiva>
Yeah, and should support full-width characters too
19:23
<MikeSmith>
and it should love them
19:24
<MikeSmith>
yes
19:24
<JulianReschke>
what, and not allowing to make this a *different* command?
19:24
MikeSmith
kicks his crappy shell
19:24
<MikeSmith>
hey, maybe I will write a new shell
19:24
<Dashiva>
MikeSh
19:24
<MikeSmith>
because clearly we don't have enough of those already
19:24
<MikeSmith>
heh
19:24
<MikeSmith>
jackassh
19:25
<JulianReschke>
please add whitespace overloading
19:26
<Dashiva>
Maybe make a javascript-based shell
19:27
<JulianReschke>
http://www2.research.att.com/~bs/whitespace98.pdf
19:30
<Dashiva>
krijnh: So how's that open sourcing of the logs page going? :)
19:43
<MikeSmith>
threads in ruby?
19:43
<MikeSmith>
ruby has a thread library?
19:44
<MikeSmith>
wow
19:44
<MikeSmith>
who thought this was a good idea?
19:44
<MikeSmith>
is this some attempt at compensating for poor overall performance?
19:45
<MikeSmith>
err, I now see python has threads too
19:45
<MikeSmith>
sign
19:45
<MikeSmith>
of problems elsewhere
19:46
<MikeSmith>
I suppose people really do use it
19:46
<Moo^_^>
MikeSmith: threads don't work well in python
19:46
<MikeSmith>
ok
19:46
<Moo^_^>
MikeSmith: global interpreter locks allows just on thread to execute python code once
19:46
<Moo^_^>
MikeSmith: if you are doing IO it is ok
19:46
<Moo^_^>
MikeSmith: otherwise you do processes
19:46
<MikeSmith>
hmm
19:46
<MikeSmith>
I see
19:47
<MikeSmith>
why don't people just use C++ or C instead, I wonder
19:47
<MikeSmith>
but oh well
19:47
<MikeSmith>
I mean, for applications that benefit from having lots of threads
19:48
<Moo^_^>
MikeSmith: I could have many arguments for that but I don't want to start flamewar before I drink the first beer of saturday :)
19:48
<MikeSmith>
heh
19:49
<MikeSmith>
cheers
19:49
MikeSmith
reads about http://en.wikipedia.org/wiki/Global_Interpreter_Lock
19:49
<Moo^_^>
MikeSmith: personally I hate it
19:49
<Moo^_^>
MikeSmith: it prevents creating low resource python web servers
19:50
<Moo^_^>
MikeSmith: because if you need processes
19:50
<Moo^_^>
MikeSmith: you go out of mem really fast
19:50
<MikeSmith>
yeah, I would think
19:50
<Moo^_^>
OTOH the set-up is more fault tolerant
19:50
<Moo^_^>
but still
19:50
<Moo^_^>
not good for shared hosting
19:50
<Moo^_^>
low end hosting
19:50
<Moo^_^>
for masses
19:51
<MikeSmith>
I thought the world had decided that message passing was the way to go these days
19:51
<MikeSmith>
for concurrency and other reasons
19:52
<Moo^_^>
MikeSmith: it's the way of the future :)
19:52
<Moo^_^>
MikeSmith: we haev a lot of legacy to deal with
19:52
<Moo^_^>
MikeSmith: Python also has lots of "tasklet" implementations
19:52
<Moo^_^>
or whatever they call them, depending on the implementation
19:52
<MikeSmith>
ok
19:52
<Moo^_^>
eventlets
19:52
<Moo^_^>
the stuff erlang is good at
19:53
<MikeSmith>
ok
19:53
<MikeSmith>
erlang sets the bar pretty high
19:53
<Moo^_^>
the thing with python, ruby et. al. is that you can get open source libaries for them
19:54
<Moo^_^>
like "rss reader"
19:54
<Moo^_^>
for erlang, you need to build everything by hand
19:54
<Moo^_^>
so the scope of the use is limited
19:54
<MikeSmith>
ah yeah
19:54
<MikeSmith>
true
20:55
<Cesarino>
someone?
21:04
<MikeSmith>
Cesarino: everybody is hiding from you
21:41
<hdhoang>
the Diaspora guys are "having problems between Websockets and Rails.", as a Mozilla guy mentioned in passing