IRC log of tp on 2007-11-07

Timestamps are in UTC.

13:41:01 [RRSAgent]
RRSAgent has joined #tp
13:41:01 [RRSAgent]
logging to http://www.w3.org/2007/11/07-tp-irc
13:41:09 [Ralph]
rrsagent, please make record member-visible
13:41:11 [mauro]
mauro has changed the topic to: Technical Plenary Day
13:41:16 [Ralph]
Meeting: Technical Plenary
13:41:25 [Zakim]
Zakim has joined #tp
13:41:25 [mauro]
mauro has changed the topic to: Technical Plenary Day - agenda at http://www.w3.org/2007/11/07-TechPlenAgenda.html
13:41:32 [Ralph]
zakim, this will be tp
13:41:32 [Zakim]
ok, Ralph, I see W3C_TP()8:00AM already started
13:41:42 [DanC_lap]
DanC_lap has joined #tp
13:42:56 [raman]
raman has joined #tp
13:43:06 [DanC_lap]
how about side chat? mix it in here or encourage it to go elsewhere by making #tp-backchannel and noting it in the /topic ?
13:43:23 [Steven]
Steven has joined #tp
13:43:25 [raman]
Morning all
13:43:32 [Steven]
hi Raman
13:43:35 [mauro]
morning raman
13:43:37 [DanC_lap]
morning raman. any opinion about side chat?
13:44:08 [raman]
I'd rather see it all in one place.
13:44:10 [Steven]
zakim, who is in the phone?
13:44:10 [Zakim]
I don't understand your question, Steven.
13:44:14 [moshe]
moshe has joined #tp
13:44:18 [Zakim]
-Captioner
13:44:24 [Steven]
zakim, who is on the phone?
13:44:24 [Zakim]
On the phone I see no one
13:44:31 [Norm]
Norm has joined #tp
13:44:42 [timbl_]
timbl_ has joined #tp
13:44:44 [PGrosso]
PGrosso has joined #tp
13:44:48 [raman]
in the spirit of side-chat == marginal notes, and that margins can often hold important proofs ;-)
13:44:53 [ht]
ht has joined #tp
13:45:03 [mauro]
+1 to raman
13:45:27 [Zakim]
+MeetingRoom
13:46:16 [clc4tts]
clc4tts has joined #tp
13:46:34 [Zakim]
+Captioner
13:47:01 [amy]
amy has joined #tp
13:47:05 [marie]
marie has joined #tp
13:47:23 [maxf]
maxf has joined #tp
13:47:33 [Mez]
Mez has joined #tp
13:47:57 [DanC_lap]
I hear pepole who haven't been given the password are gathering in #tpac
13:48:00 [mauro]
mauro has changed the topic to: W3C Technical Plenary Day - agenda at http://www.w3.org/2007/11/07-TechPlenAgenda.html
13:48:15 [barstow]
barstow has joined #tp
13:48:16 [Mez]
well go over there and tell them they're in the wrong place
13:48:19 [maxf]
zakim, who's here?
13:48:19 [Zakim]
On the phone I see Captioner (muted), MeetingRoom
13:48:20 [Zakim]
On IRC I see barstow, Mez, maxf, marie, amy, clc4tts, ht, PGrosso, timbl_, Norm, moshe, Steven, raman, DanC_lap, Zakim, RRSAgent, najib, Ralph, mauro, karl, fsasaki, olivier, ivan,
13:48:22 [Zakim]
... Keeper, Yves, ted
13:48:24 [amy]
what does "given the password" mean?
13:48:43 [ht]
this channel is pw-protected, I was told
13:48:44 [amy]
isn't it available to anyone attending?
13:48:46 [maxf]
it's not on the agenda
13:48:46 [Mez]
normal people needed a password to get here amy
13:49:10 [DanC_lap]
I went over there (#tpac); the topic doesn't suggest they're in the wrong place, and the people there are not just the disenfranchised.
13:49:21 [amy]
sorry, Mez, wasn't questioning who can get on, just questioning where the info should be! sorry, being logistical :)
13:49:34 [Mez]
sorry if I sounded rude; must be my irc voice :-)
13:49:36 [amy]
:)
13:49:46 [raman]
her irked voice;-)
13:49:52 [Mez]
ouch!
13:50:01 [DanC_lap]
amy, the password hasn't been announced, I'm pretty sure. ah. there. steve is announcing it.
13:50:02 [glazou]
glazou has joined #tp
13:50:28 [amy]
ah, ok, good
13:50:37 [raman]
one web to bind them all
13:50:41 [Hixie]
Hixie has joined #tp
13:50:42 [MSM]
MSM has joined #tp
13:50:44 [raman]
one irc chanel to gather them all
13:50:45 [Bert]
Bert has joined #tp
13:50:48 [Mez]
raman you slay me
13:50:48 [MikeSmith]
MikeSmith has joined #tp
13:50:53 [burn]
burn has joined #tp
13:51:06 [dsr]
dsr has joined #tp
13:51:07 [justin]
justin has joined #tp
13:51:10 [ferraiolo]
ferraiolo has joined #tp
13:51:16 [chibao]
chibao has joined #tp
13:51:26 [Rich]
Rich has joined #tp
13:51:28 [mchampio2]
mchampio2 has joined #tp
13:51:33 [rigo]
rigo has joined #tp
13:51:53 [anthony]
anthony has joined #tp
13:52:02 [matt]
matt has joined #tp
13:52:02 [Ashok]
Ashok has joined #tp
13:52:07 [jallan]
beantown
13:52:10 [matt]
zakim, dial matt-voip
13:52:10 [Zakim]
ok, matt; the call is being made
13:52:12 [Zakim]
+Matt
13:52:14 [jresig]
jresig has joined #tp
13:52:31 [Noah]
Noah has joined #tp
13:52:32 [AnnBassetti]
AnnBassetti has joined #tp
13:52:33 [hal]
hal has joined #tp
13:52:34 [jerome]
jerome has joined #tp
13:52:36 [caribou]
caribou has joined #tp
13:52:37 [brutzman]
brutzman has joined #tp
13:52:38 [matt]
zakim, mute me
13:52:38 [Zakim]
Matt was already muted, matt
13:52:40 [paolob]
paolob has joined #tp
13:52:50 [herve]
herve has joined #tp
13:53:00 [dbaron]
dbaron has joined #tp
13:53:04 [mauro]
===============
13:53:06 [steph]
steph has joined #tp
13:53:17 [wellsk]
wellsk has joined #TP
13:53:20 [Danny]
Danny has joined #tp
13:53:30 [Rafa]
join #tp beantown
13:53:34 [DS]
DS has joined #tp
13:53:37 [matt]
zakim, drop matt
13:53:37 [Zakim]
Matt is being disconnected
13:53:38 [Zakim]
-Matt
13:53:44 [Andrew]
Andrew has joined #tp
13:53:50 [RogerCutler]
RogerCutler has joined #tp
13:53:50 [mdean]
mdean has joined #tp
13:54:05 [RogerCutler]
RogerCutler has left #tp
13:54:11 [mauro]
Steve Bratt welcomes everybody and goes through the agenda of the week
13:54:16 [ivan]
zakim, dial ivan-voip
13:54:16 [Zakim]
ok, ivan; the call is being made
13:54:17 [jallan]
jallan has joined #tp
13:54:17 [Zakim]
+Ivan
13:54:27 [lisap]
lisap has joined #tp
13:54:52 [Rafa]
Rafa has joined #tp
13:54:57 [jun]
jun has joined #tp
13:55:02 [mikko]
mikko has joined #tp
13:55:09 [RCutler]
RCutler has joined #tp
13:55:15 [raman]
all it takes is a bad talk to get a lot of boos ;-)
13:55:16 [anne]
anne has joined #tp
13:55:22 [Mez]
bad pun man
13:55:40 [Hixie]
anyone know the username/password for the realtime captioning?
13:55:43 [janet]
janet has joined #tp
13:55:48 [Stuart]
Stuart has joined #tp
13:55:48 [fumi]
fumi has joined #tp
13:55:48 [Kangchan]
Kangchan has joined #tp
13:55:54 [maxf]
Hi ivan!
13:56:05 [matt]
zakim, dial matt-voip
13:56:05 [Zakim]
ok, matt; the call is being made
13:56:07 [Zakim]
+Matt
13:56:10 [MichaelC]
MichaelC has joined #tp
13:56:17 [DS]
help /wave
13:56:22 [Roland_]
Roland_ has joined #tp
13:56:24 [ted]
we are doing an experimental audio broadcast of today @ http://media.w3.org:8000/stream.ogg and http://media.w3.org/stream.html includes an ogg vorbis java applet client
13:56:29 [mauro]
-> http://www.w3.org/2002/09/wbs/35125/tpac2007-feedback/ Feedback on the 2007 Technical Plenary and Advisory Committee meetings week (TPAC2007)
13:56:33 [matt]
zakim, who is talking?
13:56:41 [Eman]
Eman has joined #tp
13:56:46 [KRosenbl]
KRosenbl has joined #TP
13:56:51 [Zakim]
matt, listening for 10 seconds I could not identify any sounds
13:56:52 [AaronGustafson]
AaronGustafson has joined #tp
13:57:14 [karl]
hmmmmm steve bratt didn't mention that people could blog about the session today
13:57:22 [Steven]
zakim, who is on the phone?
13:57:22 [Zakim]
On the phone I see Captioner (muted), MeetingRoom, Matt (muted), Ivan (muted)
13:57:26 [karl]
http://www.w3.org/QA/2007/11/tpac-2007-lets-start
13:57:30 [raman]
he didn't say we couldn't -- so there
13:57:35 [DanC_lap]
no? I thought he just did [encourage blogging], Karl
13:57:43 [Henny]
Henny has joined #tp
13:57:45 [karl]
If you are taking photos, blogging, etc. use one of these tags "tpac, tpac2007, w3c, w3c_tpac07"
13:57:49 [DanC_lap]
Toipc: Session 2: View from the Outside: Real World Perspectives on the W3C
13:57:49 [maxf]
Hixie, any password any login work
13:57:54 [karl]
DanC_lap: then I have to wake up ;)
13:57:59 [maxf]
http://www.captionedtext.com/client/event.aspx?CustomerID=732&EventID=863005&ParticipantId=35e14a7c-868f-4405-b1e6-fb7ec1187b14
13:58:02 [maxf]
pretty cool!
13:58:10 [caribou]
s/Toipc/Topic
13:58:24 [shadi]
shadi has joined #tp
13:58:26 [matt]
zakim, drop matt
13:58:26 [Zakim]
Matt is being disconnected
13:58:27 [Zakim]
-Matt
13:58:32 [mauro]
===========
13:58:33 [mauro]
Molly E. Holzschlag (Web Standards and Practices Education and Outreach, Molly.Com, Inc.)
13:58:41 [matt]
zakim, dial matt-voip
13:58:41 [Zakim]
ok, matt; the call is being made
13:58:42 [Zakim]
+Matt
13:58:44 [Hixie]
maxf: cool
13:58:58 [mauro]
Session 2: View from the Outside: Real World Perspectives on the W3C
13:59:10 [plh]
plh has joined #tp
13:59:21 [Zakim]
-Matt
13:59:21 [plh]
rrsagent, where am I?
13:59:21 [RRSAgent]
See http://www.w3.org/2007/11/07-tp-irc#T13-59-21-1
13:59:23 [csma]
csma has joined #tp
13:59:25 [matt]
zakim, dial matt-voip
13:59:25 [Zakim]
ok, matt; the call is being made
13:59:27 [Zakim]
+Matt
13:59:57 [Zakim]
-Matt
14:00:00 [fantasai]
fantasai has joined #tp
14:00:01 [KevinLawver]
KevinLawver has joined #tp
14:00:21 [Hideki]
Hideki has joined #tp
14:00:42 [Mez]
the tower; my least favorite card in the tarot deck
14:01:03 [Julian]
join tp beantown
14:01:15 [matt]
zakim, dial matt-voip
14:01:15 [Zakim]
ok, matt; the call is being made
14:01:15 [Jean-Gui]
Jean-Gui has joined #tp
14:01:17 [Zakim]
+Matt
14:01:23 [raman]
how do you build a tower out of ivory ... have always wondered ...
14:01:31 [matt]
zakim, mute matt
14:01:31 [Zakim]
Matt was already muted, matt
14:01:32 [Julian]
Julian has joined #tp
14:01:45 [lm]
lm has joined #tp
14:01:47 [Mez]
lots of sad, dead elephants (I don't "read" it as an ivory tower; that's Molly's call)
14:02:06 [fantasai]
raman: you'd need some other material to bind it
14:02:12 [raman]
Whos' the large elephant in the room that no one can see;-)
14:02:14 [mauro]
Molly E. Holzschlag introduces Aaron Gustafson (Easy! Designs, Inc.)
14:02:22 [fantasai]
but concrete towers aren't all concrete anyway; they're also part steel
14:02:25 [jo]
jo has joined #tp
14:02:58 [mauro]
Molly E. Holzschlag introduces Patrick Haney (Harvard)
14:03:16 [raman]
captioning should create better minutes/transcripts
14:03:17 [mauro]
Molly E. Holzschlag introduces Matthew Oliphant (MathWorks)
14:03:18 [maxf]
fantasai, we're trying to show we can do better collaboratively ;)
14:03:50 [s-mon]
s-mon has joined #tp
14:04:06 [mauro]
Molly E. Holzschlag introduces Stephanie Troeth (CloudRaker)
14:04:19 [LeeF]
LeeF has joined #tp
14:04:23 [jose_ma]
jose_ma has joined #tp
14:04:28 [AaronGustafson]
I am one of the panelists and will be trying to keep an eye on this channel for remote questions during the Q&A period. I can't guarantee we'll be able to get to all of them, but we will certainly try.
14:04:47 [Mez]
I didn't quite follow all the intros, but it looks like several folks are in http://www.webstandards.org/
14:04:47 [danz]
danz has joined #tp
14:04:50 [brutzman]
brutzman has joined #tp
14:05:31 [laurent]
laurent has joined #tp
14:05:42 [Markus]
Markus has joined #tp
14:05:43 [steve]
steve has joined #tp
14:07:31 [Mez]
steph, my specialty interest is usable security (we're working a standard in that area in WSC). Did I catch right that you are in WaSP? Is there anything going on in an area related to usable security there?
14:08:12 [raman]
would be nice if captioning were available as a simple separate IRC channel.
14:08:41 [steph]
Mez: i was in the WaSP :) perhaps let's catch up after this.
14:08:55 [Mez]
great
14:08:56 [myakura]
myakura has joined #tp
14:08:57 [jgraham_]
jgraham_ has joined #tp
14:09:33 [raman]
the default on that web site is to refresh transcripts every second --- sounds like overkill
14:09:48 [shawn]
shawn has joined #tp
14:09:50 [dom]
dom has joined #tp
14:10:15 [francois]
francois has joined #tp
14:10:24 [Liam]
Liam has joined #tp
14:11:04 [raman]
room is echoing -- acoustically not a good room
14:11:05 [MikeSmith]
AaronGustafson - having trouble hearing Molly
14:11:43 [Jonathan]
Jonathan has joined #TP
14:12:13 [raman]
w3c specifications were never intended for use by end authors -- they've always been targetted at implementors. Confusing the audience leads to bad specifications
14:12:21 [fantasai]
anne, yeah, I agree that a lot of our specs could be written better but it's very unlikely that we can write them precisely enough for implementors and metaphorically enough for authors
14:12:46 [fantasai]
authors should be reading derived material
14:12:51 [karl]
anne, if web designers and web developers do not have to read the spec. Then there is a missing piece that web developers/designers still needs. I think it is the crux of the comment. "We have to read the spec because there's no other source."
14:12:52 [anne]
i think they should work for pedantic authors, not for average authors
14:13:26 [LeeF]
In reality (well, IME), though, the derivative materials don't exist early enough to let non-implementors participate in the consensus / review process
14:13:26 [raman]
I always compile with gcc --pedantic -- but I've never read the ANSI standard for C
14:13:44 [ivan]
dom, what is the user/password for the captioning?
14:13:45 [karl]
the message is "hey we want to understand your stuff". "your stuff is created at W3C". "where on w3c site can I find help for editing."
14:13:46 [MikeSmith]
KevinLawver - probably the same thing that would happen to the anti-virus industry if ...
14:13:47 [anne]
that said, the technology should be simple enough to use for average authors, obviously
14:14:03 [Steven]
ivan, there is no user/pass
14:14:09 [anne]
intuitive, if you will
14:14:14 [fantasai]
and the terminology in the interface/API/syntax should be targetted at them
14:14:27 [dom]
ivan, you just have to sell your sour^W^W^W enter your name and companie's name
14:14:46 [soonho]
soonho has joined #tp
14:15:16 [anne]
karl, don't people write books and tutorials on how to do stuff all the time?
14:15:29 [vivien]
vivien has joined #tp
14:15:30 [karl]
anne, reference point.
14:15:42 [karl]
http://www.captionedtext.com/client/event.aspx?CustomerID=732&EventID=863005&ParticipantId=35e14a7c-868f-4405-b1e6-fb7ec1187b14
14:15:50 [Dennis]
Dennis has joined #tp
14:15:54 [steph]
anne: often it's a case of too much information, not always consistent
14:16:06 [Ralph]
zakim, who's on the phone?
14:16:06 [Zakim]
On the phone I see Captioner (muted), MeetingRoom, Matt (muted), Ivan (muted)
14:16:07 [mjs]
mjs has joined #tp
14:16:08 [AaronGustafson]
please send questions direct to me
14:16:11 [moshe]
I wrote a book that mentions the values of standards... by analogy to 150 years ago, when manufacturers accepted standard sizes for nuts and bolts.
14:16:12 [timbl_]
So W3C should produce materials to he;p developer's *managers* understand why the w3C technology is important
14:16:14 [AaronGustafson]
will make it easier to keep track
14:16:19 [raman]
the stumbling block has not been lack of educational materials -- the problem has been lack of good interoperable implementations that work like what the educational materials try to teach
14:16:33 [Jules]
Jules has joined #tp
14:17:07 [raman]
dont think we should produce things for managers ... Darwin will take care of clueless managers over time;-)
14:17:09 [fantasai]
AaronGustafson: see KevinLawver's comments above
14:17:21 [karl]
raman, that is true. ANd there is also the high variability of quality in books.
14:17:37 [Hixie]
(is there an offtopic irc channel to where we should repair the general chat?)
14:17:40 [klaus_]
klaus_ has joined #tp
14:17:46 [edm]
edm has joined #tp
14:18:15 [John_Boyer]
John_Boyer has joined #tp
14:18:32 [maxf]
that's how you learn!
14:18:44 [Nick]
Nick has joined #tp
14:19:20 [raman]
and people are still doing browser-specific code -- he's making it sound like that was all in the past...
14:19:23 [anne]
well, interop is certainly a problem
14:19:23 [MikeSmith]
Hixie - #tp ?
14:19:25 [John_Boyer]
Seem to need more participation on the *public* mailing lists. Discussion is invited in the status of the document.
14:19:34 [anne]
testcases anyone?
14:19:42 [s-mon]
+1 to raman.
14:19:49 [Hixie]
MikeSmith: fair enough :-)
14:20:14 [shepazu]
shepazu has joined #tp
14:20:47 [tantek]
tantek has joined #tp
14:21:06 [DanC_lap]
re platform for innovation... that's my motivation for chairing the HTML WG
14:21:14 [hsivonen]
hsivonen has joined #tp
14:21:27 [tantek]
greetings from San Francisco
14:21:34 [RCutler]
There are multiple levels of indirection between standards and the managers who make the $ commitmeents. That doesn't make them clueless, but it does mean that it is difficult for you to target them.
14:21:36 [raman]
Hey Tantek!
14:21:37 [moshe]
standards do not inhibit innovation. standard nuts and bolts in automobiles let the designers focus on overall form and function instead of creating custom nuts and bolts.
14:22:11 [steph]
moshe: +1
14:22:12 [CharlieW]
CharlieW has joined #tp
14:22:13 [Mez]
if you have a standards effort that is a response to a vulnerability, there can be innovation that relies on that vulnerability
14:22:22 [tantek]
hey Raman. regrets I could not be there in person.
14:22:30 [timbl_]
tantek, pity you coudn't be herein person
14:22:44 [mjs]
hi tantek
14:22:49 [j1]
j1 has joined #tp
14:22:56 [Mez]
all those custom bolt people can no longer innovate in bolts
14:23:00 [raman]
tantek: we miss you (':' intentionally used)
14:23:33 [PGrosso]
PGrosso has joined #tp
14:23:41 [steph]
our green light appears to be stuck
14:23:54 [KevinLawver]
The general web development community's (of which I'm a part) frustration is NOT with the standard itself. It's with browsers (like TV said). They're tired of waiting for things like multiple backgrounds, border radius, display: table, a "real" layout system, etc.
14:24:10 [Dowan]
Dowan has joined #tp
14:24:11 [janet]
janet has joined #tp
14:24:15 [klaus]
klaus has joined #tp
14:24:18 [moshe]
Mez, this whole panel discussion sounds like the descriptions of the first meetings when people discussed adopting standardized nuts and bolts. Same arguments.
14:24:27 [John_Boyer]
standards accelerate innovation by increasing aperture of information flow (not just data flow)
14:24:30 [mjs]
KevinLawver: sounds like when you say "browsers" you mean "Internet Explorer"
14:24:39 [KevinLawver]
We mostly understand the benefits, but the definition of "standards" to almost all of my colleagues is "the hacks I have to use to make it work in IE" not the specs themselves.
14:24:52 [Rafa]
Rafa has joined #tp
14:24:55 [raman]
browsers should focus on doing standards right -- rather than continuously aiming for bug-compatibility across browsers
14:25:01 [RCutler]
How long ago were nuts and bolts standardized? Is this really in living memory?
14:25:19 [moshe]
no, it was in the 1850's
14:25:24 [moshe]
so we have some reports
14:25:27 [tantek]
timbl_: thanks and agreed. i will do my best to participate thru IRC. i also have an iSight and Skype/iChat with sufficient bandwidth if there is someone near the front of the room willing to share.
14:25:30 [anne]
raman, supporting existing content is important, especially to our end users
14:25:48 [raman]
if nuts and bolts were done today's web-browser way, we'd all carry around bagfuls of wrenches out of fear that you'd run into a badly designed legacy bolt any minute
14:25:49 [Mez]
http://www.time.com/time/magazine/article/0,9171,741291,00.html
14:26:01 [John_Boyer]
Anne, existing content is not expressed in the new versions of the language
14:26:04 [mjs]
supporting existing content is the way you get to have enough end users to fund implementing standards
14:26:19 [SteveH]
SteveH has joined #tp
14:26:21 [John_Boyer]
good
14:26:27 [steph]
hehehe
14:26:50 [Dennis]
Dennis has joined #tp
14:27:02 [DanC_lap]
"very disheartened by the massive gaps". I think we can do better about managing expectations around HTML and CSS schedules. more on that Saturday in the HTML WG meeting.
14:27:04 [raman]
anne -- support existing content all you like -- but as long as that remains the focus, as opposed to say "doing html and css right" --- people will continue to author more junk for you to "support" and that is a death spiral
14:27:20 [MikeSmith]
I'm surprised that Alex Russell is surprised that standards are very good at innovating and inventing.
14:27:21 [raman]
think "support" == "tolerate";-)
14:27:23 [Hixie]
DanC_lap: we're going to adopt my original proposed timetable? :-D
14:27:29 [DanC_lap]
heh.
14:27:34 [hsivonen]
does dojo even work in application/xhtml+xml?
14:27:36 [MikeSmith]
s/are very good/aren't very good/
14:27:46 [DanC_lap]
I do think we should talk about your proposed timetable and mine.
14:27:59 [moshe]
During WWII, the US auto companies had to run a separate production line for the Brits, since the US and Brits had separate standards for nuts and bolts. The US standard was better, btw, from a tech standpoint.
14:28:12 [mjs]
standards are better at standardizing things that already exist than at innovating
14:28:13 [ddahl]
ddahl has joined #tp
14:28:28 [MikeSmith]
Heresy or naivetè?
14:28:30 [mjs]
(and cleaning off the rough edges)
14:28:37 [John_Boyer]
It is impossible to reach for higher goals without a higher platform.
14:28:45 [KevinLawver]
@hsivonen No, it doesn't, because no one in the "real" web uses application/xhtml+xml. Dojo is very pragmatic.
14:28:48 [Hixie]
(CSS2 would be in candidate recommendation as well if CR had existed back then)
14:28:53 [dsr]
dsr has joined #tp
14:29:00 [judy]
judy has joined #tp
14:29:07 [Klaas_Bals]
Klaas_Bals has joined #tp
14:29:07 [anne]
raman, we're actively improving the HTML and CSS standards at the W3C, but there's only so many resources
14:29:11 [aaronlev]
aaronlev has joined #tp
14:29:20 [DanC_lap]
I mostly side with mjs re standards and when to start a W3C WG [when you're ready to share test cases, basically], but I think a mix of more optimistic perspectives are also healthy.
14:29:25 [shepazu]
bingo.
14:29:26 [anne]
(improving them to meet real world constraints)
14:29:30 [tantek]
the processing model of application/xhtml+xml has never been sufficiently defined to provide similar functionality to the de facto processing model of text/html
14:29:33 [DanC_lap]
s/are also/is also/
14:29:34 [shawn]
+1 to open, but not well communicated
14:29:34 [s-mon]
indeed.
14:29:35 [Mez]
John_Boyer, I'm reminded of the stories of my ex-USSR colleagues back in the late 80's. they said in the USSR they carried tapes of their own tools they wrote (editors, compilers) from job to job
14:29:59 [Hixie]
tantek: html5 is working on that. it'll take a while though.
14:30:03 [shepazu]
the process is open, but we need to clarify how to do that... that's a big push I'll be doing with the new SVG Interest Group
14:30:05 [RCutler]
That sounds like stealing to me.
14:30:16 [DanC_lap]
the "there's only so many resources" line is a pet pieve of mine. It means "I'm tired of recruiting" or something.
14:30:28 [tantek]
Hixie, is that html5 or xhtml5? :)
14:30:32 [maxf]
it's a shame we can't complete the [indiscernible] bits in the live transcript
14:30:33 [Hixie]
tantek: yes. :-)
14:30:34 [MikeSmith]
Fun to play the heretic but not terrifically productive
14:30:43 [Hixie]
tantek: but i meant "html 5" the spec.
14:30:44 [FABLET]
join #tp beantown
14:30:55 [maxf]
http://www.captionedtext.com/client/event.aspx?CustomerID=732&EventID=863005&ParticipantId=35e14a7c-868f-4405-b1e6-fb7ec1187b14
14:30:56 [Hixie]
tantek: (as opposed to html5 the serialisation)
14:31:19 [Marcos]
Marcos has joined #tp
14:31:38 [raman]
I said it at the time "amsterdam 2000" and I still say it -- forking out the html mime time and creating application/xml+xhtml was just a major mistake. We should have put in the framework in place using the DOM as it existed to help text/html transition smoothly to clean content.
14:32:10 [Henny]
Henny has joined #tp
14:32:22 [karl]
http://flickr.com/search/?s=rec&q=w3c&m=tags
14:32:23 [tantek]
raman: +1
14:32:29 [shepazu]
w3c needs to do more "marketing" of the spec, including tutorials and working more closely with designers and developers
14:32:33 [anne]
WHATWG FTW!!!
14:32:42 [FABLET]
join #tp [beantown]
14:32:52 [Hixie]
FABLET: you're in #tp already :-)
14:32:53 [jo]
jo has joined #tp
14:33:05 [tantek]
Hixie: +1. (as happened clearly with microformats as well)
14:33:12 [Hixie]
indeed
14:33:16 [glazou]
hey tantek :)
14:33:20 [DanC_lap]
re waterfall... yeah... I made that point in public-html ...
14:33:25 [raman]
Today we're heading down the same death spiral that in the mid-90's eventually caused browsers to become incapable of handling correct content given their kluges to handle bad content
14:33:29 [FABLET]
FABLET has joined #tp
14:33:39 [anne]
raman, not really
14:33:39 [shepazu]
it depends on what you mean by marketing... it can just be a matter of communication
14:34:00 [Hixie]
shepazu: certainly there should be information available, i just mean push marketing
14:34:01 [anne]
raman, we're getting closer to a model that handles both, CSS has had that for years, HTML is getting it now
14:34:07 [brutzman]
brutzman has joined #tp
14:34:08 [tantek]
raman: what makes you say that? from where I sit, we are heading towards browsers that handle more and more semantic content on the Web. a richer experience with every veresion.
14:34:12 [mjs]
to the extent browsers are changing, it's mainly towards more standards-compliance
14:34:21 [moshe]
I dare you to read the Xforms spec and try to figure out, from the intro, just what the spec is supposed to do.
14:34:29 [DanC_lap]
"Concurrent iteration is a bit chaotic, but a strict
14:34:29 [DanC_lap]
waterfall approach seems like a poor use of available resources." -- http://lists.w3.org/Archives/Public/public-html/2007Jul/0378.html
14:34:34 [shepazu]
Hixie: I agree on the whole, but there can be coordination efforts with our members
14:34:43 [Mez]
reading specs is like reading Shakespeare; it's a specialty kind of thing that only comes with practice
14:34:54 [klaus]
klaus has joined #tp
14:34:57 [AnnBassetti]
and the average person won't do it
14:34:57 [Hixie]
shepazu: there can, my argument is that there shouldn't be.
14:34:59 [Mez]
perhaps we should perform our specs...
14:35:04 [KevinLawver]
Do we need to start doing SCRUM in working groups?
14:35:17 [Marcos]
the specs would be 20 times bigger and would take twice as long to make if we have to cater for designers.
14:35:25 [moshe]
Mez, I think I should be able to pick up a spec and figure out what it's supposed to accomplish. I shouldn't have to go to Wikipedia to figure it out
14:35:27 [glazou]
KevinLawver: agreed (about the 3-5 years)
14:35:28 [shepazu]
KevinLawver: then we'd end up with scrumlords
14:35:34 [DanC_lap]
I find studying things like SCRUM worthwhile
14:35:44 [Hixie]
shepazu: otherwise you create a fake bubble of hype and get caught up in it -- and fail to see feedback from the audience when they say the spec doesn't address their needs. we've seen that before in w3c.
14:35:46 [Mez]
why can't there be an overlay to specs for "normal"people (to Marcos comment)
14:35:47 [moshe]
Just one paragraph, Mez, that's all I ask
14:35:49 [KevinLawver]
@shepazu Isn't that what chars are supposed to be anyway?
14:35:52 [mjs]
specs are not software
14:35:57 [KevinLawver]
Sorry, chairs.
14:35:57 [John_Boyer]
Doing scrum in working groups might clarify who the specs are for (and who they're not for). This panel is illustrating the need for a spec for implementers and a primer for users of the implementations
14:36:05 [mjs]
software development methodologies do not necessarily apply
14:36:06 [shawn]
+1 to considering SCRUM like trial...
14:36:08 [Mez]
ok moshe, I'll try to get that one paragraph in WSC
14:36:11 [LeeF]
+1 to spec & primer
14:36:13 [Mez]
what does it need to address?
14:36:25 [moshe]
"This spec allows me to..."
14:36:31 [Mez]
got it
14:36:33 [KevinLawver]
I'd hate to write Molly, Eric Meyer and others out of a job. ;)
14:36:36 [raman]
I must say the IRC chat is more interesting than the panel ... or is it just me?
14:36:37 [moshe]
Or something else that explains the basic concept
14:36:40 [klaus_]
klaus_ has joined #tp
14:36:41 [DanC_lap]
"test cases, you know, use cases" . Amen! story telling and test cases.
14:36:46 [shepazu]
Hixie: agreed, it shoudl be communication of real functionality, not hype... things like real-world conformance matrices to show what works where
14:36:47 [anne]
testcases ftw too, indeed
14:36:55 [MSM]
MSM has joined #tp
14:37:14 [dom]
-> http://en.wikipedia.org/wiki/Scrum_%28development%29 agile software development method for project management.
14:37:18 [DanC_lap]
glazou, my lightning talk is about test cases... vote for it, please.
14:37:24 [shepazu]
test suites are becoming much more important, at least in the Interaction domain
14:37:25 [KevinLawver]
We're going to start doing this in CSS - looking at user-based design and coming up with personas and use cases for the specs to provide a good user-base relevant portal to CSS.
14:37:25 [Hixie]
shepazu: yeah that would be very useful
14:37:27 [glazou]
DanC_lap: sure
14:37:40 [DanC_lap]
ooh... Karl, Oliver, how about that for at tile? "Story Telling and Test Cases"
14:37:50 [karl]
hehe
14:37:57 [dom]
The mobile web test suites working group has set up a test harness that allows to collect results from test suites: http://www.w3.org/2007/03/mth/harness , results at http://www.w3.org/2007/03/mth/results
14:38:06 [clc4tts]
where is the voting anyway?
14:38:19 [s-mon]
but it seems the community is perpetuationg the perception. how do you change a perception that you, yourself, are perpetuationg.
14:38:20 [dom]
(it's primarily targeted at mobile web browsers, but works alright with desktop browsers as wellà
14:38:25 [KevinLawver]
Well, two out of three blames being accurate's not bad. ;)
14:38:41 [karl]
DanC_lap: we still have to do the slides. That will be a lightning Slides Editing from olivier and me, maybe followed by lightning talks :p
14:38:43 [KevinLawver]
I never blame myself. It's always the browser's fault.
14:38:53 [citizenspace]
citizenspace has joined #tp
14:38:57 [shepazu]
Hixie: specifically, conformance matrices that pull together different specs, like HTML, CSS, SVG, etc... so they just have to look in one place, not a dozen
14:39:02 [Bert]
Too often, when people say {W3C, a WG} doesn't listen to users, they mean it doesn't listen to "them* :-(
14:39:10 [matt]
clc4tts, voting for the Lightning Talk Wild Card was going to be done via IRC during the second session of LTs.
14:39:21 [shepazu]
... obviously, with some filtering available.... ;)
14:39:23 [clc4tts]
thanks matt
14:39:27 [s-mon]
Bert, exactly.
14:39:37 [matt]
No problem. Actually, if I could get a volunteer for tabulating the votes I would appreciate it.
14:39:55 [Mez]
does anyone know of usabiity testing that's integrated with recommendaitons in a standard? we're grappling with that in WSC
14:40:11 [Mez]
and have only found "process" oriented UT standards (this is how you should run a UT)
14:40:12 [RCutler]
I question that common goals and consistent priorities are possible or desirable in an organization like this.
14:40:32 [KevinLawver]
Why hasn't anyone just come out and said this is all Microsoft's fault? ;)
14:40:39 [karl]
http://www.technorati.com/search/www.w3.org/2007/11/TPAC
14:40:43 [klaus__]
klaus__ has joined #tp
14:40:52 [glazou]
KevinLawver, because everyone knows about it !-) no need to repeat !
14:41:24 [mjs]
KevinLawver: more polite to say "browsers"
14:41:41 [glazou]
mjs: pfffff
14:41:41 [KevinLawver]
But, mjs, it's not accurate. Is it?
14:42:14 [steph]
Mez: i imagine you're thinking of heuristics, but that's a moving target, isn't it?
14:42:15 [Kai]
Kai has joined #tp
14:42:19 [mjs]
KevinLawver: well, there are compat differences and standards compliance bugs in SafOperFox too, but they are much closer to standards and to each other than to IE
14:42:30 [mjs]
it's unclear how to define 100% support
14:42:31 [glazou]
Yves, on our dead bodies !-)
14:42:35 [raman]
Problem started withnetscape;-) in general the dominant browser has always been buggy and everyone has made the problem worse over time by being bug compatible
14:42:56 [mjs]
is there any non-trivial standard that's less than 20 years old and can claim to have 100% support from anything?
14:43:04 [mjs]
even C compilers have standards compliance bugs
14:43:13 [glazou]
mjs: electricity
14:43:18 [anne]
Yves, when they stop "sucking" and come with millions of testcases this might be feasible :)
14:43:24 [raman]
if gcc were like browsers, we would first ask all failing CS students to submit their programs -- and then build a compiler that compiled all the code from the failing student's ;-)
14:43:28 [Mez]
steph, we're trying to figure out both ways to "verify" that our recommendations will be sufficiently "usable", and also what to tell people about ensuring their implementation of our recommendations are suffiiciently "usable"
14:43:34 [MikeSmith]
It seems like there is maybe an expectation among some users/readers that all specs should somehow be understandable/readable by non-implementors or those who don't understand the implementation complexities ... an unrealistic expectation. The answer is maybe: You're not going to understand some specs unless you have certain prior knowledge that it may likely be you don't have unless you're an implementor or or have worked in some way a browser or other U
14:43:34 [MikeSmith]
A or developed complex web apps or whatever.
14:43:41 [mjs]
glazou: I don't think that meets the "non-trivial" standard
14:43:43 [Yves]
anne: well it more than millions of test cases that are needed ;)
14:43:48 [hsivonen]
btw, the model Molly just referred to keeps the test results secret
14:43:54 [mjs]
raman: gcc is different because what's delivered to end-users is binaries, not source
14:44:18 [mjs]
therefore breaking source compat for standards compliance can be livable, after all, users still have the old binaries
14:44:19 [jo]
jo has joined #tp
14:44:20 [shepazu]
"SafOperFox"... ugh... Mopari, maybe? ;)
14:44:20 [glazou]
mjs: ah ? have you ever been at a CEI standardization meeting about electricity ?
14:44:31 [raman]
gcc delivered as a binary ? wonder what rms would say ...
14:44:32 [citizenspace]
raman: and 10000x more people can write HTML than can write C. surely having a publishing technology that is accessible to more authors is a good thing?
14:44:44 [mjs]
raman: not of gcc itself, of programs it compiles
14:44:52 [mjs]
that's the difference between a complier and interpreter
14:44:52 [anne]
Yves, well, based on impl experience standards prolly have to be fixed
14:45:06 [mjs]
perhaps it's no coincidence that popular interpreted languages have just one dominant implementation and no spec
14:45:23 [glazou]
mjs: standardization in electricity is far more complex than ours, FYI
14:45:26 [shepazu]
CSS is not the whole of W3C
14:45:29 [KevinLawver]
The panel excluded, but a lot of trainers I've seen just don't know the subject. They spread FUD and misinformation instead of teaching the specs, benefits and then work-arounds.
14:45:36 [raman]
it is easier to teach authors simple rules to create correct content, as opposed to today's mess of you can do x, but not y if you've done x ...
14:45:42 [Yves]
anne: it depends on the issue, as always it's only on a case-by-case basis
14:46:03 [mjs]
glazou: I'm sure the standards are way more thorough, but the technical complexity is much less - 120v tri-phase could be described sufficiently on one page with some equations
14:46:09 [anne]
Yves, I suppose, but if the standards don't get implemented they're not really standards imo
14:46:24 [DanC_lap]
re making HTML more accesible to authors, i'm recruiting tutorial writers.. pls see http://www.w3.org/2002/09/wbs/40318/tasks83/results#xtasks
14:46:32 [Yves]
anne: so not even tcp should be considered as a std then :)
14:46:33 [Marcos]
less whingeing, more specn!
14:46:33 [youenn]
youenn has joined #tp
14:46:43 [shepazu]
MikeSmith: I agree that specs aren't usually adequate for users... that's why we need tutorials
14:46:58 [PHB]
PHB has joined #tp
14:47:01 [Marcos]
Shepazu, are you going to write the tutorials?
14:47:07 [moshe]
electrical code standards are thick, thick books. and there are code inspections. what's the lesson from this difference?
14:47:08 [fantasai]
tantek, there's a transcript somewhere, have you found that yet?
14:47:18 [tantek]
fantasai - no not yet
14:47:27 [SeanP]
SeanP has joined #tp
14:47:29 [PHB]
Waterfall models work wonderfully - For consultants!
14:47:36 [Steven]
http://www.captionedtext.com/client/event.aspx?CustomerID=732&EventID=863005&ParticipantId=35e14a7c-868f-4405-b1e6-fb7ec1187b14
14:47:37 [anne]
Yves, I'm not familiar enough with it I'm afraid
14:47:43 [glazou]
mjs: bwahahaha. there's so much more than that. for example, electrical control is relayed on electricial wires, that's also standardized there ; +interferences ; exchange between countries ; +fallback ; ...
14:47:46 [shepazu]
Marcos: I wrote SVG tutorials and built the SVG Wiki for years, so yes :)
14:47:54 [ArtB]
What we need is an entirely new Declarative Format for Applications and User Interfaces (DFAUI). Ooops, just remembered We've Been There, Done That :-)
14:47:58 [fantasai]
tantek: ^Steven
14:48:13 [dsinger]
dsinger has joined #tp
14:48:23 [PHB]
Watefall model means that at the end of the engagement the customer has the choice of paying for more time or only getting an architecture nobody else can implement for their money
14:48:28 [Marcos]
Shepazu will not officially write all tutorials for all working groups!
14:48:29 [mjs]
glazou: well, TCP/IP is also specced and implemented much more thoroughly than web standards, but I guess for anything sufficiently low-level, if you don't get it right the world stops
14:48:31 [Henny]
Henny has joined #tp
14:48:36 [anne]
Yves, but I do think that at least standards applicable to the Web are not set in stone but evolve with usage and implementations
14:48:37 [Marcos]
s/not/now
14:48:37 [mjs]
the world doesn't stop if you have a float rendering bug though
14:48:53 [shepazu]
Marcos: but now that I'm with the W3C, I'm planning on moving that forward through the SVG Interest Group
14:49:01 [KevinLawver]
mjs, your world doesn't stop, but mine does.
14:49:07 [anne]
Yves, how this maps to the W3C Recommendation track is not exactly clear to me
14:49:09 [hsivonen]
I don't think specs can be precise for implementors and tutorials for authors at the same time
14:49:11 [dorchard]
dorchard has joined #tp
14:49:13 [John_Boyer]
But shouldn't the community be responsible for pointing out those ambiguities by actually sending the feedback to the public lists, as they are invited to do?
14:49:15 [s-mon]
that slide is unreadable.
14:49:15 [moshe]
YELLOW on WHITE? who can read this slide?
14:49:19 [Kai]
Kai has joined #tp
14:49:19 [glazou]
mjs: but last time a bug occured in command control at PG&E, 9 western US states had a power outage...
14:49:22 [s-mon]
c'mon!
14:49:22 [mjs]
KevinLawver: it's a lot easier to patch around a browser bug than a power grid bug
14:49:28 [anne]
Yves, then again, all stuff I work on never reached that stage :)
14:49:29 [fantasai]
hsivonen++
14:49:33 [hsivonen]
better to have a separate primer doc
14:49:39 [Marcos]
hsivonen++
14:49:41 [Rich]
slide lacks contrast - accessibility problem :-)
14:49:46 [Marcos]
heheh
14:49:48 [jgraham_]
hsivonen, indeed. You can have specs that make good implementations or specs that are eay to read, not boh
14:49:48 [John_Boyer]
hsivonen++
14:49:52 [shepazu]
primers++
14:49:59 [dsinger]
dsinger has joined #tp
14:50:02 [raman]
but if you patch over a bug, then the bug never gets fixed, and over time you get stuck with buggy content. equivalent to passing all failing students
14:50:15 [Yves]
anne: do you mean that when a std is reasonnably well implemented at time T, then a popular implementation at T+10 years starts implementing the same std badly proves that the spec was bad from day one? It may be then that another std is needed.
14:50:30 [Dennis]
Dennis has joined #tp
14:50:41 [mjs]
the web is not a university class
14:50:45 [mjs]
it is an infosystem
14:50:51 [anne]
Yves, if the T+10 impl significantly impacts deployed content, maybe
14:50:53 [mjs]
the goal is to get information from authors to users
14:50:54 [raman]
today, it's not even up to first grade;-)
14:51:03 [mjs]
not to give out marks
14:51:19 [luis]
luis has joined #tp
14:51:20 [KevinLawver]
Pointing at Dan Connolly, only because he asked for it.
14:51:24 [Yves]
anne: in that case another std defined is ok, saying that old spec was bad is not right as it was proven ok before
14:51:24 [Lachy_]
Lachy_ has joined #tp
14:51:28 [Marcos]
hehe
14:51:36 [moshe]
mjs, impossible to improve process without metrics
14:51:39 [kaz]
kaz has joined #tp
14:51:53 [anne]
Yves, how you solve that problem politically is not really of interest to me
14:52:07 [timbl_]
Molly asks for the TAG (or someone) to make a top-down effort to make more consistency between specs ... but does that mesh with allowing WGs to be more autonomous. ... hmmm . ther is no free coorination. CGs are suppoosed to work in a less centralized way, the TAG watches out for cracks.
14:52:22 [plh]
oh oh, we lost the captioner
14:52:27 [MikeSmith]
amen to what MSM said about if "non-dominant browsers decide NOT to be bug-compatible ... users complain" -- that's the market reality that drives browser vendors ... if a user can get to broken/non-conforming site in browser A but not in browser B, then as far as the user is concerned, browser B is broken and he/she will switch to using browser A instead (at least to get to that site)
14:52:32 [tH]
tH has joined #tp
14:52:57 [Yves]
anne: if the specs were clearer from day one, then those issues would not even be on the table, so there is a big QA effort to do on specs first
14:53:00 [caribou]
which reality? user reality? market reality?
14:53:03 [jose-lap]
jose-lap has joined #tp
14:53:09 [clc4tts]
the captioning seems to have stopped
14:53:24 [glazou]
yep
14:53:25 [Liam]
Liam has joined #tp
14:53:32 [mjs]
moshe: agreed, but the metric must be meaningful
14:53:33 [IanJ]
IanJ has joined #tp
14:53:38 [lbolstad]
lbolstad has joined #tp
14:53:40 [anne]
Yves, specs being correct doesn't mean they get implemented correctly
14:53:49 [beowulf]
beowulf has joined #tp
14:53:57 [Yves]
specs being clearer helps then being implemented correctly
14:54:04 [anne]
it also depends on how you define correct
14:54:07 [s-mon]
it goes a long way.
14:54:13 [anne]
some people think error handling doesn't need to be part of a spec for instance
14:54:14 [ed]
ed has joined #tp
14:54:15 [rigo]
anne: then, the technology is dead and will go back fragmentation = pre-standard
14:54:21 [hsivonen]
specs that aren't compatible with market realities fail
14:54:31 [kenny]
kenny has joined #tp
14:54:31 [DanC_lap]
Bert, did you hear that story about tables and that professor?
14:54:34 [anne]
rigo, I'm not saying we shouldn't have standards, I think
14:54:35 [karl]
http://flickr.com/photos/mollyeh11/1878846606/
14:54:36 [citizenspace]
specs that have too high a barrier to implementation fail
14:54:41 [DanC_lap]
"teach your children well" comes to mind, Noah.
14:54:48 [s-mon]
he he.
14:54:54 [mjs]
is tables vs. css interesting technically as anything but a shibboleth?
14:54:54 [raman]
espiranto anyone;-)
14:55:00 [Marcos]
technology are tools?
14:55:16 [ivan]
The university course situation is really bad overall...
14:55:30 [ivan]
maybe W3C should have a separate arm concentrating on Un curriculae?
14:55:53 [glazou]
MikeSmith: patiently ?-)
14:56:03 [rigo]
anne: but after all you generate remarks like the one from hsivonen, where the market does not come together to get consensus on a solution, but where some magicien runs after some cats to write down where the cats currently are .. ;)
14:56:05 [KevinLawver]
There is also a HUGE difference between: the standard, the implementations, and then the best practices. I think the conversation gets confused when we don't call the three out.
14:56:06 [fantasai]
ivan: it would have to be done in conjunction with the community,
14:56:08 [DanC_lap]
yes, connecting W3C to curriculum development is a goal of mine
14:56:13 [kenny]
Is there any way that the w3c specs could be made easier to read and synthesise for developers and designers? Would listing the books describing the respective spec make a difference?
14:56:35 [anne]
anne has left #tp
14:56:43 [anne]
anne has joined #tp
14:56:52 [ivan]
fantasai: of course. But building curriculae requires a special knowledge and talent, not necessarily present in the WG-s
14:57:10 [karl]
tantek: that would be interesting to see for html 5, I think it is one of the biggest spec ever being written
14:57:13 [DanC_lap]
"real world" is an attack phrase; let's find another way to talk
14:57:19 [anne]
rigo, the market is coming together no?
14:57:29 [s-mon]
i can't stand that term.
14:57:29 [MikeSmith]
AaronGustafson - molly got a mike on her left as well
14:57:30 [fantasai]
ivan: I'm not suggesting the WGs do it. :) That wouldn't work, clearly.
14:57:38 [DanC_lap]
I'd like people to talk about "my experience" rather than "the real world"
14:57:42 [tantek]
DanC_lap: s/real world/URLs to real world examples on the public Web
14:57:44 [s-mon]
who lives n the "real world".
14:57:44 [anne]
rigo, I think I'm missing your point
14:57:51 [Hixie]
i don't think molly's answer answered the question
14:57:57 [hsivonen]
you can't deploy in the real world if browser competition has not resulted in something to deploy on
14:58:28 [rigo]
anne: yes it is coming together, but we are talking about the gap (and time-lag) between market and specification and how to fill that gap
14:58:31 [karl]
"My real world is not yours"
14:58:50 [anne]
rigo, not wait for Recommendation level with implementing
14:58:53 [anne]
we're already doing that
14:59:00 [shepazu]
my real world includes Joost and intranets and set-top boxes
14:59:02 [KevinLawver]
The e-mail problem is a really good case of bad implementations.
14:59:04 [Stuart]
mjs mentions shibboleth.... interesting crack in the floor at Tate Modern, London: http://www.tate.org.uk/modern/exhibitions/dorissalcedo/default.shtm
14:59:15 [rigo]
anne: and whether the market takes up some things designed by committee and whether the committee is taking up something done in the market and I say it's both :)
15:00:00 [raman]
See the following from Lawrence Lessig: Lawrence Lessig: I have been doing this for about two years--more than
15:00:00 [raman]
100 of these gigs. This is about the last one. One more and it's over
15:00:00 [raman]
for me. So I figured I wanted to write a song to end it. But then I
15:00:00 [raman]
realized I don't sing and I can't write music. But I came up with the
15:00:03 [raman]
refrain, at least, right? This captures the point. If you understand
15:00:06 [raman]
this refrain, you're gonna' understand everything I want to say to you
15:00:10 [raman]
today. It has four parts:
15:00:13 [raman]
15:00:16 [MikeSmith]
kenny - If a spec is intended to be a precise and unambiguous functional specification for implementors, it probably can't also be made easier to read for designers.
15:00:16 [raman]
* Creativity and innovation always builds on the past.
15:00:20 [raman]
* The past always tries to control the creativity that builds upon it.
15:00:20 [raman]
* Free societies enable the future by limiting this power of the past.
15:00:23 [raman]
* Ours is less and less a free society.
15:00:26 [raman]
15:00:29 [DanC_lap]
(I guess a certain amount of venting frustration about history is in order, but I hope we get past it soon, Hakon.)
15:00:33 [Kangchan]
Ivan : W3C can not the building curriculae, but can propose the well organized curriculae (just link the existing curriculae pages) for reffering as outreach activitirs of W3C
15:00:41 [DanC_lap]
yes, ACID2 rocks
15:00:45 [hsivonen]
shepazu: Joost is single client. hence, not client interop standardization issues
15:00:48 [steph]
MikeSmith: I don't think the idea is one all encompassing document :)
15:00:49 [raman]
Lessig's articulation is the best expression I've heard of my unease over bending over backwards to support broken legacy bugs.
15:01:16 [Dennis]
Dennis has joined #tp
15:01:21 [raman]
The web is *much* bigger than the browser.
15:01:22 [dom]
oh, I never thought of applying lessig's refrain to our situation with standards, nice, raman
15:01:23 [dglazkov]
dglazkov has joined #tp
15:01:30 [shepazu]
hsivonen: it uses Web protocols, Web formats, and it's free and open to use...
15:01:39 [ivan]
Kangchan: yes. But that would require specialized groups that know how to build curriculae...
15:01:42 [Kai]
Kai has joined #tp
15:01:43 [kenny]
mike: i agree. However, I have always found books much easier to read than specs. if the spec pages were to have a list of books describing the spec, won't it make it easier for everyone?
15:01:43 [klaus__]
klaus__ has joined #tp
15:02:00 [MikeSmith]
steph - actually, I seems for some readers that is the expectation; they want to read the core specs and understand them, and if they can't, they feel "dumb" ... at least that is something that I thought one of the panelists was saying
15:02:19 [hsivonen]
karl: do they not needs browsers to deploy web content in bangladesh and singapore?
15:02:22 [glazou]
Ann++
15:02:28 [raman]
The web was built on http, URIs and HTML that Web is *much* bigger ...
15:02:41 [KevinLawver]
People don't upgrade, which only exaggerates the problem of browsers being slow to implement.
15:02:47 [anne]
huh, no innovation?
15:02:51 [gavin_]
gavin_ has joined #tp
15:02:51 [Marcos]
yes. lets freeze innovation because companies are not agile!
15:02:55 [mjs]
we can't tolerate innovation?
15:03:00 [kenny]
clear
15:03:03 [steph]
MikeSmith: i think it's because at the moment it's not easy to understand which document is target at whom, so when you're a designer coming to the site to know ... how they are supposed to know more :)
15:03:04 [rigo]
anne: this is the past trying to control you :)
15:03:06 [Marcos]
STOP ALL INNOVATION PEOPLE!!!
15:03:06 [mjs]
is this the World Wide Telegraph Consortium?
15:03:20 [kenny]
clear
15:03:22 [raman]
world-wide-waiting consortium
15:03:39 [moshe]
don't shoot the messenger
15:03:40 [plh]
[the captioner lost internet connection, they're trying to set up a new captioner]
15:03:48 [DanC_lap]
I'm glad Anne brings the large enterprise perspective; it's real. But yes, it's only one part of the conversation.
15:03:50 [anne]
maybe Boeing can use curl, that's quite consistent I believe
15:03:59 [DanC_lap]
s/Anne/Ann B/
15:04:00 [KevinLawver]
I hate to say it, but if people implemented their documents the right way the first time, they wouldn't need to mandate an old browser.
15:04:04 [moshe]
the message is: legacy content of 1 zillion web pages can't be re-written every tuesday
15:04:04 [MikeSmith]
kenny - the books are written after the spec, based on the spec ... proper place to list such books is not in the spec itself but at corresponding W3C site for the related working group or whatever
15:04:42 [glazou]
Marcos, mjs: have you any idea what means deploying a next-gen browser into a 140,000 employees company world-wide ?
15:04:45 [plh]
me [the captioners aren't based at the same place]
15:05:06 [mjs]
glazou: do you think an end to innovation would make upgrades less risky?
15:05:06 [Zakim]
+ +1.309.329.aabb
15:05:08 [Zakim]
-Captioner
15:05:14 [rigo]
anne: big corps are part of the market and the picture, so how could we ignore them?
15:05:18 [Bert]
+1 to KevinLawver
15:05:20 [mjs]
if there wasn't any innovation, I guess there wouldn't need to be upgrades at all
15:05:26 [rigo]
...and their needs
15:05:34 [Roland_]
Tantek: not exactly what you asked for, but nonetheless interesting paper : An Empirical Study of Open Standards -- http://web.si.umich.edu/tprc/papers/2007/692/finaltprc.pdf
15:05:44 [glazou]
mjs: I'm just saying software developers, software vendors and users do not live in the same time space
15:05:45 [AaronGustafson]
thank you all very much
15:05:53 [Marcos]
glazou, there are billions of people in the world.
15:05:57 [MikeSmith]
AaronGustafson - thank you all. Great panel
15:06:01 [hsivonen]
the conclusion I draw from the enterprise comment is that new specs must not break compat
15:06:04 [AaronGustafson]
thanks MikeSmith
15:06:11 [steph]
thanks everyone
15:06:11 [AxelPolleres]
AxelPolleres has joined #tp
15:06:14 [Marcos]
Companies that can't keep up can't stop innovation!
15:06:19 [anne]
hsivonen, that works for me :)
15:06:33 [mauro]
===============
15:06:40 [moshe]
marcos, the problems of big companies concenrate and make visible the problems of small companies
15:06:41 [Ralph]
[break -- 24 mins]
15:06:42 [Zakim]
-Ivan
15:06:43 [mauro]
break until 10:30AM
15:06:51 [glazou]
amazing, truely
15:07:09 [Marcos]
tell it to the facebook people
15:07:21 [beowulf]
well that was interesting
15:07:34 [csma]
csma has left #tp
15:08:31 [tt]
tt has joined #tp
15:09:12 [karl]
http://www.w3.org/QA/2007/11/tpac_2007_real_world_perspective
15:09:22 [cblouch]
cblouch has joined #tp
15:10:06 [Martin]
Martin has joined #tp
15:19:20 [KevinLawver]
KevinLawver has joined #tp
15:22:45 [encarna]
encarna has joined #tp
15:25:48 [Zakim]
+ +1.847.925.aacc
15:27:08 [Philip]
Philip has joined #tp
15:27:54 [Zakim]
+Terry_Morris (was +1.847.925.aacc)
15:28:50 [ivan]
zakim, dial ivan-voip
15:28:50 [Zakim]
ok, ivan; the call is being made
15:28:52 [Zakim]
+Ivan
15:30:59 [Marcos]
I wonder why Steven is not on the HTML5 VS XHTML2 panel? Or why none of the (X)HTML Editors are not on the panel?????
15:31:25 [pv]
pv has joined #tp
15:31:49 [mjs]
marie: I thought it was supposed to be AND, not VS
15:32:03 [karl]
Marcos: there is no editors of HTML 5 as well
15:32:04 [Marcos]
um, yeah right :P
15:32:11 [Marcos]
Can we add a seat for Hixie and Steven up there?
15:32:42 [amy]
can anyone point out John Schneider and Matt Womer to each other?
15:33:23 [karl]
Editors:
15:33:23 [karl]
Jonny Axelsson, Opera Software
15:33:23 [karl]
Mark Birbeck, x-port.net
15:33:23 [karl]
Micah Dubinko, Invited Expert
15:33:23 [karl]
Beth Epperson, Websense
15:33:24 [karl]
Masayasu Ishikawa, W3C
15:33:26 [karl]
Shane McCarron, Applied Testing and Technology
15:33:28 [karl]
Ann Navarro, WebGeek, Inc.
15:33:30 [karl]
Steven Pemberton, CWI (HTML Working Group Chair)
15:33:42 [tH]
tH has joined #tp
15:33:54 [jgraham_]
jgraham_ has joined #tp
15:34:03 [Marcos]
Well, can any of those people please be on the panel?
15:34:17 [herve]
herve has joined #tp
15:34:25 [Hixie]
Marcos: i'm happy to let henri and anne do their thing, i don't think there's really anything that needs to be said that they won't say
15:34:53 [hal]
hal has joined #tp
15:35:04 [cvenezia]
cvenezia has joined #tp
15:35:19 [amit]
amit has joined #tp
15:35:26 [glazou]
glazou has joined #tp
15:35:35 [Marcos]
I still think Steven should be up there :)
15:36:27 [mjs]
well this is gonna be a fun panel
15:36:27 [Dennis]
Dennis has joined #tp
15:36:28 [Zakim]
+ +1.972.296.aadd
15:36:40 [j1]
j1 has joined #tp
15:36:43 [brutzman]
brutzman has joined #tp
15:36:44 [DanC_lap]
"since the web exploded in the big bang" ;-) -- Al G
15:36:56 [Marcos]
It be more fun with Steven :)
15:37:13 [DanC_lap]
yeah, that screen is blurry
15:37:15 [LeeF]
LeeF has joined #tp
15:37:27 [Marcos]
hehe
15:37:27 [Kai_]
Kai_ has joined #tp
15:37:37 [glazou]
ivan, rotfl
15:37:40 [Martin]
Martin has joined #tp
15:37:42 [Norm]
Norm has joined #tp
15:37:50 [brutzman]
might the slides be linked on the agenda page?
15:37:52 [Zakim]
-Terry_Morris
15:37:59 [lm]
lm has joined #tp
15:38:00 [Zakim]
- +1.309.329.aabb
15:38:17 [Zakim]
+ +1.309.329.aabb
15:38:46 [glazou]
karl, since you're on last row, can you ask the people behind you to fix the blurry image on left screen ?
15:38:48 [myakura]
myakura has joined #tp
15:39:06 [Kai_]
Kai_ has left #tp
15:39:19 [Kai_]
Kai_ has joined #tp
15:39:19 [raman]
ssh ou
15:39:34 [MikeSmith]
glazou - will ask
15:39:41 [MSM]
MSM has joined #tp
15:39:46 [glazou]
MikeSmith: thx
15:39:52 [Steven]
What I actually said was that Tantek had agreed to do it. It was his action item
15:40:02 [najib]
najib has joined #tp
15:40:12 [steph]
steph has joined #tp
15:40:16 [judy]
judy has joined #tp
15:40:20 [tantek]
Steve - which action item was that?
15:40:21 [clc4tts]
clc4tts has joined #tp
15:40:22 [anne]
those minutes: http://www.w3.org/2004/04/webapps-cdf-ws/minutes-20040601.html
15:40:32 [glazou]
who's the speaker ?
15:40:44 [anne]
Richard S. from IBM
15:40:44 [dbaron]
Richard Schwertfeger (sp?)
15:40:44 [John_Boyer]
Richard Schwerdtfeger
15:40:44 [aaronlev]
glazou: Rich Schwerdtfeger
15:41:08 [raman]
call him sword weielder
15:41:24 [kenny]
clear
15:41:25 [kenny]
clear
15:41:35 [DanC_lap]
glad to hear Rich feels his time collaborating with W3C was well spent!
15:42:00 [kenny]
appologies for the previous two messages. it was a unsuccessful attempt at clearing the screen.
15:42:10 [DanC_lap]
RS: I'd like to see the two groups merge.
15:42:16 [Hixie]
html5 is not looking to tomorrow?
15:42:19 [anne]
WHATWG: http://www.whatwg.org/ W3C HTML WG: http://www.w3.org/html/
15:42:33 [Marcos]
HTML5 should stop living in the past :P
15:43:13 [John_Boyer]
Good to innovate on html, but never understood why not contribute to the W3C working group?
15:43:29 [mjs]
the strongest objections I've heard to HTML5 are to remove the forward-looking parts like <canvas> and <video>
15:43:39 [anne]
John_Boyer, because that proposal was initially denied
15:43:48 [anne]
John_Boyer, during that workshop, it was voted down
15:43:51 [anne]
we're doing it now though
15:43:57 [Marcos]
and I'm sure the offline stuff will probably get objections too
15:43:58 [mjs]
or the parsing spec which ends the death spiral of bug compatibility reverse engineering
15:44:00 [tlr]
tlr has joined #tp
15:44:02 [DanC_lap]
I think the HTML 5 spec does reasonably well re consuming content, but I'll say again I'm recuriting tutorial writers, to help authors
15:44:07 [John_Boyer]
Denied by whom?
15:44:07 [ChrisL]
ChrisL has joined #tp
15:44:09 [glazou]
John_Boyer: you mean the xhtml2 wg ?
15:44:19 [John_Boyer]
yes, which was the html wg
15:44:37 [Rich]
Rich has joined #tp
15:44:38 [karl]
Creating "HTML 5 for authors" will make your life healthier. Please join
15:44:38 [kenny]
is html going to be phased out eventually in favour of xhtml?
15:44:43 [glazou]
do you have any idea how many times we said that the successor to html 4 has to be built upon html 4 itself ?
15:44:54 [MikeSmith]
bravo hsivonen
15:45:01 [mjs]
I already know I agree with Henri :-)
15:45:11 [anne]
that's my excuse too :)
15:45:23 [Marcos]
me three :)
15:45:46 [MikeSmith]
Al: "WAP1 was a scandalous failure"
15:45:50 [raman]
a good reflection of why there has been such poor communication across the different camps in this debate. Each person believes he is right and knows the answer alas
15:45:52 [glazou]
John_Boyer: the problem was not on the contributors' side, but on the WG side
15:46:02 [DanC_lap]
I don't think "html 5 for authors" needs to be a book-length tome; maybe it's planethtml5, an edited journal of authoring techinques. (a la a list apart)
15:46:21 [clc4tts]
the captioning is really bad
15:46:22 [glazou]
raman: belief ? right ? can you please tell me where is xhtml 2 on the web ATM ?
15:46:22 [cwiecha]
cwiecha has joined #tp
15:46:27 [anne]
maybe it's all of that
15:46:31 [John_Boyer]
If one is *within* the working group, then one gets a vote on how it proceeds, so it's only a problem within the WG if you have a priori decided not to join it.
15:46:45 [Rich]
yes. merge the groups and get everyone on the same page. Don't ignore today's browsers, move us forward
15:46:57 [molly]
molly has joined #tp
15:47:06 [Steven]
Ebay UK Mobile is in XHTML2
15:47:15 [Steven]
I am told
15:47:19 [Hixie]
uri?
15:47:20 [Kai]
Kai has joined #tp
15:47:34 [mjs]
Rich: merging groups that disagree is a follow-on to getting on the same page, not a prerequisite for it
15:47:38 [karl]
DanC_lap: an aggregation of content produced by people would be a neat idea
15:47:44 [janet]
janet has joined #tp
15:47:51 [ChrisL]
steven - authored in or served as?
15:47:59 [Steven]
authored in
15:48:11 [Dennis]
Dennis has joined #tp
15:48:17 [glazou]
John_Boyer: http://www.glazman.org/weblog/dotclear/index.php?post/2003/04/14/2880-untitled
15:48:19 [Steven]
then transformed to whatever the device accepts
15:48:19 [Rich]
you won't get them to agree if they are in separate groups and allowed ot not collaborate
15:48:29 [shepazu]
"The Web that goes to people"?
15:48:34 [Hixie]
wait, what was the question?
15:48:36 [tantek]
what does "more orthodoxy" mean here?
15:48:48 [DanC_lap]
(I missed the question; I miss the traditional W3C scribe.)
15:49:00 [glazou]
DanC_lap, agreed
15:49:01 [dsinger]
ortho - straight, doxy - opinion
15:49:11 [raman]
there we go with "real world" again. whose real world?
15:49:12 [John_Boyer]
glazou, nice blog. Join the group, contribute solutions.
15:49:17 [jo]
jo has joined #tp
15:49:22 [DanC_lap]
glazou is a founding member
15:49:32 [glazou]
John_Boyer: that's a joke I presume ?
15:49:39 [KevinLawver]
karl, where would the money come from then?
15:49:44 [MikeSmith]
karl - Let's replace it with "Solve Real Problems"
15:49:50 [s-mon]
hmm, real world again.
15:49:53 [molly]
karl: Is it the nomenclature?
15:50:04 [glazou]
John_Boyer: http://disruptive-innovations.com/zoo/slides/200611-Tokyo/
15:50:07 [jose_ma]
sometimes "real world" sounds like "average user" ;)
15:50:09 [DanC_lap]
glazou hasn't been very loud in the public-html coctail party, bu the's been around
15:50:16 [shepazu]
s-mon, stop breaking the web
15:50:17 [MikeSmith]
Keep it Real
15:50:22 [karl]
MikeSmith: I will solve the real world trouble you are just at my left side. If you hear screaming, I just punched MikeSmith
15:50:59 [molly]
I ask for good reasons - if that way of articulating the issue doesn't express well, that's helpful to know for future conversations
15:51:35 [karl]
molly: it is just that real world has no real shared meaning across Web people
15:51:43 [DanC_lap]
"real problems" ... is that still in the TOC? reminds me of "real world", an attack phrase as I noted above
15:52:06 [molly]
what would be better terminology? Suggestions welcome!
15:52:08 [shepazu]
"don't do error handling at all"?
15:52:11 [KevinLawver]
It's the Ivory Tower vs The Plebes! Let them eat markup!
15:52:13 [Hixie]
yeah tim! you didn't define error handling! :-)
15:52:30 [DanC_lap]
hmm... yes, it's still in the TOC. http://www.w3.org/html/wg/html5/principles/#solve-real-problems
15:52:33 [Bert]
The audience doesn't ask questions, because they are busy with the real world :-)
15:52:41 [shepazu]
Bert++
15:52:47 [molly]
haha!
15:52:48 [mjs]
where "the real world" means IRC?
15:52:59 [danz]
danz has joined #tp
15:53:01 [mjs]
I dunno if I can agree with that definition
15:53:06 [Markus]
Markus has joined #tp
15:53:07 [shepazu]
mjs++
15:53:09 [molly]
how about "practitioners" ?
15:53:09 [kenny]
would it help the conformance to w3c standards if browsers were less fault tolerant with html and css? Would that not force the authors to comply?
15:53:10 [glazou]
the audience has no questions but the speakers are slackers ;-)
15:53:22 [aurelien_levy]
aurelien_levy has joined #tp
15:53:25 [ted]
ted has left #tp
15:53:29 [MikeSmith]
more questioners up at the microphones would be nice
15:53:33 [klaus__]
klaus__ has joined #tp
15:53:34 [tantek]
kenny, no, it would simply raise barriers to authoring = fewer authors, less content
15:53:37 [Bert]
"The real world" is "my world" for every instance of "my" in this crowd.
15:53:41 [mjs]
css has fault tolerance as a conformance requirement
15:54:11 [KevinLawver]
I expected more boxing gloves for this panel.
15:54:13 [shepazu]
SVG has well-defined error handling (though not perfect... still need to fill in the cracks)
15:54:24 [Hixie]
IanJ: for parsing, for instance, the first implementation of the error handling is html5lib, which is not a browser tool
15:54:40 [shepazu]
CSS has error-handling (though not implemented universally)
15:54:45 [molly]
I agree with Tantek on that. There has to be a way to address Mom's desire to post something to her site versus company XYZ's need to have highly secure, manageable documents
15:55:06 [mjs]
validation should be on the authoring end, not the content consumption end
15:55:14 [KevinLawver]
Molly, that's why we need better authoring tool, and have things like HTML tidy.
15:55:18 [tantek]
the "fragile content" attitude has failed on the Web, why are people still pursuing it seriously?
15:55:23 [srv4661]
beantown
15:55:27 [ted]
ted has joined #tp
15:55:30 [Katsu]
Katsu has joined #tp
15:55:36 [molly]
Yes Kevin, but as every man knows
15:55:41 [molly]
one must be the master of one's tool.
15:55:57 [IanJ]
I haven't looked closely at the error-handling bits in html 5.
15:56:23 [mjs]
HTML5 has split document and UA conformance
15:56:26 [IanJ]
Just wondering (in my ignorance) how it is going describing error handling in a way that works across big spectrum of apps.
15:56:27 [shepazu]
tantek: sorry, can you explain the "fragile content" attitude?
15:56:30 [karl]
molly: about terminology. "real world" is a way to ostracize people then it creates a community. More than just talking about "issue to solve".
15:56:36 [tantek]
shepazu: XML
15:56:38 [KevinLawver]
Molly, if that's true, then your mom needs to master the tol...
15:56:41 [KevinLawver]
tool, even.
15:57:04 [molly]
Karl, well, for the misuse of terminology I apologize, but I do think the points remain clear
15:57:21 [molly]
and I have tens of thousands of people to point to as examples worldwide
15:57:29 [Hixie]
ht: again just looking at the parsing/syntax section, http://whatwg.org/html5/section-writing0.html is the language, and http://whatwg.org/html5/section-parsing.html
15:57:29 [IanJ]
Does the error-handling go beyond parsing?
15:57:30 [herve_]
herve_ has joined #tp
15:57:31 [shepazu]
tantek: ah, ok... a little spinny, but point taken
15:57:39 [Hixie]
ht: ...is the error handling algorithm
15:57:51 [Hixie]
IanJ: yeah
15:57:53 [tantek]
fragile = error->stop processing
15:57:53 [IanJ]
ok
15:58:04 [Hixie]
IanJ: though it's more obvious in the parsing section
15:58:19 [IanJ]
thanks Hixie
15:58:37 [aurelien_levy]
hi
15:58:51 [raman]
HST's question is not being answered.
15:58:52 [shepazu]
tantek: yes, that restriction could be loosened up in a defined way
15:58:54 [IanJ]
parsing seems like a good area for generic error handling.
15:58:55 [ChrisL]
as opposed to fragile = silent error recovery and thus error accumulation until one small change breaks rendering mysteriously, type of fragiility
15:58:58 [IanJ]
my guess is rendering is not as much.
15:59:40 [karl]
http://www.w3.org/html/wg/html5/#parsing
15:59:41 [tantek]
ChrisL - one of those has succeeded on the Web, the other has failed.
15:59:44 [Hixie]
IanJ: e.g. search for "invalid value default". there's also cases where the error handling is automaticly implied by the way the algorithms work
16:00:00 [Hixie]
IanJ: e.g. "take the first foo element" handles the error of what if there are two foo elements
16:00:01 [MikeSmith]
couldn't hear the organization who the current questioner is from ...
16:00:13 [Hixie]
IanJ: (as opposed to "take the foo element", which is ambiguous)
16:00:19 [dsinger]
glenn adams, representing samsung (though his own company is xsfi)
16:00:21 [mjs]
IanJ: it's layered - there's a way to parse even erroneous token streams into a DOM, and a way to render even erroneous DOMs
16:00:25 [karl]
interesting the discussion has became meta
16:00:29 [ChrisL]
Tantek - I would argue which one failed. you mean you never edited a page and got an unexpected result that seemed unrelated to what you just changed?
16:00:32 [karl]
about error handling
16:00:47 [karl]
it is almost a general discussion about all W3C technologies
16:00:55 [karl]
s/W3C//
16:01:06 [MikeSmith]
Glenn Adams from Samsung?
16:01:24 [mjs]
MSM: there are ~30 billion HTML documents targeted for display to end users on the public web
16:01:25 [dbaron]
I think the problem was that error handling wasn't defined, not whether or not the error handling was liberal.
16:01:26 [tantek]
ChrisL - any sufficiently complex system will produce unexpected results.
16:01:39 [mjs]
MSM: I think XML is quite a few orders of magnitude off from that
16:01:42 [dom]
+1 to dbaron
16:02:04 [raman]
who will specify the errors in the implementation of the error handling?
16:02:05 [MSM]
mjs, yes -- targeted for one specific use -- and not easily reusable, precisely because people have abused Postel's Law
16:02:41 [mjs]
MSM: "people consuming documents and applications" seems like a pretty broad use to me
16:02:44 [tantek]
it's the difference between the CSS parsing methodology towards error handling vs. the XML parsing methodology towards error handling.
16:03:00 [mjs]
as far as I'm concerned it is *the* use, everything else is implementation details for that
16:03:06 [tantek]
MUST recover vs. MUST fail.
16:03:24 [tantek]
recovery = adoption. failure = abandonment.
16:03:27 [PHB]
If you didn't want xml to resync automatically, why have te start tag repeated in the end tag?
16:03:51 [PHB]
<p> stuff </p> has no point unless you want </p> to allow for resync
16:03:54 [MSM]
[Repeating an off-the-log comment: responding to a remark of Tantek's. I think you get a more pervasive form of fragility from divergent and inconsistent approaches to error recovery -- I note that in the XML community, it has never been necessary to charter a WG to run after the implementations to find out how they do error recovery, so that authors can figure out which errors they must fix and which they can safely leave4 in place.]
16:04:15 [DanC_lap]
re postels law, we just need somebody to take the "be conservative about what you produce" part of the HTML 5 spec and tune it to the authoring audience
16:04:27 [raman]
real world again.
16:04:37 [molly]
"real world" content
16:04:39 [DanC_lap]
and some help from some popular authoring tools like dreamweaver
16:05:02 [raman]
haven't seen imaginary world content ... what does it look like? if you know, then it cant be imaginary;-)
16:05:06 [Liam]
phb - it turns out that </> makes it very ahrd for humans to fix faulty documents
16:05:07 [ChrisL]
phb - you are correct that having the name in the end tag allows better error pinpointing. The aim was to make it easier to fix by the author, not by auto-guessing in the browser
16:05:21 [karl]
DanC_lap: agreed. An authoring document can/should be stricter than the parsing section
16:05:27 [raman]
complex content must have an imaginary part, otherwise it wouldn't be complex, it would be real;-)
16:05:33 [shepazu]
raman: it's the Platonic Ideal of Content
16:05:38 [tantek]
MSM, optimizing for chartering or not a WG, or for adoption of the technology by maximum # of authors? I choose the latter.
16:05:43 [Marcos]
"reality"
16:05:48 [DanC_lap]
no, the technical content should be the same... or the same negotiation, karl. but the document audiences are different
16:06:02 [MikeSmith]
Melinda Grant from HP
16:06:27 [IanJ]
MG: "Will browsers of the future support both html5 and xhtml2?"
16:06:28 [molly]
with a real world question
16:06:31 [molly]
badumpump!
16:06:33 [Bert]
(Not sure </> makes it hard to fix errors, but </foo> makes it hard to avoid errors.)
16:06:35 [IanJ]
MG: "If not, what are the implications?"
16:06:49 [MSM]
PHB, there was a longish discussion of whether to allow empty end-tags, during the design of XML 1.0. The browser vendors eventually got on their knees and begged the WG to specify draconian error handling, because (they argued) the alternative was the same kind of race to the bottom that was visible in HTML.
16:07:16 [karl]
DanC_lap: ;) not sure about that. I can say hey author you should write <strong class="strict">boo</strong>, and say to browsers You have to be ready to understand clumsy people who write <strong class=strict>boo</strong>
16:07:21 [Marcos]
hehe
16:07:22 [MSM]
Tantek, I have never regarded popularity as the true test of truth, beauty, or technical utility. YMMV.
16:07:31 [KevinLawver]
Melinda Grant is my hero.
16:07:42 [tantek]
MSM - evolution would disagree with you.
16:07:55 [tantek]
as would the market
16:08:19 [tantek]
also, popularity = accessibility to more authors.
16:08:28 [MSM]
Tantek, and this means what? Evolution is hardly guaranteed to produce truth, beauty, technical quality, or progress.
16:08:42 [tantek]
evolution has very much produced progress
16:08:46 [jallan]
transcript http://www.captionedtext.com/client/event.aspx?CustomerID=732&EventID=863005&ParticipantId=35e14a7c-868f-4405-b1e6-fb7ec1187b14
16:08:47 [Lachy_]
many of the good ideas in XHTML2 have been introduced in HTML5 already
16:08:51 [tantek]
and produces quality, if incrementally, through iteration
16:09:03 [Lachy_]
the rest can be mostly ignored, IMHO
16:09:13 [DanC_lap]
yes, karl, that's what I'm saying
16:09:16 [Hixie]
Rich: i assure you that all innovations developed by the xhtml2wg as well as other wgs (e.g. xforms) are closely studied by the html5 contributors and we've taken many ideas already (e.g. <section>)
16:09:16 [mjs]
xml events don't reduce JavaScript in the markup afaict
16:09:22 [karl]
tantek: do you think that britney spears is singing beautiful songs. (sorry for the fans of Britney, just taking her as an example)
16:09:40 [ed]
mjs: that's what I thought too...didn't get that statement
16:09:47 [raman]
glad I dont have to listen to music that is "evolved" and especially not as it is "evolving"
16:09:51 [karl]
ChrisL, there is a real transcript
16:09:51 [DanC_lap]
s/Rich:/Rich,/
16:10:04 [DanC_lap]
(i.e. hixie is using direct address, not attribution)
16:10:11 [ChrisL]
I was just hoping the question being asked would be noted
16:10:22 [karl]
DanC_lap: then we agreed, but expressed it in different ways ;)
16:10:22 [tantek]
karl, marketing != evolution
16:10:27 [DanC_lap]
(I hate the fact that IRC defaults to : for direct address, after the chicago manual of style and such set the standard)
16:10:33 [MSM]
Tantek, I think you are confusing "quality" with "success in reproduction". A common fallacy in the nineteenth century among social Darwinists, but I had thought it was exinct.
16:10:44 [tantek]
raman, most music is evolved / built-upon previous music
16:10:56 [tantek]
as is most art
16:11:00 [tantek]
c.f. Picasso
16:11:13 [anne]
anne has joined #tp
16:11:15 [mjs]
art or music that wasn't in some way derivative would just be confusing
16:11:42 [DanC_lap]
if you want to see music built upon previous music, I think I'm going to play guitar and ask people to sing along at an HTML WG unconference break-out
16:12:01 [karl]
eeek discussing philosophy on IRC is like discussing politics drunk in a bar
16:12:03 [MSM]
Is spam really the highest quality of email? It seems to be what email has evolved to, and it's way more common that real email.
16:12:09 [tantek]
DanC - really sorry I would miss that ;)
16:12:29 [karl]
+1 to glazou!!!!
16:12:31 [DanC_lap]
lessig said "I can't sing..." and then told a powerful story; well, I can play guitar, and I've got some excellent singers coming along
16:12:37 [tantek]
MSM email is obsolete ;)
16:12:43 [karl]
editor vendors and CMS vendors on the HTML WG
16:12:51 [KevinLawver]
+10 to glazou
16:13:00 [John_Boyer]
+100 to need for editing tools need
16:13:01 [mjs]
Apple is an editor vendor
16:13:05 [DanC_lap]
yes, CMS folks... I need head/@profile support in CMS tools
16:13:21 [mjs]
Mail, Dashcode, iWeb, Aperture, Safari itself...
16:13:27 [glazou]
mjs: so have your editing solution implement html 5 when safari is ready for html 5 too
16:13:29 [glazou]
:)
16:13:39 [DanC_lap]
ah... that reminds me, mjs... the apple mailer doesn't support hyperlink authoring
16:13:56 [MSM]
ah. It would seem then that evolution and popularity are the test of quality sometimes, but not all the time? Is it off base to suggest they are the test of quality when you like their results, but not when you don't?
16:14:14 [DanC_lap]
what is the test of quality, MSM?
16:14:30 [mjs]
DanC_lap: Leopard version does
16:14:37 [DanC_lap]
gold star, mjs
16:14:39 [mjs]
DanC_lap: I don't know about Tiger, haven't run it in a while
16:14:44 [MSM]
I'll borrow Tantek's approach, and say that for some authors, invalid HTML that makes you care about browsers' error recovery strategies is simply obsolete. You don't have to care about it any more. Freedom is just a decision you have to make.
16:14:51 [mjs]
Edit > Link > Add...
16:14:57 [DanC_lap]
tantek, I hope we get to jam sometime. bummer indeed that you're not here this time.
16:15:08 [Lachy_]
the things Stephen is talking about can be done in HTML5, XHTML2 isn't needed for that stuff
16:15:30 [tantek]
MSM, you actually don't. Author valid POSH and browsers will handle it just fine.
16:15:38 [karl]
mjs: why people from iWeb are not participating to the WG
16:15:44 [Marcos]
why author in xhtml2 to serve as html? why not just use html for the whole process?
16:15:48 [cblouch]
Part of evolution is natural selection where some new branches turn out to be bad and die
16:15:55 [Hixie]
the first hit for "ebay xhtml2" on google is an e-mail i wrote. sheesh. -_-
16:15:57 [glazou]
cblouch: exactly
16:16:06 [mjs]
karl: too busy typing?
16:16:08 [raman]
hear hear!
16:16:21 [MikeSmith]
Marcos - perhaps because it attempts to provide another level of abstraction
16:16:28 [mjs]
karl: I will note that I specifically and the Safari team in general talks a lot to other teams at Apple that do web-technology-relevant things
16:16:29 [raman]
hear hear to what Steven said I meant
16:16:32 [John_Boyer]
Unfortunately, the above comment from Marcos is the same as saying that the things Steven is talking about can be done in machine language, so no need for a higher level language.
16:16:34 [MSM]
Tantek, one question that then arises is "Why have a WG worrying about an obsolete problem that no one actually needs to care about?"
16:16:34 [MikeSmith]
... the last part of what Steven mentioned
16:16:39 [shepazu]
Steven++
16:16:41 [molly]
my question is why isn't he on this panel???
16:16:43 [karl]
mjs: having *real world* authoring tools developers on the group would be more than welcome
16:16:56 [John_Boyer]
+1 molly
16:16:56 [karl]
ooops I did it again
16:17:07 [tantek]
I actually agree with Steven's point. It is worth encouraging various different approaches, and testing them against the market.
16:17:08 [s-mon]
oh karl.
16:17:13 [aaronlev]
someone representing xhtml2 should be on the panel
16:17:14 [s-mon]
:^)
16:17:19 [shepazu]
well stated, Steven
16:17:19 [karl]
;)
16:17:27 [Marcos]
John_Boyer, MikeSmith, I don't disagree, I'm just asking.
16:17:30 [s-mon]
+1 Steven.
16:17:35 [raman]
good question: why wasnt Steven on the panel? it's a symptom of the larger problem within the W3C. we've spent six years standing on the "xml" leg; now, we're standing on the "html" leg, and the xml leg is not fashionable any more ...
16:17:36 [molly]
Karl, I'm going to hit you later. I promise.
16:17:46 [glazou]
karl: is that for me ?-)
16:17:57 [molly]
raman: I asked Steven and he said he had no idea why he wasn't on the panel
16:17:59 [tantek]
MSM, perhaps ask the WG - I think browser vendors care about handling invalid HTML. Modern web designers however, are doing just fine authoring valid POSH.
16:18:06 [John_Boyer]
marcos, Steven answered your question though with the anecdote about reducing development time from 150 person years to 10
16:18:11 [shepazu]
raman: he wasn't on the panel because we wanted this to be a member-driven panel, no W3C Team members
16:18:13 [ChrisL]
molly, I was surprised that Steven was not on the panel, given the title and all
16:18:32 [ChrisL]
... although DanC is also not on the panel
16:18:32 [Marcos]
There are plenty of chairs up on the stage, maybe he could go and sit up there?
16:18:51 [mjs]
defining how to handle invalid content is important given the huge amount of existing content on the web and the need to support casual authors
16:18:52 [MikeSmith]
molly - original plan was to not have anybody from the W3C team on the panel
16:18:53 [karl]
glazou: you are one of the rare on the group
16:18:55 [John_Boyer]
shepazu, Roland Merrick is co-chair and not a team member
16:18:58 [anne]
ChrisL, supposedly it's about people in the industry
16:19:01 [aaronlev]
shepazu: why isn't there anyone at all representing xhtml2 though
16:19:02 [karl]
not enough unfortunately
16:19:06 [mjs]
you have to support these things as a loss leader for supporting standards-compliant content
16:19:07 [John_Boyer]
perhaps you should have invited him and Chris Wilson
16:19:17 [MSM]
Raman, it's OK, many XML people are used to making technical decisions based on their technical judgement, and not worrying overmuch about fashion. (Not all, of course, and it's not much fun to be held up as among the World's Worst Dressed. But we can bear being unfashionable, because the technology works for us.)
16:19:20 [raman]
I think the view Steven represents is best articulated by him and losing him because he is a team member is just wrong.
16:19:23 [Bert]
I'm sure if Steven were on the panel, people would complain about it being always the same people on the same panels.
16:19:29 [glazou]
karl: apple has an editor, msft has web expression, adobe has dw
16:19:32 [shepazu]
John_Boyer: I mean for this panel, not all of them
16:19:32 [Marcos]
John_boyer, who is you?
16:19:41 [dbaron]
Those public sites may claim to be XHTML, but they're actually serving XHTML as HTML.
16:19:46 [John_Boyer]
Forms WG Chair and XForms editor
16:19:52 [John_Boyer]
Also not a team member...
16:19:52 [MikeSmith]
was a mistake to not have Steven on the panel I think
16:19:56 [shepazu]
aaronlev: because the guy who was going to be up there had to cancel
16:19:58 [karl]
glazou: yes, but not represented in the group unfortunately
16:20:10 [raman]
if W3Cwere dynamic and agile, we would add Steven to panel *now*;-)
16:20:15 [glazou]
karl: old story
16:20:19 [shepazu]
Rich is on the XHTML2 WG
16:20:19 [raman]
as someone pointed out there are chairs there.
16:20:19 [John_Boyer]
+1000 Raman!
16:20:27 [MSM]
raman++
16:20:29 [molly]
Go Raman!
16:20:38 [aaronlev]
i agree with raman for once
16:20:42 [Marcos]
Steven, get up there ;)
16:20:43 [karl]
I want dreamweaver (adobe) on the HTML WG.
16:20:48 [raman]
Go Steven Go
16:20:50 [ted]
ted has left #tp
16:20:58 [ChrisL]
Remember that we are missing a panel member (who uses xhtml2 in his business to server to multiple mobile targets) so it was intended to have a commercial, industry user
16:20:59 [glazou]
karl: ping our old friend sho kuwamoto ?
16:21:19 [cwiecha]
or just have all these panel participants sit down and have a conversation among the community
16:21:32 [Bert]
dynamic, agile, without focus and unprepared, yes :-)
16:21:39 [karl]
glazou: I don't know sho kuwamoto but would be happy to be put in contact.
16:21:54 [glazou]
karl: but he could give us a name
16:22:02 [ChrisL]
karl, I know sho but haven't seen him around in nearly a decade
16:22:24 [tantek]
sacrificing accessibility of authoring for accessibility of reading = fewer people authoring = fail.
16:22:34 [Lachy_]
as an author, ARIA isn't a good long term solution. It's a quick fix, band-aid solution
16:22:48 [ted]
ted has joined #tp
16:22:48 [karl]
Matt May (Adobe) will try too
16:22:52 [glazou]
ChrisL: I pinged him a while ago, he's still around
16:22:53 [ted]
ted has left #tp
16:22:57 [Marcos]
I'm going to go home and write a ruby on rail app that creates xhtml2 and then use XSLT to transform it to HTML :P
16:23:03 [raman]
Agree with lachie for a change, ARIA is a papering over the pig solution
16:23:03 [shepazu]
Lachy_: I don't think you understand ARIA, then
16:23:17 [aaronlev]
Lachy_: it's a little more complicated than that
16:23:23 [steph]
Lachy_, raman: is it better not to have a solution at all? :)
16:23:30 [ChrisL]
Marcos, a similar app that converted xhtml2 to html5 would actually be a very nice thing. Care to give it a go?
16:23:33 [raman]
history teaches that such patches survive longer than they deserve to
16:23:38 [ted]
ted has joined #tp
16:23:42 [mjs]
ARIA is the analogue of low-level accessibility API
16:23:48 [anne]
yeah
16:23:54 [mjs]
many high-level toolkits provide built-in AX
16:23:55 [Lachy_]
I understand ARIA just fine. It's just not something that's really well designed for authors to use
16:24:04 [anne]
which is problematic, because authors in general don't care
16:24:08 [aaronlev]
raman: we should remove as much of the need for ARIA as we can, but we still need it
16:24:16 [Keeper]
Keeper has joined #tp
16:24:18 [anne]
makes sense to support it anyway though, at least as short-term solution
16:24:30 [raman]
today's ARIA will lead to solutions that smell of twenty blind men feeling an elephant and concluding different things.
16:24:42 [mjs]
being able to customize standard controls more than currently is important
16:24:45 [klaus]
klaus has joined #tp
16:24:47 [gavin_]
is it just me, or is the right-channel on the audio feed distorted?
16:24:50 [Kai]
Kai has joined #tp
16:24:57 [dorchard]
dorchard has joined #tp
16:25:09 [ted]
ted has left #tp
16:25:13 [ted]
ted has joined #tp
16:25:20 [aaronlev]
raman: so tell google to stop making google web toolkit, it's not accessible
16:25:25 [mjs]
with Cocoa there is much less need to break out of the standard controls than with HTML
16:25:33 [ted]
ted has left #tp
16:25:37 [ted]
ted has joined #tp
16:25:55 [FabienG]
FabienG has joined #tp
16:26:12 [cblouch]
Are a preponderance of "authors" creating widgets from scratch and therefore need to understand ARIA implementation?
16:26:20 [raman]
aaron, as long as accessibility defines itself as "stopping people doing useful things" you'll fail. I dont live in that world.
16:26:21 [Steven]
+1 to fantasai re constraints
16:26:33 [ted]
ted has left #tp
16:26:34 [ted]
ted has joined #tp
16:26:44 [aaronlev]
raman: you can innovate and be accessible
16:26:55 [Lachy_]
having constraint-based conformance in the spec would be useful for authors, not for implementers
16:27:02 [steph]
development by constraints would probably achieve more consistent end results
16:27:11 [dbaron]
I think constraints vs. algorithms depends on which is simpler / easier to understand for the particular case.
16:27:12 [mjs]
it's hard to write enough constraints to be as specific as an algorithm in terms of conformance requirements
16:27:17 [steph]
there could always be suggested algorithms
16:27:22 [mjs]
depends on for what
16:27:23 [Hixie]
dbaron: indeed
16:27:27 [matt]
The voting for the lightning talk wild-card slot is now open: http://www.w3.org/2002/09/wbs/1/TP-LT-2007/
16:27:42 [Stuart]
constraints are requirements on alogorithms ie. an expression of what the algorithm is intended to accomplish (maybe)).
16:27:54 [matt]
matt has changed the topic to: W3C Technical Plenary Day - agenda at http://www.w3.org/2007/11/07-TechPlenAgenda.html - lightning talk voting: http://www.w3.org/2002/09/wbs/1/TP-LT-2007/
16:27:55 [dbaron]
In the particular case of table width balancing, it was simple to describe as a linear combination of bracketing value sets.
16:27:55 [MikeSmith]
matt - is mjs still eligible for wild-card slot for those talks?
16:28:15 [matt]
It's been filled up for weeks and weeks...
16:28:20 [MikeSmith]
OK
16:28:23 [matt]
okay, maybe not weeks and weeks,but a while now :)
16:28:42 [Bert]
Constraints is a way of thinking that some people like, others can't get their head around it. We've got both constraints and algorithms in the CSS spec, dpeending on who wrote that section...
16:28:57 [mjs]
MikeSmith: I dunno if there's enough coffee in this place to make up for my sleep deficit in any case
16:28:59 [IanJ]
[Discussion of "view source" principle]
16:29:26 [Liam]
mjs, you are not alone :)
16:29:32 [s-mon]
Steven's points on abstraction and extraction were presented at apacheconeu 2007 - http://www.w3.org/2007/Talks/05-02-steven-apachecon/ .
16:29:37 [tantek]
it's not a matter of "should". it's a matter of how "view source"able a technology affects its adoption.
16:29:43 [DanC_lap]
oh! speaking of mjs and slots... will you demo this CSS cool stuff in the "short talks" to the whole HTML WG on Thu, mjs?
16:29:48 [amit]
amit has joined #tp
16:30:00 [mjs]
DanC_lap: I can do that, though it won't be strictly on-topic
16:30:04 [cwiecha]
quality of this day so far is way down relative to last year
16:30:09 [mjs]
though making demos made me think of things that could be useful in HTML
16:30:18 [Hixie]
mjs: oooh, do tell us
16:30:26 [Hixie]
(ideally in the talk and then by e-mail, not here!)
16:30:26 [mauro]
=============
16:30:28 [mauro]
Session 4: Lightning Talks
16:30:31 [mauro]
=============
16:30:36 [DanC_lap]
it's sufficiently on-topic for my tastes, mjs. I'm sure my co-chair will agree. cool.
16:30:40 [Lachy_]
how long will the lightning talks go for?
16:30:44 [glazou]
someone has my lightning talk countdown page on stage ?
16:30:46 [mjs]
Hixie: mainly a way to have toggles that expose a pseudo-class for CSS styling but can contain arbitrary content
16:31:03 [mauro]
Lachy_, scheduled to last half an hour
16:31:12 [mattmay]
mattmay has joined #tp
16:31:17 [DanC_lap]
how much time, mjs? is 20 minutes good? more? less?
16:31:20 [mjs]
Hixie: because :hover and :active cover a lot of animation triggers without script, but toggle is also a very common case that shouldn't require script
16:31:24 [mjs]
DanC_lap: less!
16:31:27 [dom]
vote for the last lightning talk at http://www.w3.org/2002/09/wbs/1/TP-LT-2007/
16:31:30 [DKA]
DKA has joined #tp
16:31:32 [mjs]
but I can try to throw some video in the demo maybe too
16:31:39 [dom]
go directly to