"Future of the Web Page" session, Banff WWW Conference

09 May 2007

Track Schedule - Session Info and Talks' abstracts - Raw session log

Table of Contents (session timeline)

  1. Panel Intro
  2. Talk: HTML Reloaded (Chris Lilley)
  3. Talk: Next Step for Forms (Dave Raggett)
  4. Talk: CSS, 10 years after (Bert Bos)
  5. Talk: Ensuring Accessibility of the Developing Web (Michael Cooper)
  6. Talk: Enriching The Web App Model (Arun Ranganathan)
  7. Panel: Chris Lilley, Dan Connolly, Arun Ranganathan, Dave Raggett, Bert Bos, Michael Cooper
  8. Panel Discussion: Innovation? History and Future of HTML
  9. Audience Question: Flash, linking
  10. Audience Question: SMIL, XHTML2, HTML
  11. Panel Discussion: Versioning, Ease of adoption
  12. Audience Question: W3C's messaging toward non-developers
  13. Audience Question: on the probabilty of unforeseen uses, spec architecture

Session Transcript

with thanks to Mike Smith for scribing, with help from IRC participants.

Panel Intro

Scribe: Session on "Future of HTML" starting at 3:30 in Cascade. It's the hip place to be.

Session Description

abandon whatever other sessions you're in and come over here instead or forever regret it.

olivier thereaux (chair for the "Future of HTML" session is introducing the session now

Panel: Chris Lilley, Dan Connolly, Arun Ranganathan, Dave Raggett, Bert Bos, Michael Cooper

<Bert_lap> You're forgot to mention the panel's chair: Olivier Thereaux, the panel chair with the coolest glasses.

Talk: HTML Reloaded (Chris Lilley)

[Chris Lilley coming up to podium now]

Chris's slides: HTML Reloaded

"99.99999% of the Web was invalid HTML. W3C pretended that didn't exist. This isn't a workable solution."

ChrisL: there were benefits, if you had valid XHTML, you had a consistent DOM
... was a benefit, but really it wasn't that great ...

<Mez> nomination for understatement of the track "Chartering HTML and Forms... Probably the most complex rechartering at W3C"

<Bert_lap> Bert_lap is now known as Bert_Bos

ChrisL asks for show of hands, "How many people here are invited experts in the HTML WG"

3 hands go up

<Mez> they all didn't need this session :-)

<DanC_lap> [[

<DanC_lap> # 403 group participants,

<DanC_lap> # 403 in good standing,

<DanC_lap> # 56 participants from 20 organizations

<DanC_lap> # 347 Invited Experts

<DanC_lap> ]]

<DanC_lap> -- http://www.w3.org/2000/09/dbwg/details?group=40318&amp;public=1

ChrisL: "Pretending to base HTML on SGML is no longer funny"

Talk: Next Step for Forms (Dave Raggett)

[on to next talk, from Dave Raggett, on Forms]

DaveR: one of the problems of forms is that for more than 10 years now, it's embarrassing now to think of some of the mistakes we made ... now time to fix them

Dave's Slides: Next Step for Forms

DaveR asks how many people are comfortable with writing regular expressions

not many hands go up

<ChrisLilley> regexp is a write-only syntax

<Mike> heh

<Mike> one-way cipher

<dajobe> isn't that xslt?

<Mike> you mean xquery

DaveR: despite what some feel, XForms has been a big success ...
... in enterprise ... but not with browser vendors
... not all of us are skilled programmers, not good at scripting

DaveR trots out his Spreadsheets 2.0 hobbyhorse

DaveR: need to give the Web back to users for creating applications

<Steven_> :-O

Talk: CSS, 10 years after (Bert Bos)

[next speaker: Bert Bos on CSS]

Bert's slides: CSS, 10 years after

Bert: people do want to print from browser and have it look good ... footnotes, x-refs, other stuff

<DanC_lap> there's a public WD of mathml3 out recently, yes?

<DanC_lap> 27 April 2007: Mathematical Markup Language (MathML) Version 3.0 http://www.w3.org/TR/2007/WD-MathML3-20070427/

<ThomasT> * ThomasT notes that there is a poster presented about math in pure CSS+HTML in this conference (today in the CC)

Bert: we have been talking about math in HTML pages since at least 1996 (confernence in Darmstadt) and it still doesn't work ... but we are working on it

Talk: Ensuring Accessibility of the Developing Web (Michael Cooper)

[next speaker: Michael Cooper, to talk about accessibility concerns]

<Mez> Darmstadt was WWW3 in April 1995

Michael's slides: Ensuring Accessibility of the Developing Web

<Mez> * Mez feels old in web years

<Bert_Bos> Yeah, I remember places better than dates :-)

<Mez> you can find them all for the series at http://www.iw3c2.org/conferences/

<DanC_lap> Darmstadt was fun. I'll never forget the difference in style of the Apple and Microsoft keynotes (Alan Kay, Thomas Reardon)


<Bert_Bos> That makes us, how much? 48 web years? :-)

<Mez> I threw out that grey umbrella within the last year (I should have savedit for web history!)

<Bert_Bos> Re mathml3: yes WD is out, and also WD of mathml-for-CSS

<Steven_> Steven_ is now known as Steven__

<ChrisLilley> I never got funding to go to WWW2 or 3. But I was at 1,4,5,6,7,8,9,10,11,12 and then stopped going for a bit

<Mez> * Mez notes that Steven has a long tail...

<dajobe> ChrisLilley: I found a pic of you and timbl from boston95, come find me and I'll show you

<Steven__> Steven__ is now known as Steven_


<ChrisLilley> * ChrisLilley asks dajobe to wave irl if in the w3c track

<Bert_Bos> Long tail... Steven 2.0? :-)

<Mez> 1 was before my time, and I missed 5 and 7 (due to conflicts and budget respectively), but was on the PC for 5 (Hopgood; greatest organized program chair ever) and steering for 7, so I feel involved with all but the 1st

<dajobe> * dajobe waves in irc from another room watching GeoTracker

<ThomasT> Bert_Bos: have a look at http://www2007.org/posters/poster1015.pdf

<Mez> Steven, the niche market?

<ChrisLilley> the first and second were in a race to be first, of course :)

<Mez> I totally missed that back room aspect back then

<Bert_Bos> * Bert_Bos looks at ThomasT's link...

MichaelC: Web 2.0 is a paradigm shift similar to shift from DOS to Windows

[in terms of accessibility concerns]

<Steven_> * Steven_ hopes that Mez is trying to get WWW2 back online

<Mez> we must learn to graduate from training to tools

<Bert_Bos> * Bert_Bos thanks ThomasT, will try and find the author.

Talk: Enriching The Web App Model (Arun Ranganathan)

Arun's slides: Enriching The Web App Model

<Mez> ... just like CHI did at least 15 years ago

<Mez> foo; have no idea what happened with www2

[Olivier jokes that Arun was invited to speak because he works for AOL, which is sponsoring the WWW 2007 conference]

<Mez> and I had a poster there...

Arun: I have deep, empathic bond with the W3C ...

[Arun is wearing W3C T-shirt]

<ChrisLilley> arun has swallowed the kool-aid

[Arun talking about various widget environments]

<ChrisLilley> content is getting modular

<Mez> * Mez gives props to AOL and other sponsors, who enable things like the high quality chocoholic break we just had

<Steven_> Mez, it used to be at http://www.ncsa.uiuc.edu/SDG/IT94/Proceedings/WWW2_Proceedings.html

Arun talking about how webpages consist of stuff pulled in (aggregated etc.) from all the hell all over the place

Arun: how many of you build Flash apps?

[no hands?]

<Mez> trying to think who was involved. Hardin of course. I've asked steering what happened (I'm sure we went looking, since the link isn't broken, it's gone on iw3c2.org)

<ChrisLilley> few hand go up, yay!

<Steven_> Wayback hasn't got it: We're sorry, access to http://www.ncsa.uiuc.edu/SDG/IT94/Proceedings/WWW2_Proceedings.html has been blocked by the site owner via robots.txt.

Arun: Future of webapps will likely include many instances of proprietary [crap] messed in with [good] stuff based on standards

* MikeSmith is editorializing a bit

<Bert_Bos> Making it more objective? :-)

<ChrisLilley> all those proceedings are lost in time like ... tears, in rain

Arun: some of you have heard of Ajax ...

<Bert_Bos> Next year's conf should not only have Web History, but also Web Poetry. Featuring ChrisLilley as guest of honour.

<DanC_lap> WWW2 proceedings have gone 404? harrumph. what good is all the web history foo here at the conference if we can't keep the flargin proceedings online.

<ChrisLilley> its not mine. its a bladerunner quote

<Mez> could reconstituting www2 be an interesting web history wiki like project?

<Steven_> too right DanC

<Bert_Bos> (22C3 - great conference, b.t.w. - had a poetry track)

Arun: web-apiS group working clipboard (copy/paste) and file-upload APIs ...
... also DOM3 events, Selectors API ...
... and network API gives you, say, an TCP/IP connection object ...

<Bert_Bos> (23C3 was the name of the latest, actually)

<Mez> btw, don't know if you remember, but for www2, submissions were abstracts. I don't remember if papers came out afterwards, or just abstracts and presentations.

[Arun closes]

Panel: Chris Lilley, Dan Connolly, Arun Ranganathan, Dave Raggett, Bert Bos, Michael Cooper

Speakers are joined by Dan Connolly

[time for panel]

olivier: discussion among the panelists

Panel Discussion: Innovation? History and Future of HTML

olivier: I was asking if there is still innovation in the realm of the webpage ...
... so question to ChrisL

ChrisL: what we need is to say that CR = "is it implementable"
... REC = "go out and use this now"

MichaelC: innovation will happen whether we as standards people do anything about it not
... what is important for us is to stay on top of it ...
... and make sure that accessibility is addressed asap ...

<schepers> .. needs to be top-down and bottom-up

<ChrisLilley> no, i still think it should go to rec, which says its passed CR. But then we need a Web Standard later, with a second CR-like period

DanC_lap: we need an HTML standard that levels the playing field [baseline of interoperability]

Arun talks about previous attempts to "rescue HTML from the clutches of the W3C"

DaveR: fixing problems of scripting and CSS are issues ...

<ChrisLilley> Dave: bigger problems with scripting and CSS than with HTML

DaveR: [we always need HTML even if we have higher-level authoring languages for generating HTML]

<ChrisLilley> bert refers to his postultimate slide

Bert: maintenance is essential

damn, chaals at the mic

Chaals: we need both [innovation and maintenance?]
... we have all individually been implementing some of this stuff (clipboard, etc). based on black magic and guessing what each other (browser vendors) have done
... [so we need standards for interoperability]

[ChrisL makes comment

chaals: to those of you have been "dragged out of the museum" to come here, what would have done differently?
... [now that you know what mistakes you made]

ChrisL: why did we have to so this SGML things?

<Bert_Bos> Naming and shaming. Yes, I'm doing that and have always done that, but can only be ahead of the implementations by a little bit, otherwise people will blame my site instead of the browsers.

ChrisL: I often wonder what would have happened if we had specified the syntax we had actually been using [not imposing SGML on top of it]
... which is what we are finally getting around to doing now [in the HTML WG]

<Mez> wikis irritate me for the same reason; no relative links

Audience Question: Flash, linking

<Mez> asked about the inability to link into Flash "do something! do something!" (a quote from, I think, Comedy of Errors)

<ThomasT> Jacek Kopecki, Deri Innsbruck

DanC: Flash is great for some things, but if you use it for things that HTML is better for [you are making a poor design choice]

<Bert_Bos> (Flash is only useful at leaf nodes on the Web.)

Arun: can't use semantic stuff in Flash -- for example, microformats

<schepers> * schepers notes that it's not up to W3C to make technical decisions for Flash... that's up to Adobe

Arun: W3C should be a place for those who are interested in "massive interoperability"

ChrisL: session states is what you are asking for and something that the W3C should be working on ...
... [Chris gives example of "playing with a webapp for 20 minutes, then you want to link to it, but you can only currently link to the start point, not the 20-minutes later point]

Audience Question: SMIL, XHTML2, HTML

Attendee: Do you see a future for SMIL?
... What is relationship between the HTML WG and XHTML2 WG ...
... what is your recommendation to Web authors about that? (what should we use) ...

[sound of can of worms opening]

Chris suggests that Steven Pemberton and DanC should answer that.

Steven: XHTML2 was designed on input from many communities

<JacekK> * JacekK voices apologies to Arun for calling him "the AOL guy"

Steven: trying to find an architecture for addressing a variety of problems ...
... and putting the solution into a single package ...
... it has been adopted in DIAL because it is an extremely good authoring format ...

<Bert_Bos> SMIL is doing fine, version 3 is under development. But for wider use, we need some good and free audio and video formats, and W3C doesn't make those.

Steven: one for which you don't necessarily need browser support ...

DanC: I am much less ambitious

<Bert_Bos> The Ogg formats look promising though.

DanC: my goal is how to specify the stuff you are already seeing fly by on the wire

RhysLewis: DIAL is my fault ...
... we decided what we wanted to do was come up with a "clean" authoring language ...
... in this model, what authors write is not what goes to the device ...
... I think at the moment, the power of XHTML is that it provides the basis for that clean authoring language ...

[attendee not very satisfied with that answer]

<ChrisLilley> attendee wonders, if it can be transformed, why use xforms

Attendee: You can't transform the complex features of XForms to HTML forms

Rhys: I'm sorry, you're wrong. You can.

<ChrisLilley> answer is that what you transform to is deeply ugly (but works)

<dajobe> * dajobe points at http://oatmealstout.wordpress.com/2007/05/09/www2007-dave-raggett-on-next-steps-for-html-forms/ onwards for some summaries of these talks

<DanC_lap> after the muck-a-muck reception, I bumped into somebody who seemed to be at the heart of video codec standardization. I wonder if I wrote down the name.

Panel Discussion: Versioning, Ease of adoption

[Olivier talking]

olivier: relaying question from panelist who could not make it today (S. Troeth from WASP) - multiplication of techologies, concurrent versions. Hard to advocate and teach.

<Mez> * Mez chuckles

Bert: in CSS, there is really only one standard that matters, and that is the latest version
... but users only care about what is the latest one that actually works

<ChrisLilley> low water mark and high water mark?

<DanC_lap> (er... w.r.t. advocacy and advice, saying "it's just a communication problem" seems goofy)

Bert: We tryed to use the word "level" instead of "version" to help make things more clear

Chris: relates to what I was talking about earlier with how we need something more between CR and Rec

<ChrisLilley> but it didn't, really

Arun: versioning on specs is bane of anybody who's trying to figure out what to do
... use a good authoring tool ...
... so you are then insulated from that problem ...

RickEfromUofWashington: I was involved in WCAG 2, where we did a lot of reviews of draft ...
... made me wonder how are we doing on process ...
... is the current W3C process "doing the job"

DanC: there was a time when the 200 people in this room would have been "it" ...

[that would have been the whole set of people you needed to convince of anything

scribe: now it's a certain amount of "shoot in the dark" ...
... because of hugeness of community ...

ChrisL: I'm not quite as pessimistic as DanC ...
... it's not all volunteers ...
... there are people in many orgs who are paid [full time] to work on standards-related activities

DanC: it works well when W3C gets involved at the right time ...
... example: GRDDL, which is still a relatively small community ...

<Mez> the way that businesses convince is a lot of people's to have marketing and sales, get feedback from customers, and charge them :-)

Chaals: we are the guys in the middle [between low-water mark and high-water mark]
... I quite like XHTML2 ...
... (for reasons StevenP mentioned) ...

<ChrisLilley> actually, you guys are lifting the low water mark

Chaals: on the other hand, we (Opera) make a browser for people to use ...
... we address what /users/ want ...

<ChrisLilley> html 3.2 is submerged. below the low water mark. hopefully it will drown

Chaals: it's a popularity contest ...

StevenP: I think it is difficult for browser builders ...
... there is not a good economic model for building a browser ...
... so I think you can't blame them for that ...

<Bert_Bos> Charles seems to imply that browsers (or at least Opera) are not the driving force for innovation anymore...

StevenP: though I do remember hearing one guy say that XSLT would only go into his company's browser "over his dead body" ...
... and it's in there now ...

<ChrisLilley> Steven notes that Hakon is not in fact dead. Despite XSLT being in Opera

* MikeSmith says damn right to that ... XSLT++

<Klaus_> Klaus_ is now known as Klaus

<Steven_> * Steven_ notes that he didn't actually name the person, CHris did ;-)

<ChrisLilley> "naming and ...."

Audience Question: W3C's messaging toward non-developers

[VirginiaFromTorontoStockExchange asking a question]

<Steven_> (For the record: I didn't name the person, despite the record above)

Virginia: As a product manager and usability expert, I find that your (W3C) communication is not geared to me ...

<Mez> * Mez loves that comment

Olivier: W3C has historically been about getting specs out the door and hoping people are there waiting to do something with it

DaveR: the new HTML WG has many public participants and volunteers ...

<Mez> any comm team in the house?

DaveR: hopefully some people within that group will volunteer to help with improving communication

Virginia: Realize that there are people keenly working within the Web space who are not developers.

<Mez> it's the "why it's like this" that's really valuable

<DanC_lap> (usability is another thing that's squeaking into the W3C program of work. esp. the new security context WG.)

Audience Question: on the probabilty of unforeseen uses, spec architecture

NoahM from IBM and W3C TAG asking a question

<Mez> for example, I was just reading the "allow cross site reading" spec, and Art had said it was supposed to enable specific use cases. I really wanted to read some examples, in some other document, linked from the spec (since they're not spec material)

<Bert_Bos> Several W3C Offices also have excellent material, summaries, etc.

Noah: I think it's harder that it's harder than we are letting on ...

<mauro_banff> [yes, Mez, 5 in the podium + Olivier, and at least 4 or 5 on the audience]

<ChrisLilley> I find that if the spec isd too terse and formal, that developers second guess and 'improve'; its important to say why a choice was made

Noah: I worked on Lotus and well understand the problem of not letting architecture getting too far ahead

[somebody please help with logging what Noah is saying]

<Mez> * Mez not getting it yet; sorry

Noah: [example of soundness of GET design]
... to just at the right time notice the few places where, if you don't take the trouble now to make better architecture ...

<ChrisLilley> i think noah is saying, allow unforseen uses. pages authored for humans being harvested for bots. otherwise it closes out innovation

<ChrisLilley> I *think* that is his point

Noah: you are locking into limited architecture that is not going to be extensible later ...

[we are closing the session]

<Bert_Bos> About misuse of GET and POST, that is not clearly explained to developers in a hurry.

olivier: discussion continues on the QA weblog

olivier Thereaux - Session Chair, with many thanks to Mike Smith - Scribe