- From: Jonathan Rees <jar@creativecommons.org>
- Date: Tue, 2 Mar 2010 16:59:13 -0500
- To: www-tag@w3.org
Draft minutes available online as
http://www.w3.org/2001/tag/2010/02/25-minutes.html
and below in text form. -Jonathan
----------------------------------
[1]W3C
[1] http://www.w3.org/
- DRAFT -
Technical Architecture Group Teleconference
25 Feb 2010
[2]Agenda
[2] http://www.w3.org/2001/tag/2010/02/25-agenda
See also: [3]IRC log
[3] http://www.w3.org/2010/02/25-tagmem-irc
Attendees
Present
Timbl, noah, DanC, Raman, John_Kemp, Ashok_Malhotra,
Jonathan_Rees, Ht, jar, DKA, Larry
Regrets
Chair
Noah Mendelsohn
Scribe
Jonathan_Rees
Contents
* [4]Topics
1. [5]Convene
2. [6]Administrative items
3. [7]ISSUE-33 & ACTION-332: Mixing SVG & MathML in HTML 5
4. [8]ISSUE-41 & ACTION-395: TAG Response on HTML
Decentralized extensibility
5. [9]Face to Face Meeting Agenda Preparation
* [10]Summary of Action Items
_________________________________________________________
<inserted> Scribenick: jar
<inserted> Scribe: Jonathan_Rees
<trackbot> Date: 25 February 2010
Convene
<DanC> +1 approve [11]http://www.w3.org/2001/tag/2010/02/18-minutes
[11] http://www.w3.org/2001/tag/2010/02/18-minutes
<johnk> +1 to approve the minutes
RESOLUTION: approve minutes of Feb 18
Administrative items
(F2F agenda discussion)
noah: Writing quarterly TAG status, input welcome
(Noah discussing today's agenda)
ISSUE-33 & ACTION-332: Mixing SVG & MathML in HTML 5
[12]http://lists.w3.org/Archives/Public/www-tag/2009Nov/0012.html
[12] http://lists.w3.org/Archives/Public/www-tag/2009Nov/0012.html
<timbl> The HTML 5 design puts MathML elements in the MathML
namespace
<timbl> (and likewise SVG) with no syntactic cost to authors:
danc: Opinions on the HTML WG design for MathML and SVG?
<timbl> Given that the algorithm as far as I can see can be
separated into a quite generic one, and a bit of metadata which
defines which element names trigger which namespcaes, and given that
it generates an XML DOM basically, this seems a nice design.
timbl: Good feature: it's not specific to HTML. So there is
potential for follow-your-nose.
<DanC> "There isn't a general-purpose mechanism
<DanC> for mixing other UI-related namespaces in the spec, but any
mechanism
<DanC> of that sort that should come along should be consistent with
<DanC> the HTML 5 design, IMO."
<Zakim> noah, you wanted to ask how this issue is different from
dist. extensibility discussion?
Noah: what I still trip over is who decides what these elements are,
how they get updated
... there's no distributed extensibility story [but maybe there
could be]
<noah> Dan, we'll have you scribe another time.
danc: DKA, Do you know about XBL or XBL2?
DKA: negative
<Zakim> ht, you wanted to make two points: text/html vs.
application/xml+html; extensibility aspects of this design
<noah> NM: I pointed out that the lack of a syntactic hook for
distributed extensibility bothers me. Otherwise, it seems OK.
ht: my 2nd point is what Noah says, there's centralised
extensibility, not distributed extensibility
... But also it's another wedge between html and xhtml
serialization. [...] doesn't work
<noah> Raman is saying people depend on common prefixes. Maybe. Not
sure I completely believe that.
raman: with this design pattern, a particular way of doing things is
being pushed
<timbl> <svg:html><html:math><html<mi>x<//html:ma>
<DanC> (no; don't go into namespace binding stuff... at least: don't
go into stuff that doesn't address the *User Interface* aspects;
screen real-estate, event bubbling, etc.)
ht: a casual reader would think that there are these two magic
elements <math> and <svg>. once you get to one of these you're using
mathml or svg. this reasoning would be faulty
<timbl> Other cross-over points? Please elaborate
ht: the price is quadratic
danc: you have to think about html inside svg, etc. - all
combinations
<ht> ht: OK, DanC, so, yes, cubic already, and likely to grow
timbl: might be possible that extension points are common, hub
architecture ...
larry: the plugin interface is an example of an extensibility
mechanism that ... dom ... [scribe slow]
<timbl> The plugin interface needs to specify a set of elements
which it adopts.
<timbl> If you use this for plugins then you get search path madness
within a few steps.
<DanC> (binding elements to plug-ins... that's how XBL works, yes?
does anybody here know?)
larry: extensibility has many aspects. graphic contexts, user
requests re caching or security, dom
... looking at requirements for plugin interface is a good way to
think about extensibility
... mathml and svg are about rendering, but if that's all you're
looking at you may be missing things
<masinter> different kinds of extensibility: rendering
extensibility, DOM integration extensibility, security integration,
state integration
timbl: So (you're saying) looking at the plugin API [requirements]
is a good test.
<masinter> math and svg integration also share events, not just
rendering
raman: how do you write a processor that is extensible over time...
doesn't solve the problem of how svg calls back to the host
container... that's an unsolved problem
<ht> Tim, for multiple magic, see
[13]http://dev.w3.org/html5/spec/syntax.html#parsing-main-inforeign
[13] http://dev.w3.org/html5/spec/syntax.html#parsing-main-inforeign
timbl: you mean it just hasn't been written yet
<DKA> We tried to address this problem in the CDF working group,
FYI: [14]http://www.w3.org/2004/CDF/
[14] http://www.w3.org/2004/CDF/
noah: There are people invested in this kind of generality, e.g.
microsoft rendering html on surfaces
<DanC> (yes, CDF was chartered to solve the general problem...)
noah: To do this you need an integrated rendering model that all the
pieces buy into
raman: authoring, rendering, eventing all need to be coordinated
timbl: This was done in Amaya, years ago, but i'm not sure whether
you could apply a shear to something with html in it and have that
shear affect the embedded HTML
<DanC> hmm... interesting example tim just gave... a graph shear
(sp?) ... whether it applies to HTML... and whether hit detection on
HTML buttons in there work
<ht> See also the example in
[15]http://dev.w3.org/html5/spec/syntax.html#foreign-elements
[15] http://dev.w3.org/html5/spec/syntax.html#foreign-elements
<DanC> is that different from the example in my msg? looking...
masinter: part of the css/html to embedded object interface is
[scribe lost]
noah: They closed this, but left distributed extensibility open
<DanC> "html-svg-mathml State: CLOSED" --
[16]http://www.w3.org/html/wg/tracker/issues/37
[16] http://www.w3.org/html/wg/tracker/issues/37
ht: The situation is appalling. Don't think we should push back on
this. SVG and MathML WGs have accepted it
... There will be DOM incompatibilities because attribute names are
lowercased
danc: DKA, what about the CDF experience?
DKA: (compound document format WG)
<DanC> [17]WICD Core 1.0 W3C Candidate Recommendation 18 July 2007
[17] http://www.w3.org/TR/WICD/
dka: WICD didn't have implementations
[scribe sorry for missing Dan's summary. in a noisy room]
DKA: the approach was that there would be separate DOMs ...
raman: but they have consistent eventing, so you could graft
dka: but there was a security issue, so we decided not to do that
noah: let's wrap up
ISSUE-41 & ACTION-395: TAG Response on HTML Decentralized extensibility
<DanC> action-357?
<trackbot> ACTION-357 -- Henry S. Thompson to elaborate the DPD
proposal to address comments from #xmlnames and tag f2f discussion
of 2009-12-10, particularly wrt integration with XML specs and wrt
motivation -- due 2010-03-16 -- OPEN
<trackbot> [18]http://www.w3.org/2001/tag/group/track/actions/357
[18] http://www.w3.org/2001/tag/group/track/actions/357
noah: Deadline is March 23
ht: we've seen no evidence of any constituency inside the htmlwg
interested in this. pushing from outside would probably be fruitless
noah: Being on record is of some value even so, sometimes
ht: Any of these proposals would be an improvement on the current
draft
noah: (Discussion of TAG relation to change proposal preparation)
<noah> I thought I'd heard particular flaws raised with at least
some of the proposals. Is it not appropriate for the TAG to get to
the bottom of whether that's true?
raman: Why should TAG take a position on one option vs. another? Put
all 3 on the table
<DKA> +1 to Raman's suggestion.
<johnk> yes, also +1 to Raman's suggestion
ht: I don't have time to do what Noah asked, that is, to compare the
strengths and weaknesses of the status quo and the three proposals,
but I do have time to put in a change proposal. Encourage other
proposal authors to do the same
<Zakim> johnk, you wanted to ask whether we can simply send an email
to WG pointing out the set of proposals we know about
<Zakim> noah, you wanted to ask about flaws in the designs
johnk: We've done work here, we could say we looked at the three,
couldn't someone in WG take a look at them
noah: don't think we should appear to endorse one if we think it had
a serious flaw.
... did our discussion peter our because of our criticisms?
raman: The proposals got shouted down, so authors went away
noah: didn't we say "but it doesn't do this or that"?
<Zakim> DanC, you wanted to ask what suggests the HTML WG hasn't
looked at these? The trick is to get a critical mass of support,
which involves somebody who's interested to at least
danc: I don't think the WG hasn't looked at them. The trick is to
get a critical mass of support, someone to code them up
<ht> Boy, setting the bar for _proposals_ at having an
implementation is pretty high
danc: When you said no support in the WG - don't know what happened
to Microsoft - and there was someone else
<Zakim> ht, you wanted to disagree with NM
danc: the wg never adopts something unless someone's convinced to
code it up
<DanC> (WG participants who own code and see a proposal that they're
not willing to code up regularly reject the proposal.)
ht: In the discussion about all 3 proposals, as in the discussion of
the status uo, there was consideration of strengths and weaknesses.
... but I don't think any flaws are fatal. pros and cons are
contextual
<DanC> ("rough consensus and running code", no?)
<ht> Describes the IETF. . .
ht: the community that ought to be looking at this, is the HTML WG
<DanC> and the W3C, largely.
<masinter> "running code" is given a priority, yes, but not
"shipping code"
<DanC> the code doesn't have to get written before a proposal is
adopted, but implementors have to be *willing* to code it up
ht: I would like noah, for the TAG, to write to Liam Quin and the MS
proposal authors, to ask them to get a change proposal in
<Ashok> +1
<ht> +1
+1
<DanC> well, what i meant by "prototype in the context of a shipping
browser" is "running code that's integrated with code that handles
the complexity of the modern web"
<johnk> +1
<bubbles> 1+
<DKA> to what?
DKA, see above 'ht: I would like"
<DKA> +1
<masinter> +1 but note requirement may not to get CP in by deadline
<ht> trackbot, status?
<ht> ACTION to Henry S to draft emails for NM to send to HTML WG
chairs and to Liam+MS authors encouraging a change proposal wrt
distr. extensibility by 23 March
<trackbot> Sorry, couldn't find user - to
<DanC> ACTION-356?
<trackbot> ACTION-356 -- Noah Mendelsohn to work with Carine Bournez
to schedule followup meeting on xmlnames followup discussion -- due
2010-02-20 -- OPEN
<trackbot> [19]http://www.w3.org/2001/tag/group/track/actions/356
[19] http://www.w3.org/2001/tag/group/track/actions/356
<ht> ACTION Henry S to draft emails for NM to send to HTML WG chairs
and to Liam+MS authors encouraging a change proposal wrt distr.
extensibility by 23 March
<trackbot> Created ACTION-396 - S to draft emails for NM to send to
HTML WG chairs and to Liam+MS authors encouraging a change proposal
wrt distr. extensibility by 23 March [on Henry S. Thompson - due
2010-03-04].
Face to Face Meeting Agenda Preparation
ht: Request time at F2F to talk about domain name permanence
<DanC> ACTION-351?
<trackbot> ACTION-351 -- Henry S. Thompson to look into a workshop
on persistence... perhaps the June 2010 timeframe -- due 2010-03-16
-- OPEN
<trackbot> [20]http://www.w3.org/2001/tag/group/track/actions/351
[20] http://www.w3.org/2001/tag/group/track/actions/351
<DanC> web apps architecture product
[21]http://www.w3.org/2001/tag/group/track/products/7
[21] http://www.w3.org/2001/tag/group/track/products/7
raman: Web app arch. As background, look at webapps [?] and geo
<noah> [22]http://www.w3.org/TR/hash-in-uri/
[22] http://www.w3.org/TR/hash-in-uri/
raman: Next steps on hash-in-uri
ashok: What is status of our issue on geoloc/geopriv? Open or not?
<DanC> (we don't have an issue on geoloc/geopriv)
ashok: ... we've been contacted by IETF
DKA: I think it would be useful to talk about geo in context of the
wg rechartering
<DanC> ACTION-380?
<trackbot> ACTION-380 -- Daniel Appelquist to draft response to
Fredrick, short and to the point. Larry to review. -- due 2010-02-17
-- OPEN
<trackbot> [23]http://www.w3.org/2001/tag/group/track/actions/380
[23] http://www.w3.org/2001/tag/group/track/actions/380
DKA: also consider going back to the WGs to ask for changes
action ashok Frame F2F discussion on geolocation and geopriv, with
help from DKA
<trackbot> Created ACTION-397 - Frame F2F discussion on geolocation
and geopriv, with help from DKA [on Ashok Malhotra - due
2010-03-04].
<DanC> action-380: done 3 Feb
[24]http://www.w3.org/mid/C78FAD46.92F4%25daniel.appelquist@vodafone
.com
[24] http://www.w3.org/mid/C78FAD46.92F4%25daniel.appelquist@vodafone.com
<trackbot> ACTION-380 Draft response to Fredrick, short and to the
point. Larry to review. notes added
<scribe> scribenick: noah
JAR: I'm working on issue-57 and maybe issue-62
... I think there will be another draft of the HTTP semantics note.
<DanC> ACTION-201?
<trackbot> ACTION-201 -- Jonathan Rees to report on status of AWWSW
discussions -- due 2010-03-02 -- OPEN
<trackbot> [25]http://www.w3.org/2001/tag/group/track/actions/201
[25] http://www.w3.org/2001/tag/group/track/actions/201
JAR: 2) possible other work on ISSUE-57 (address bar on
redirections, etc.)
<DanC> ACTION-282?
<trackbot> ACTION-282 -- Jonathan Rees to draft a finding on
metadata architecture. -- due 2010-03-10 -- OPEN
<trackbot> [26]http://www.w3.org/2001/tag/group/track/actions/282
[26] http://www.w3.org/2001/tag/group/track/actions/282
JAR: 3) metadata architecture (ISSUE-62) action to draft finding.
Won't know until the 8th.
<DanC> ACTION-282 due 8 March
<trackbot> ACTION-282 Draft a finding on metadata architecture. due
date now 8 March
<DanC> action-390?
<trackbot> ACTION-390 -- Daniel Appelquist to review ISSUE-58 and
suggest next steps, due 2010-03-03 -- due 2010-03-03 -- OPEN
<trackbot> [27]http://www.w3.org/2001/tag/group/track/actions/390
[27] http://www.w3.org/2001/tag/group/track/actions/390
<DanC> (chinese menu expansion: scalabilityOfURIAccess-58)
<jar> issue-58?
<trackbot> ISSUE-58 -- Scalability of URI Access to Resources --
OPEN
<trackbot> [28]http://www.w3.org/2001/tag/group/track/issues/58
[28] http://www.w3.org/2001/tag/group/track/issues/58
DKA: I could do an overview of mobile issues (widgets, etc.)
... Should device access APIs (e.g. camera) be Javascript and/or
REST?
<jar> scribenick: jar
dka: Competing architecture proposals in the DAP WG. one from
Google. Javascript vs. REST
noah: I'm interested in mobile, let's talk
<scribe> scribenick: noah
LM: I'll be at TAG F2F into Thurs morning.
<jar> masinter: IRIEverywhere issue
<DanC> issue-27?
<trackbot> ISSUE-27 -- Should W3C specifications start promoting
IRIs? -- OPEN
<trackbot> [29]http://www.w3.org/2001/tag/group/track/issues/27
[29] http://www.w3.org/2001/tag/group/track/issues/27
<DanC> action-343?
<trackbot> ACTION-343 -- Larry Masinter to discuss petname
application to IRI spoofing in public-iri and www-tag -- due
2010-02-25 -- OPEN
<trackbot> [30]http://www.w3.org/2001/tag/group/track/actions/343
[30] http://www.w3.org/2001/tag/group/track/actions/343
LM: I want to talk at F2F about IRI everywhere.
NM: And what would we actually talk about wrt IRIs
LM: On reflection, probably not much. But we could talk about things
like whether to close the issue.
<jar> masinter: Hoping that by then there may be more clarity on the
IRI issue. How do we go about closing it
LM: I'd like to prepare an action plan for resolution.
<scribe> ACTION: Larry to prepare plan for resolving issue-27
IRIEverywhere for F2F discussion [recorded in
[31]http://www.w3.org/2010/02/25-tagmem-irc]
[31] http://www.w3.org/2010/02/25-tagmem-irc
<trackbot> Created ACTION-398 - Prepare plan for resolving issue-27
IRIEverywhere for F2F discussion [on Larry Masinter - due
2010-03-04].
JK:
<jar> johnk: We need further discussion on sniffing
<DanC> (HT an LMM have the actions on sniffing
[32]http://www.w3.org/2001/tag/group/track/issues/24
[32] http://www.w3.org/2001/tag/group/track/issues/24
<DanC> )
<DanC> ACTION-386?
<trackbot> ACTION-386 -- Larry Masinter to review
draft-barth-sniff-4 and send comments, cc TAG -- due 2010-02-25 --
OPEN
<trackbot> [33]http://www.w3.org/2001/tag/group/track/actions/386
[33] http://www.w3.org/2001/tag/group/track/actions/386
<jar> . ACTION johnk to frame F2F discussion on sniffing
<DanC> trackbot, status?
<scribe> ACTION: John to prepare F2F discussion of sniffing, being
sure to check status of other pertinent actions [recorded in
[34]http://www.w3.org/2010/02/25-tagmem-irc]
[34] http://www.w3.org/2010/02/25-tagmem-irc
<trackbot> Created ACTION-399 - Prepare F2F discussion of sniffing,
being sure to check status of other pertinent actions [on John Kemp
- due 2010-03-04].
<DanC> action-355?
<trackbot> ACTION-355 -- John Kemp to explore the degree to which
AWWW and associated findings tell the interaction story for Web
Applications -- due 2010-03-10 -- OPEN
<trackbot> [35]http://www.w3.org/2001/tag/group/track/actions/355
[35] http://www.w3.org/2001/tag/group/track/actions/355
<DanC> action-355 due 8 March
<trackbot> ACTION-355 Explore the degree to which AWWW and
associated findings tell the interaction story for Web Applications
due date now 8 March
<DanC> action-352 due 8 March
<trackbot> ACTION-352 Integrate whiteboard drawings into a prose
document about ways to distribute applications due date now 8 March
JK: I may have progress on ACTION-355 and/or ACTION-352 for the F2F
DC: I've been doing some stuff on redirection. Will coordinate with
JAR.
<DanC> ACTION-368?
<trackbot> ACTION-368 -- Dan Connolly to write up version change
ontology as blog item -- due 2010-02-26 -- OPEN
<trackbot> [36]http://www.w3.org/2001/tag/group/track/actions/368
[36] http://www.w3.org/2001/tag/group/track/actions/368
TBL: some interest in persistent domains, but probably can't do any
preparation
DC: I might manage to do some tabulator stuff
<DanC> HTML+RDFa, Microdata, and Canvas in HTML WG Tim Berners-Lee
(Thursday, 25 February)
[37]http://lists.w3.org/Archives/Public/public-html/2010Feb/0871.htm
l
[37] http://lists.w3.org/Archives/Public/public-html/2010Feb/0871.html
DC: Tim, do you want to discuss some of your responses to the HTML
WG?
NM: Tim, do you want to spend F2F time on this?
TBL: Don't know yet. Depends in part on wishes of the TAG.
Any objection to me adjourning now?
<DanC> +1 adjourn now.
TBL: I think we should push for distributed extensibility.
<DKA> ciao
We are adjourned.
<masinter> distributed extensibility requires orthogonality of
specifications
<masinter> without it you don't have modulairity
<jar> gotta go, sorry
<timbl> Fred Brooks, Mythical Man Month
big committee -> big language
[38]http://www.w3.org/2001/tag/2010/02/18-minutes#item02
[38] http://www.w3.org/2001/tag/2010/02/18-minutes#item02
<raman> n
Summary of Action Items
[NEW] ACTION: John to prepare F2F discussion of sniffing, being sure
to check status of other pertinent actions [recorded in
[39]http://www.w3.org/2010/02/25-tagmem-irc]
[NEW] ACTION: Larry to prepare plan for resolving issue-27
IRIEverywhere for F2F discussion [recorded in
[40]http://www.w3.org/2010/02/25-tagmem-irc]
[39] http://www.w3.org/2010/02/25-tagmem-irc
[40] http://www.w3.org/2010/02/25-tagmem-irc
[End of minutes]
_________________________________________________________
Minutes formatted by David Booth's [41]scribe.perl version 1.133
([42]CVS log)
$Date: 2010/03/02 21:51:58 $
[41] http://dev.w3.org/cvsweb/~checkout~/2002/scribe/scribedoc.htm
[42] http://dev.w3.org/cvsweb/2002/scribe/
Received on Tuesday, 2 March 2010 21:59:48 UTC