TAG Weekly -- 19 Jun 2008

Convene

<Stuart> scribe: Dan Connolly

agenda review... TVR asks that we make sure to give tagSoup considerable time

<DanC_> 12 Jun minutes

RESOLUTION: to approve 12 Jun minutes

SKW: propose to meet 26 Jun as usual... nominate DO to scribe, but can't confirm, so NDW?

RESOLUTION: to meet again 26 Jun, NDW to scribe

AM: meeting 3 July?

SKW: haven't looked forward that far; might meet with XRI TC then

News and New Items: widget: URI scheme

SKW: I note www-tag discussion of widget: URI scheme

NDW: I'm interested in a suggestion from LMM about addressing into zips/packages

<Stuart> From: http://lists.w3.org/Archives/Public/www-tag/2008Jun/0090

<Stuart> Regardless, we need the TAG's support resolving this

<Stuart> issue sooner than later.

[missed some...]

some discussion of itunes:

and ado:

<timbl> daap: ?

TVR: one end-to-end architecture is: (1) start with HTTP, (2) get a media type, (3) get a namespace. But due to software bugs, sometimes a URI scheme is easier to deploy

<timbl> I am interested in us getting involved

<timbl> :)

<DanC_> (is that WG chartered to do a new URI scheme?)

several show interest in getting involved

SKW: ok, I'll let them know we're interested to get involved

<scribe> ACTION: stuart respond to Marcus re widget: URI scheme [recorded in http://www.w3.org/2008/06/19-tagmem-irc]

<trackbot> Created ACTION-162 - Respond to Marcus re widget: URI scheme [on Stuart Williams - due 2008-06-26].

scalabilityOfURIAccess-58

<Stuart> http://lists.w3.org/Archives/Public/www-archive/2008Jun/0126.html

TimBL: question is, if the W3C systems team says "for the good running of the web, please install these catalogs"... is that good?
... I'm concerned that some catalog systems use public identifiers rather than URIs

<Zakim> Norm, you wanted to say that supporting XML Catalogs ought to be sufficient

NDW: XML catalogs seems sufficient; SGML catalogs would be overkill...
... and while XML catalogs allow public identifiers, it seems reasonable to stick to URIs

<raman> as google's AC rep, I'd like the W3C to do something here because we and many large companies get periodically locked out of the W3C site because some piece of test code inadvertantly hammers the W3C site in the process of implementing a spec

<Zakim> ht, you wanted to ask what the point I'm missing is

<timbl> The namespace name defined by this Standard is "urn:oasis:names:tc:entity:xmlns:xml:catalog".

DanC: in the past, the cache key for catalogs wasn't a URI; but catalogs now have URI support; any cache mechanism that uses URIs as the cache key is OK by me

ht: what am I missing? what good is a catalog without...

DanC: the catalog comes along with the cache hit data

<Zakim> Norm, you wanted to suggest that Ted should build a sample catalog as he might distribute and to volunteer to review it.

<timbl> I think he should build THE calalog

<Norm> ACTION: Norman to coordinate with Ted to build a sample catalog [recorded in http://www.w3.org/2008/06/19-tagmem-irc]

<trackbot> Created ACTION-163 - Coordinate with Ted to build a sample catalog [on Norman Walsh - due 2008-06-26].

<Zakim> timbl, you wanted to express concern about the use of ur: all over the spec

<timbl> <catalog xmlns="urn:oasis:names:tc:entity:xmlns:xml:catalog"

NDW: yes, the catalog spec pre-dates my change in opinion from urns to http URIs for xml namespaces

<Norm> Norm changed his mind, http://norman.walsh.name/2004/03/03/266NorthPleasant

HT: bummer, that it starts urn: but the software we're interested to exploit knows that namespace name

<timbl> Maybe we should serve http://legacy.w3.org/urn/oasis:names:tc:entity:xmlns:xml:catalog

tagSoupIntegration-54

SKW invites TVR to summarize last week's discussion http://www.w3.org/2008/06/12-tagmem-minutes#item06

TVR: a provocative question, following ht's attempt to get namespaces into HTML design: does it make sense to work toward convergence of XML and HTML? [tvr quickly gives lots of context at a rate the scribe fails to match]
... [explains 2 extreme positions and some middle ground; scribe would like help recovering them]

TimBL: thanks for the summary... reminds me of discussion about RDFa; the spec sticks to XHTML, but a lot of the practical deployment looks like it's in HTML. [?]
... likewise, I'd expect that if good tags are added to HTML, they'd get adopted in XHTML
... there's an issue in that in HTML, there just aren't any namespaces. But you can get around that by saying the root HTML namespace is implicit, based on the media type

<Stuart> action-145?

<trackbot> ACTION-145 -- Tim Berners-Lee to add public prose around his slides at the AC meeting to make the case for extensiblity and flexible XML -- due 2008-06-26 -- OPEN

<trackbot> http://www.w3.org/2001/tag/group/track/actions/145

TimBL: some people are going to need some things from HTML and some things from XML, and keeping those separate isn't going to help
... pls continue that action; I do hope to get to it

TVR: I agree RDFa is an example of something with bottom-up motivation that works pretty well, techinically; but attempts to integrated it into HTML 5 don't seem to be received well
... [more interesting stuff that consumes too much of the scribe's brain for him to summarize in real time]

<Stuart> Raman pose two questions to encourage convergence: what changes to namespaces could be accomodated; likewise for XML. In each case also consider who benefits from what changes (IIRC)

<Zakim> Norm, you wanted to waffle in the same general space as TVR

[scribe missed some from TVR and TimBL]

NDW: I've been thinking/writing about the situation with HTML and XML; I can see technical rationale for convergence, but I don't see that the respective communities have sufficient motivation

TVR: I think this fork in the web is a really bad thing... one possibility is that the two languages come together such that when you're writing, you don't think of two languages but just one
... but there's also... let's call it the cohabitation question
... since we haven't asked the co-habitation question, things from one language tend to pollute the other, usually to the detriment of the XML side

<ht> The cohabitation question is precisely the question we started this issue with: "Is the indefinite persistence of 'tag soup' HTML consistent with a sound architecture for the Web?"

TVR: so I think it will help to think about how these two things can sit side by side on the web
... much/most Atom/RSS feeds aren't valid

<ht> I just tried two random atom feeds, the first was well-formed, the second had an un-escaped & . . . :-(

<Norm> :-(

<ht> +10 to DC about innovators: I just spent the morning with a startup who just got 2nd round funding who have a very cool, with customers, xhtml+css+xml schema web site management tool

<Zakim> Stuart, you wanted to mutter about open/closed language syntax and extensibility v interoperability

<ht> We'll trade you namespaces lite for document.write :-)

<timbl> I understand that the XML whitespace parsing rules are an external parameter given to the parser., an not available anywhere in a follow your nose way. This seems to be a big XML bug. It seems o me to be that the HTML pasring rules are in a similar place.

<Zakim> timbl, you wanted to talk about the white space xml mparser flag having 3 values

discussion of design constraints on RDFa which may be due to problems in other specs or bugs in software...

in particular, whether whitespace is preserved in XSLT, and getting at namespace info in XSLT

<Zakim> raman, you wanted to point out xhtml + css is a hands-down winner with authors and to add html tag soup is pushed by the browsers who have the implementations

<Stuart> ack raman:

<Zakim> raman:, you wanted to add that document.write is the actual dividing issue -- not surface syntax

TVR: also, much of the tag soup started as clean XML in the author's hand, but then downstream tools emit tag soup
... various syntax fix-ups are manageable, but document.write is fundamentally different.
... document.write tangles things up... that's why HTML 5 is so contorted

<Norm> It seems to me that AJAX replaces the need for it, but...

NDW: document.write is too far deployed to say "don't do that"?

HT: it's in google ads and such

DanC: TV gave me some hope that's not permanent, that it's manageable over years

TVR: yes... much of the stuff that was necessary for interop in the 2001-2003 timeframe isn't necessary in current browsers, but there's a feedback loop that is hard to break

ACTION-7?

<trackbot> ACTION-7 -- Dan Connolly to work with Olivier and Tim to draft a position regarding extensibility of HTML and the role of the validator for consideration by the TAG -- due 2008-05-15 -- PENDINGREVIEW

<trackbot> http://www.w3.org/2001/tag/group/track/actions/7

<Stuart> Does anyone have an architecture for an extensible validator?

yes, they announced a "unicorn" design

action-7?

<trackbot> ACTION-7 -- Tim Berners-Lee to draft a position regarding extensibility of HTML and the role of the validator for consideration by the TAG -- due 2008-06-26 -- PENDINGREVIEW

<trackbot> http://www.w3.org/2001/tag/group/track/actions/7

action-7?

<trackbot> ACTION-7 -- Tim Berners-Lee to draft a position regarding extensibility of HTML and the role of the validator for consideration by the TAG -- due 2008-07-31 -- OPEN

<trackbot> http://www.w3.org/2001/tag/group/track/actions/7

XMLVersioning-41

Alternative language versioning formalism Jonathan Rees (Tuesday, 27 May)

http://lists.w3.org/Archives/Public/www-tag/2008May/0155.html

Def. L is extensible if there is an L' that extends it.

action-149?

<trackbot> ACTION-149 -- Henry S. Thompson to henry to help Jonathan with ACTION-148 and ACTION-158 -- due 2008-05-26 -- OPEN

<trackbot> http://www.w3.org/2001/tag/group/track/actions/149

action-158?

<trackbot> ACTION-158 -- Jonathan Rees to write up thoughts on versioning and share with the group -- due 2008-05-28 -- PENDINGREVIEW

<trackbot> http://www.w3.org/2001/tag/group/track/actions/158

close action-158

<trackbot> ACTION-158 Write up thoughts on versioning and share with the group closed

close action-149

<trackbot> ACTION-149 Henry to help Jonathan with ACTION-148 and ACTION-158 closed

action-148?

<trackbot> ACTION-148 -- Jonathan Rees to see if he can develop a formal basis for the definition of extensibility, possibly includiing definitions of forwards/backwards compatibility -- due 2008-05-26 -- CLOSED

<trackbot> http://www.w3.org/2001/tag/group/track/actions/148

action-158?

<trackbot> ACTION-158 -- Jonathan Rees to write up thoughts on versioning and share with the group -- due 2008-05-28 -- CLOSED

<trackbot> http://www.w3.org/2001/tag/group/track/actions/158

action-158?

<trackbot> ACTION-158 -- Jonathan Rees to write up thoughts on versioning and share with the group -- due 2008-05-28 -- PENDINGREVIEW

<trackbot> http://www.w3.org/2001/tag/group/track/actions/158

SKW: I hope to pick this up again when DO is around