Technical Architecture Group Teleconference -- 16 Sep 2010

<trackbot> Date: 16 September 2010

<noah> scribenick: masinter

<noah> meeting: TAG Teleconference of 16 Sept 2010

<noah> agenda: http://www.w3.org/2001/tag/2010/09/16-agenda

Larry talking about getting brief "MIME and the web" on today's agenda just to set direcdtions

agenda

scheduling meetings

ashok won't be on some telcons (missed which ones)

noah wants people to get some writing done

approval of minutes

<Ashok> I will miss the next 3 telcons ... next telcon I can attend is Oct 14

<noah> Minutes of Sept 2.: http://www.w3.org/2001/tag/2010/09/02-minutes

<johnk> +1 also

resolution: minutes of Sept 2 approved

administrative items

noah: F2F coming up

<ht_home> I am coming

<jar> I'm planning on it

Larry, Ashok, Jonathan will be there

<johnk> I can't say yet, but *should* be there

noah: Contact from Mark Nottingham (IETF liaison to W3C), suggesting that Alexey M. IETF application area director is planning on being at TPAC, suggested possible meeting with TAG
... suggest responding to MNot suggesting meeting at TPAC on Thursday

<noah> YL: I can take the action

<noah> ACTION: Noah to respond positively to Mark Nottingham regarding meeting with Alexey at TPAC [recorded in http://www.w3.org/2010/09/16-tagmem-minutes.html#action01]

<trackbot> Created ACTION-463 - Respond positively to Mark Nottingham regarding meeting with Alexey at TPAC [on Noah Mendelsohn - due 2010-09-23].

<ht_home> I will unfortunately now not going to be in Lyon at all

<noah> HT: I will, unfortunately, not be in Lyon

<jar> I won't be in Lyon either

<noah> ACTION: Yves to coordinate agenda for TAG/IETF meeting at TPAC [recorded in http://www.w3.org/2010/09/16-tagmem-minutes.html#action02]

<trackbot> Created ACTION-464 - Coordinate agenda for TAG/IETF meeting at TPAC [on Yves Lafon - due 2010-09-23].

<noah> ACTION-464 due: 2010-10-23

<noah> ACTION-464: due 2010-10-23

<trackbot> ACTION-464 Coordinate agenda for TAG/IETF meeting at TPAC notes added

<jar> ACTION-464 due 2010-10-23

<trackbot> ACTION-464 Coordinate agenda for TAG/IETF meeting at TPAC due date now 2010-10-23

<noah> scribenick: Ashok

Noah: We need to do some writing. Will talk to people one-on-one

Metadata Architecture

<noah> http://lists.w3.org/Archives/Public/www-tag/2010Sep/0016.html

Noah: Jonathan would like to pass this on to someone else

jar: I cannot think about anything non-trivial to say ...

<jar> masinter: survey what w3c is doing?

Larry: Can we survey what W3C does about metadata and problems thereon

Noah: Should we follow the links from the action

jar: I cannot do this ... would be better if we got some fresh eyes looking at it
... or postpone to January

Noah: We could close it but wait till Tim is around
... or someone follow the links

<jar> agenda: http://www.w3.org/2001/tag/2010/09/16-agenda

<noah> Take a look at email: http://lists.w3.org/Archives/Public/www-tag/2009Nov/0028.html

<noah> Tracker, this discussion is about ACTION-282

Ashok: That's about metadata access ... this issue is about Metadata architecture

<noah> I think Larry is proposing: "Draft a finding on which metadata technologies (RDFa, photo EXIF, etc.) do I use when?"

<noah> Maybe we should retitle the action?

Larry: People don't know how to write metadata ... lots of confusion

<jar> masinter: A success criterion for this activity. RDFa, microdata, XIF, XMP, GPS, privacy policy. People are confused about what to use when. It's a confusing mess. What happens when they conflict? When I want to change it? Operational issues among W3C and other specs.

<noah> NM: I think we also discovered that the advice in http://lists.w3.org/Archives/Public/www-tag/2009Nov/0028.html is misleading; ACTION-282 is mostly NOT about metadata access.

<noah> LM: I'd like an architecture which is a framework into which these specific mechanisms can be placed.

Larry: We need a framework into which these mechanisms can fit

jar: That matches my memory of how this came up

<noah> jar: F2F discussion would be good...we're ready

jar: Perhaps duscuss at f2f

Larry: Perhaps create an outline

noah: I can put on an agenda

Larry: May be good to have a reading list
... I will send mail

<noah> . ACTION: Noah to schedule discussion of ACTION-282, which metadata mechanisms to use when. Get reading list from Larry and www-tag.

<noah> ACTION: Noah to schedule discussion of ACTION-282, "which metadata mechanisms to use when". Get reading list from Larry and www-tag. [recorded in http://www.w3.org/2010/09/16-tagmem-minutes.html#action03]

<trackbot> Created ACTION-465 - Schedule discussion of ACTION-282, "which metadata mechanisms to use when". Get reading list from Larry and www-tag. [on Noah Mendelsohn - due 2010-09-23].

Generic fragment id processing

Noah: There is inconsistency in 3023bis and frag-id processing for RDF+XML
... disagreement about what should be done
... TAG recommended that 3023bis back off but we got strong pushback saying that generic processing was desirable

<noah> Larry's email: http://lists.w3.org/Archives/Public/www-tag/2010Aug/0034.html

Larry: Is there usecase for putting Xpointer in the frag-id?

Noah: Recaps the generic processing argument

Yves: Issue is not have XPointer supported but all media types must have a syntax compatible with XPointer

<noah> Martin Duerst's Note: http://lists.w3.org/Archives/Public/www-tag/2010Jul/0022.html

<noah> Roy Fielding's note: http://lists.w3.org/Archives/Public/www-tag/2010Jul/0021.html

<noah> Norm Walsh's note: http://lists.w3.org/Archives/Public/www-tag/2010Jul/0020.html

<noah> Norm writes:

<noah> >> I find that unsatisfactory. It leaves generic XML processors out in

<noah> >> the cold once again by expecting them to be aware of all of the media

<noah> >> type registrations for all +xml formats.

Henry: By using +xml they buy into XML and thus folks expect XML processing

<Zakim> noah, you wanted to go back to talking about conneg

Noah: So, people can read different documents and process the content differently

Larry: You want frag-id processing be generic

<jar> ht, noah: the URIref 'identifies' one thing when the processor has read the rdf+xml spec, a different thing when it reads 3023bis. maybe that's ok?

<jar> ... a single document, same media type in both cases

<jar> masinter: I want to push back on Norm's and Roy's use cases

Larry: I want to push back on their usecases ... don't understand why frag-id processing has to be generic

<noah> HT: XSL stylesheets that format rdf+xml

<jar> I don't buy that one. It would be meaningless (almost always) in the RDF interpretation

<ht_home> Note that XInclude is _not_ an example, because it does not allow the use of FragIDs

Larry: I will ask Norm or Roy for concrete example

<noah> ACTION: Larry to ask Norm, Roy and Martin for concrete use cases where generic processing of fragment ids is important [recorded in http://www.w3.org/2010/09/16-tagmem-minutes.html#action04]

<trackbot> Created ACTION-466 - Ask Norm, Roy and Martin for concrete use cases where generic processing of fragment ids is important [on Larry Masinter - due 2010-09-23].

<ht_home> jar, the example I have in mind preserves the RDF semantics but restructures the RDF XML

<jar> grumph. need details.

AWWW and the Web interaction model

<noah> 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 2010-06-07 -- PENDINGREVIEW

<trackbot> http://www.w3.org/2001/tag/group/track/actions/355

<noah> John did some writing: http://lists.w3.org/Archives/Public/www-tag/2010Jun/0034.html

Discussion of John's document

<noah> AWWW section on

<noah> AWWW section on "Interaction": http://www.w3.org/TR/webarch/#interaction

Noah: AWWW slices into 3 themes, identification, interaction, data formats

John: How does client know it can use a particular URI ... client/server agreement
... update of state w/o user interaction e.g. web sockets
... need to define 'user agent' ... may be more than a browser
... acts on behalf of server also

Larry: Web apps break our model of knowledge ... downloaded code knows stuff client does not know
... client runs downloaded code on behalf of server

<noah> Larry: what's interesting is the client is executing code not (just) on behalf of the client, but rather downloading code that acts on behalf of the server

<noah> I think that's a really useful way to put it.

John: Clients act like servers and servers like clients

Larry: Client has to know security boundaries not the semantics

Noah: John, can you write a document ... maybe a sketch

John: Yes

<noah> LM: Be careful about terminology about "client", "server" and "knowing"

Larry: Be careful about terminology -- client, server and knowing
... only need URIs for states you need to come back to

<noah> What I asked John to do was an early draft of either an updated version of http://www.w3.org/TR/webarch/#interaction, or else perhaps a finding that would be the start of a complementary section in some TAG publication

<noah> AM: Larry, you commented on my writeup saying that, and I agree with you. Raman came back and pushed back, but I've forgotten what his point was.

<noah> AM: Do you know?

<noah> LM: No.

MIME and the Web

Larry: I wrote a blog post on use of MIME and how it could be aligned better
... suggested some improvements to MIME registry

<noah> LM: Thinking of splitting into two documents.

Larry: may be easier to get agreement on status and history than the improvements

<noah> We are discussing: http://lists.w3.org/Archives/Public/www-tag/2010Sep/0027.html

<noah> LM: Also might relate to versioning issues

Larry: also about sniffing
... probly separate versions from variants
... finally about distrbuted extensibility
... I want to bring learning from other MIME types to HTML
... Pehaps descriptive expositions of what has happenned

Noah: Can you do some work on this

Larry: I will have another document in a week

<noah> ACTION: Larry to revise Mime & Web blog entry, perhaps leading to Internet Draft [recorded in http://www.w3.org/2010/09/16-tagmem-minutes.html#action05]

<trackbot> Created ACTION-467 - Revise Mime & Web blog entry, perhaps leading to Internet Draft [on Larry Masinter - due 2010-09-23].

<noah> http://www.w3.org/2001/tag/group/track/actions/pendingreview

Pending Review Actions

http://www.w3.org/2001/tag/group/track/actions/pendingreview

<noah> NM: I've closed ACTION-322

<noah> ACTION-322?

<trackbot> ACTION-322 -- Jonathan Rees to brief Science Commons counsel re forbidding hyperlinking -- due 2010-08-20 -- CLOSED

<trackbot> http://www.w3.org/2001/tag/group/track/actions/322

<noah> ACTION: Noah to invite Thinh to telcon where Tim will be available to discuss "forbidding of hyperlinking" [recorded in http://www.w3.org/2010/09/16-tagmem-minutes.html#action06]

<trackbot> Created ACTION-468 - Invite Thinh to telcon where Tim will be available to discuss "forbidding of hyperlinking" [on Noah Mendelsohn - due 2010-09-23].

<noah> ACTION-468?

<trackbot> ACTION-468 -- Noah Mendelsohn to invite Thinh to telcon where Tim will be available to discuss "forbidding of hyperlinking" -- due 2010-09-23 -- OPEN

<trackbot> http://www.w3.org/2001/tag/group/track/actions/468

<noah> 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 2010-10-05 -- OPEN

<trackbot> http://www.w3.org/2001/tag/group/track/actions/355

<noah> http://www.w3.org/2001/tag/group/track/actions/overdue

Adjourned

- DRAFT -

Technical Architecture Group Teleconference

16 Sep 2010

Attendees

Contents