W3C

- DRAFT -

Technical Architecture Group Teleconference

15 Jul 2010

Agenda

See also: IRC log

Attendees

Present
DKA, +1.413.458.aaaa, Jonathan_Rees, Ashok_Malhotra, Noah_Mendelsohn, Yves, johnk
Regrets
Chair
SV_MEETING_CHAIR
Scribe
johnk

Contents


trackbot-ng: start telcon

<trackbot> Date: 15 July 2010

<scribe> scribe: johnk

<scribe> ScribeNick: johnk

minutes approval

RESOLUTION: approve minutes of the 24th June

Administration

nm: no teleconference next week, discuss following weeks
... I'm at risk for the 29th July
... any requests, otherwise I'll leave as tentative
... ok, hearing no requests, will leave the next teleconference as tentatively for the 29th July
... any changes to the agenda?

(hears none)

siteData-36

<noah> http://lists.w3.org/Archives/Public/www-tag/2010Jul/0005.html

ACTION-451

ACTION-451?

<trackbot> ACTION-451 -- Jonathan Rees to attempt to close ISSUE-36 (dummy action so that this shows up in agenda planning) -- due 2010-07-09 -- PENDINGREVIEW

<trackbot> http://www.w3.org/2001/tag/group/track/actions/451

ISSUE-36 is http://www.w3.org/2001/tag/group/track/issues/36

nm: I assume that RFC says "this is a good way of doing this" (dealing with well-known URIs management)

jar: the RFC sets up a registry, and from reviewing the issue, I believe that the RFC and the registry resolves it
... the RFC should carry a lot of weight

PROPOSAL: close ISSUE-36

<noah> +1 to close

<jar> +1 close

<Yves> +1 to close

+1

<DKA> +1

should add notes to the issue, refing the RFC

RESOLUTION: we believe RFC5785 provides an appropriate means of using site metadata, and are thus closing ISSUE-36

Generic fragment id processing

ACTION-443?

<trackbot> ACTION-443 -- Jonathan Rees to chase down what specs say regarding looking up fragid in 2nd representation if not found in 1st representation -- due 2010-06-23 -- PENDINGREVIEW

<trackbot> http://www.w3.org/2001/tag/group/track/actions/443

nm: what happens when conneg interacts with frag processing?

<noah> close ACTION-451

<trackbot> ACTION-451 Attempt to close ISSUE-36 (dummy action so that this shows up in agenda planning) closed

jar: in theory you look at any or all of the available pieces of information, you figure out what the fragid refers to
... you could get the information from conneg or other sources
... so there is the possibility for confusion

yl: is there a relationship to connecting fragment ids?
... similar to issue in HTTP combining fragment ids

jar: I think it's a different issue

nm: in any particular interaction you get back only one representation
... issue is when you get multiple different representations
... agree the redirection case is different

<Yves> redirection allows frags on the redirected URI, and indeed it's a different issue

nm: what do we want to do with the issue related to your action (jar)?

<noah> close ACTION-443

<trackbot> ACTION-443 Chase down what specs say regarding looking up fragid in 2nd representation if not found in 1st representation closed

jar: I don't think there's anything particularly new, suggest to close the action

<Yves> see http://trac.tools.ietf.org/wg/httpbis/trac/ticket/43

nm: should look at Yves' issue too

yl: mime type definition is not well-defined in the case of fragment ids

nm: you do an interaction and get a redirect
... follow the redirect, and second interaction gives you a representation
... and you are suggesting that it is the second representation's mime type that rules

yl: yes

<jar> GET A#B -> 307 C#D, GET C#D -> 200 Content-type M

jar: I think that is common practice

<Yves> in fact you do GET A

<jar> what is meaning of A#B, given GET A -> 307 C#D, GET C#D -> 200 Content-type M

<jar> what is meaning of A#B, given GET A -> 307 C#D, GET C -> 200 Content-type M

nm: does everyone agree with the simple case?

<jar> #(A,'B') = #(#(C,'D'),'B')

<noah> Not what I first thought, but now that I think about it, what you're proposing makes a lot of sense. +1

jar: what resource is identified by C#D, and then determine what resource is identified within that context for B

<noah> (Granting that, in the end, we may have to just spec what browsers are doing, regardles of what seems cleanest architecturally).

<Yves> if in html (or xml) the id 'D' appears before 'B' and hence is not in the subtree, should the redirect locate it?

<jar> I would say "within that secondary resource" not "within that context"

<noah> Challenge: we have rules for resolving a fragid vs. a media typed doc, I.e., read the mime type registration; what's the rule for resolving against a secondary resource, which in general may not have mime type

<noah> So, in your example, which spec would I read to find out how to resolve a B against #(C,'D')

jar: I agree it could be solved by the mime type registration

<timbl> ? GET A#B -> 307 C#D ??

jar: if primary resource is an element, what does it mean to have a secondary resource that is also an element?

<timbl> Can you redirect to C#D ?

<timbl> I though that C#D was illegal in a redirect

<Yves> timbl, it is allowed

<jar> no it's not... not in rfc 2616

<jar> but yes in 2616bis i think

nm: you're proposing to solve this in MIME type reg, but with a particular architectural solution

<jar> (worth checking)

<timbl> It is done by purl

nm: "I will tell you how the second resource resolves within the context of the first"

jar: would like the idea of composition of these to be dealt with somehow

<noah> NM: But, implicitly, the media type registration for the primary resource is a good place to specify that two level resolution?

<timbl> So what does it mean to redirect to a fragment?

<Yves> agree that the composition needs to be defined, at least with a default

jar: you don't redirect to a fragment, just to a resource

<jar> you don't redirect to a fragment, you redirect to a (secondary) resource

<timbl> I can imagne it means "A#B and C#D identify the same thing, go fetch C to learn mode" but I don'ytk know where that is written

<jar> this is a use/mention issue

<noah> But, the point is, that the second frag resolution should be defined in terms of the first.

<timbl> "In terms of"?

<jar> C#D#B

<Yves> in fact A redirect to C#D, so A#B and C#D are not the same

<timbl> Aaaagh

jar: what would a URI with two '#' in it mean (conceptually)?

<jar> ok, then use my #(,) notation if ## is so scary

nm: it's repeated operations, not parsing of syntax
... having noted all this, should we do more?

<jar> how does this compositional idea compare to actual browser behavior? can a new media type reg resolve the dissonance?

YL: would suggest we track an issue

NM: carrying this under ISSUE-39 currently

ISSUE-57?

<trackbot> ISSUE-57 -- The use of HTTP Redirection -- open

<trackbot> http://www.w3.org/2001/tag/group/track/issues/57

ACTION on Noah to schedule discussion of redirecting to secondary resources

<trackbot> Sorry, couldn't find user - on

<noah> . ACTION: Noah to schedule, when Tim is available, discussion of redirection from A#B to C#D

<noah> ACTION: Noah to schedule, when Tim is available, discussion of redirection from A#B to C#D [recorded in http://www.w3.org/2010/07/15-tagmem-minutes.html#action01]

<trackbot> Created ACTION-452 - Schedule, when Tim is available, discussion of redirection from A#B to C#D [on Noah Mendelsohn - due 2010-07-22].

ACTION-449

<jar> (that's meaning of A#B given redirection from A to C#D)

ACTION-449?

<trackbot> ACTION-449 -- Noah Mendelsohn to schedule discussion of pushback on generic handling of fragment IDs in application/xxx+xml media types (self-assigned) -- due 2010-07-13 -- PENDINGREVIEW

<trackbot> http://www.w3.org/2001/tag/group/track/actions/449

<noah> http://lists.w3.org/Archives/Public/www-tag/2010Jun/0125.html

nm: drafted email saying the TAG thinks you should back off from generic processing of frag ids
... got pushback

<noah> ACTION-450?

<trackbot> ACTION-450 -- Yves Lafon to investigate generic processing of svg+xml and XHTML+xml -- due 2010-07-01 -- OPEN

<trackbot> http://www.w3.org/2001/tag/group/track/actions/450

YL: it was possible to create valid fragments that looked like XPointers, but which led to nothing
... I need to investigate further

NM: speaking as an individual, I would like to see the TAG change its advice

<jar> +1 we should probably change our advice

NM: indicating that generic process is part of RFC 3023
... and making clear that future specs. will support generic processing of frag ids

<Yves> and also media type wanting to define fragment must do so in order to avoid conflict with generic processing

JAR: we should think hard before changing advice

<DKA> +1 to thinking twice and getting Henry's and Tim's feedback on this.

<noah> ACTION-449?

<trackbot> ACTION-449 -- Noah Mendelsohn to schedule discussion of pushback on generic handling of fragment IDs in application/xxx+xml media types (self-assigned) -- due 2010-07-13 -- PENDINGREVIEW

<trackbot> http://www.w3.org/2001/tag/group/track/actions/449


. ACTION: Noah to schedule discussion of this when Tim, Larry and Henry are around and send an email to 3023 group describing our action

<noah> . ACTION: Noah to let 3023bis folks know that we are aiming to resolve generic processing concerns in Sept, after Larry returns.

<noah> ACTION: Noah to let 3023bis folks know that we are aiming to resolve generic processing concerns in Sept, after Larry returns. [recorded in http://www.w3.org/2010/07/15-tagmem-minutes.html#action02]

<trackbot> Created ACTION-453 - Let 3023bis folks know that we are aiming to resolve generic processing concerns in Sept, after Larry returns. [on Noah Mendelsohn - due 2010-07-22].

NM: will edit due date on 449 to match, and link to minutes

Web Developer Camp

DKA: sent email suggesting this idea to TAG
... combining the TAG Web App Architecture initiative and the prevalence of Web APP developers in CA seems like a good environment for exploring relevant issues

<noah> I suspect your workshops on other topics, like security, specifically drew world-class experts. "camp" doesn't quite suggest that same model to me.

DKA: could be an opportunity for us to learn
... could be an opportunity to create understanding about what the TAG/W3C is doing in this area
... would be useful to see it as a W3C event, rather than TAG
... but should also be seen as being related to the TAG work
... scope should focus on what I would call loosely "architectural issues"

<noah> Wasn't there some sort of developer-focused event at the Bay Area TPAC last Nov? Am I misremembering?

DKA: what are boundary problems, or issues faced by Web developers?

<noah> I'm thinking of: http://www.w3.org/2009/11/TPAC/DevMeeting

<noah> W3C invites the public to gather for an afternoon of discussion and networking. This Developer Gathering takes place during W3C's annual Technical Plenary (TPAC) Week, when W3C Working Groups meet face-to-face and work to resolve the most challenging technical issues facing the Consortium.

<noah> How did that work out?

DKA: we can work on a scope and then start on logistics

NM: how would it differ from or be the same as previous workshops?
... is this a bunch of experts to a roundtable so that we make the TAG more effective, that's one thing?
... or more grassroots-level event?

DKA: perhaps both?

NM: what is success?

<noah> I see, e.g. learning about points of confusion.

DKA: action points for us on Web architecture?

<noah> FWIW: I often give classes that focus on some of the subtle points in our TAG findings.

DKA: a formal workshop might not be appropriate

<noah> That does tend to tease out discussion of what people don't understand, and what they disagree with.

NM: I think we need a bit more structure in order to make this successful
... one way to do this would be to have a series of presentations, structured around TAG findings
... including about stuff we haven't yet resolved
... and then open the floor after each presentation...

DKA: I think that would be valuable... but would like to find a way to listen to what concerns are
... spend some of the time in listen mode, and some in talk mode telling people what has been done

<Yves> doing two-ways communication and learn form issues we might not realize people bump in would be good

DKA: merging unconference with scheduled presentations is difficult

JK: (discusses how we might do that)

NM: scheduling would be difficult

DKA: was thinking about either Sun/Mon

NM: we need to have good attendance from the TAG
... we need to iterate on the success criteria - why do this, who should come, etc.
... OK, so should we spend TAG time on this event?

<DKA> +1

<DKA> :)

<Yves> +1

+1

<noah> .5

<noah> Nervous

<jar> +1 but I think most of the work of organizing can/should be done offline. tag should support efforts of organizers

and I'm happy to actually ask at least one person whether he would be interested in such an event

volunteers to list potential topics and people as a start

<DKA> I'm happy to take lead on this with TAG support on topics and format, reaching out to W3C team folks to help with logistics...

<jar> oracle?...

<jar> DKA has my vote


. ACTION DKA to take lead in organizing possible Web application workshop

<DKA> . ACTION DKA to take lead in organizing possible Web apps architecture camp / workshop / openday

<DKA> ACTION DKA to take lead in organizing possible Web apps architecture camp / workshop / openday

<trackbot> Created ACTION-454 - Take lead in organizing possible Web apps architecture camp / workshop / openday [on Daniel Appelquist - due 2010-07-22].

privacy workshop

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

<DKA> action noah to schedule discussion on privacy workshop outcomes.

<trackbot> Created ACTION-455 - Schedule discussion on privacy workshop outcomes. [on Noah Mendelsohn - due 2010-07-22].

pending review items

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> ACTION-363?

<trackbot> ACTION-363 -- Jonathan Rees to inform SemWeb CG about market developments around webfinger and metadata access, and investigate relationship to RDFa and linked data -- due 2010-06-23 -- PENDINGREVIEW

<trackbot> http://www.w3.org/2001/tag/group/track/actions/363

<noah> close ACTION-363

<trackbot> ACTION-363 Inform SemWeb CG about market developments around webfinger and metadata access, and investigate relationship to RDFa and linked data closed

ACTION-435?

<trackbot> ACTION-435 -- Jonathan Rees to consult Tyler Close regarding UMP-informed web storage vulnerability analysis -- due 2010-06-22 -- PENDINGREVIEW

<trackbot> http://www.w3.org/2001/tag/group/track/actions/435

close ACTION-435

<trackbot> ACTION-435 Consult Tyler Close regarding UMP-informed web storage vulnerability analysis closed

NM: any other business?
... hearing none, ADJOURN

rrsgent, generate minutes

Summary of Action Items

[NEW] ACTION: Noah to let 3023bis folks know that we are aiming to resolve generic processing concerns in Sept, after Larry returns. [recorded in http://www.w3.org/2010/07/15-tagmem-minutes.html#action02]
[NEW] ACTION: Noah to schedule, when Tim is available, discussion of redirection from A#B to C#D [recorded in http://www.w3.org/2010/07/15-tagmem-minutes.html#action01]
 
[End of minutes]

Minutes formatted by David Booth's scribe.perl version 1.135 (CVS log)
$Date: 2010/07/15 18:27:09 $

Scribe.perl diagnostic output

[Delete this section before finalizing the minutes.]
This is scribe.perl Revision: 1.135  of Date: 2009/03/02 03:52:20  
Check for newer version at http://dev.w3.org/cvsweb/~checkout~/2002/scribe/

Guessing input format: RRSAgent_Text_Format (score 1.00)

Succeeded: s/Yve's/Yves'/
Succeeded: s/appaers/appears/
Succeeded: s/confict/conflict/
Succeeded: s/the work/the work of organizing/
Found Scribe: johnk
Inferring ScribeNick: johnk
Found ScribeNick: johnk
Default Present: DKA, +1.413.458.aaaa, Jonathan_Rees, Ashok_Malhotra, Noah_Mendelsohn, Yves, johnk
Present: DKA +1.413.458.aaaa Jonathan_Rees Ashok_Malhotra Noah_Mendelsohn Yves johnk
Agenda: http://www.w3.org/2001/tag/2010/07/15-agenda

WARNING: No meeting chair found!
You should specify the meeting chair like this:
<dbooth> Chair: dbooth

Found Date: 15 Jul 2010
Guessing minutes URL: http://www.w3.org/2010/07/15-tagmem-minutes.html
People with action items: noah

WARNING: Input appears to use implicit continuation lines.
You may need the "-implicitContinuations" option.


[End of scribe.perl diagnostic output]