W3C

Technical Architecture Group Face-to-face, day 3

09 Oct 2012

See also: IRC log [incomplete]

Attendees

Present
Dan Appelquist (in part), Robin Berjon (in part, by 'phone), Tim Berners-Lee, Yves Lafon, Peter Linss, Ashok Malhotra, Larry Masinter, Noah Mendelsohn, Jonathan Rees (in part, by 'phone), Jeni Tennison, Henry S. Thompson
Regrets
Jonathan Rees
Chair
Noah Mendelsohn
Scribe
Yves Lafon, Larry Masinter, Jeni Tennison, Henry S. Thompson

Contents


Approval of minutes

<noah> http://www.w3.org/2001/tag/2012/09/20-minutes

RESOLUTION: minutes of 2012-09-20 approved

<noah> http://www.w3.org/2001/tag/2012/09/27-minutes

RESOLUTION: minutes of 2012-09-27 approved

<noah> http://www.w3.org/2001/tag/products/index-2012-09-26.html

TAG priorities

<noah> http://www.w3.org/2001/tag/products/index-2012-09-26.html

Noah: for Fragment identifier document, do we need CR?

Yves: yes CR is mandatory as you must prove interoperability (technically, CR can be skipped if there is evidence)

Larry: the specification contains lots of recommendations for media type registration for example
... happy to change the REC date to be the CR date (then wait for "implementation" evidence)

Ashok: we need to spell out the exit criteria

Larry: as it can be implemented, we should go through CR

Yves: we can already have implementation, so we need first to outline the exit criteria

Noah: we need to have an action to create exit criteria

<JeniT> ACTION: Jeni to with Larry work out what the exit criteria from CR for fragids best practices should be [recorded in http://www.w3.org/2001/tag/2012/10/09-minutes.html#action01]

Larry: we need to put attention to steps that engage the community, so it makes sense to push our schedule to accomodate communicating with the community

Publishing and Linking

Noah: someone working on starting a community group?

Larry: wondering if it is the right step

Ashok: what are the exit criteria here?

Yves: is REC the right track? shouldn't it be a TAG finding?

Larry: when we put the schedule together, we had best practise in the document. We moved incrementally best practises out of the document

Noah: we promised a REC, we may have now more insight and decide that it was not the right format

Larry: we will do another editorial pass on the document. Thinking of the exit criteria after the next iteration is a good idea

<noah> ACTION: Larry with help from Ashok (and TAG), 1) decide exit criteria on Publishing & Linking 2) Rec track vs. Finding 3) Update product page to match - Due 2012-10-16 [recorded in http://www.w3.org/2001/tag/2012/10/09-minutes.html#action03]

URI Documentation Discovery

product page should be updated before TPAC

Noah: there are also lower priority items that might (or not) become higher priority topic

Ashok: Web application storage should be transformed into offline applications, we will then decide what to do with it

JeniT: we might look at distributed web applications, with data exchanged between different services (at different locations)

about registered protocol handlers, web intents, and accessing your own data from an online service

(can be tied to open data)

Ashok: will it involve synchronization?

local data to global data

Tim: it's an independant problem

Noah: is the TAG interested in this topic, and should Jeni work on a more detailed proposal?
... yes

Tim: another interesting thing is to look at a specific technology or topic and find out what are the missing pieces

Larry: how do we update the architecture document to match what people need

<JeniT> ACTION: Jeni to draft rough product page / briefing pape for "distributed web applications" [recorded in http://www.w3.org/2001/tag/2012/10/09-minutes.html#action04]

Noah: it is good information to have a timeline and the kind of document we want to produce

Jeni: do we need to change the arch doc wrt issue-57?

Noah: the goal by itself is not updating webarch

<masinter> perhaps we need to add, before we close an item, to review webarch & web presence

Larry: before we close an item, we should add another step: go back and update webarch if needed

keep this an alive document

Noah: the goal of a topic is not to publish a document but to help the community, we failed to get input from the community one year after to see if it helped or not

ht: I agree with Tim, AWWW is an historical document.

NM: Maybe I should come up with a plan for going over our recently closed work to check on whether success criteria are being met

ht: the world moved on, with new complexities

<noah> ACTION: Noah to think about how to evaluate results vs. success criteria on closed work - Due 2013-01-01 [recorded in http://www.w3.org/2001/tag/2012/10/09-minutes.html#action05]

<Zakim> masinter, you wanted to add 'persistence' and 'security infrastructure & architecture' as potential projects

Larry: persistence of identifier is too narrow, persistence of document is important

LM: Not sure I have time to lead this

Larry: we need also work on security

Ashok: who should work on that?

Noah: who wants to work with the security community to figure out where the TAG could be helpful?

<noah> ACTION: Ashok working with experts in security community, to suggest projects TAG might undertake relating to security - Due: 2012-11-20 [recorded in http://www.w3.org/2001/tag/2012/10/09-minutes.html#action07]

Noah: we should go through our issue list and check if there are issues we could make progress as work items

larry: dwim?

Noah: it would be good also to strongly promote the "be conservative in what you send" part of Postel's law

<masinter> http://www.w3.org/wiki/Evolution

<masinter> A third rail is a method of providing electric power to a railway train, through a semi-continuous rigid conductor placed alongside or between the rails of a railway track. It is used typically in a mass transit or rapid transit system, which has alignments in its own corridors, fully or almost fully segregated from the outside environment. In most cases, third rail systems supply direct current electricity.

Larry: error handling is part of the theory about managing evolution. should be on the list of candidate topics

<masinter> there is a debate in the W3C community about how to write specifications, which is whether you specify a narrow normative language and leave it implementation-dependent how to handle it

<masinter> http://masinter.blogspot.co.uk/2011/06/irreconcilable-differences.html

<noah> ACTION: Larry to frame for telcon discussion possible TAG work relating to DWIM and Issue errorHandling-20 - Due 2013-11-13 [recorded in http://www.w3.org/2001/tag/2012/10/09-minutes.html#action08]

<masinter> on long-term permanence, see http://larry.masinter.net/0603-archiving.pdf

TAG effectiveness

<noah> ACTION-725?

<trackbot> ACTION-725 -- Jeni Tennison to with help from Peter, to create big picture overview coming out of analysis of TAG effectiveness -- due 2012-06-21 -- OPEN

<trackbot> http://www.w3.org/2001/tag/group/track/actions/725

<noah> Objectives matrix: http://www.w3.org/2001/tag/doc/objectives-matrix-20120930

<noah> From the intro to the matrix: "http://www.w3.org/2001/tag/doc/objectives-matrix-20120930"

<noah> From the intro to the matrix: " If it were 1988 and one had to design the Web, what would the design objectives be?"

Jeni: it is important to look at the big picture, and assess how effective we are in improving the web. not had time to flesh out a proposal

Jeni: it is an important thing to do, we might also come back to it after the election

LM: Remind me where we stood with the June session

NM: Jeni and Peter took action to distill. Still good to do, no progress yet.

Noah: there might be discussion with the AB at TPAC, ready to facilitate such a meeting

<noah> ACTION: Noah to send Steve Zilles note saying Henry, Larry, Peter, Ashok, Tim and Jeni (Wed) could meet with AB [recorded in http://www.w3.org/2001/tag/2012/10/09-minutes.html#action09]

<noah> 4 min timeout to review: http://www.w3.org/2001/tag/doc/objectives-matrix-20120930

discussion

timbl: this is an incredible amount of work that jar has done

tbl: it's nice also that it ties these to our issues
... the objectives were taken by scraping them out of the document
... did these grow into properties of the web, do they group into high-level goals or properties of the web?
... some of the ones about extensibility can be grouped together

noah: this is a good piece of work, but it doesn't talk to the areas where we are having trouble.
... what we fail to do is bring the community together
... i don't think fleshing out this matrix is going to convince the community
... we're not respected, how do we fix that
... this is potentially useful, but even in areas where we know stuff and articulate it well
...we aren't getting the traction we should.

jeni: i think it's a two-way process. This matrix is about, one of the things we talked about was expressing "Why Architecture is Important", these could be grouped together

Jeni: the matrix can be used to define what are the main goals for the web
... i.e. explain why the architecture is important

larry: to be effective starts with picking the right item to work on, do a good work, then promote it. in this multi-stage pipeline, the matrix is about one part of it.

<masinter>

timbl: the tag tries to define good architectural principles, architecture of the web, and also work on crisis.

noah: we published a finding a while ago on webapps state. I have nor heard back from the community about it and I don't see how the matrix is helping us here

<jar> ummm... the matrix helps because it says we ought to know the objective(s)... ?

<jar> ... or because it provides a list of candidate objectives? ...

larry: what would be the exit criteria for application state? (if it was on REC track) we might review our finding in the light of "what would it take to go to REC", or "should it be obsoleted"

<masinter> "implementation" of findings might be "W3C recs that recommend or reference (or could reference)" ?

timbl: some document might need a push (like going to WGs to talk about it)

<masinter> http://www.w3.org/2001/tag/findings

Noah: the 'success criteria' should not be 'publish a document' (and we tried to stay away from that)

<masinter> benchmark TAG against IAB?

<noah> ACTION-757?

<trackbot> ACTION-757 -- Noah Mendelsohn to think about how to evaluate results vs. success criteria on closed work -- due 2013-01-01 -- OPEN

<trackbot> http://www.w3.org/2001/tag/group/track/actions/757

<Zakim> noah, you wanted to talk about communcation vs. recs

JeniT: on 'managing the crisis', we need to be more reactive where people are actually arguing rather that be only in the mode "this looks interesting, I want to work on this"

<Zakim> JeniT, you wanted to talk about outreach

Ashok: the application state finding is targeted as devs, people we usually don't speak directly to. It would be good to better engage the dev community

<masinter> why don't WG chairs ask TAG for advice on disputes?

<Zakim> noah, you wanted to ask whether we should work with AB on process changes

noah: do we have ideas on changing charter or the process?

<jar> (re "targeted at devs": there is little that individual developers can do or not do to "lead the web to its full potential". rarely they can "innovate". being a "good citizen" doesn't scale. usually potential is advanced and preserved through consensus, and that means WGs)

[DKA joins]

peter: the css wg brought issues in front of the tag, would be happy to have TAG answers being more formal

<noah> LUNCH BREAK

3023bis & fragids

ht: if the +suffix definition for a fragid doesn't resolve to a fragid, then a specific media type can take over to provide a definition
... question is whether this (should) only apply to barenames, or to all fragid syntaxes
... eg in +xml, to XPointer syntax fragids
... if it did apply to all XPointer syntaxes, what would that mean?
... XPointer allows multi-part parts with failover
... if you support the scheme in the first part, and it defines a subresource, then that one, otherwise fall on to the next part
... syntax of XPointer includes barenames
... looks like three failure modes:
... 1. fragid that doesn't match XPointer syntax productions
... that's declarative based on syntax
... 2. 'doesn't map'
... 2a. unbound scheme prefix
... ie no namespace declaration for the prefix used in the scheme
... 2b. unsupported scheme
... ie unsupported by particular implementation
... 3. doesn't resolve to a fragment -- scheme-specific, failover to next part of the XPointer
... XPointer only talks about synactic failure and failure to resolve
... options for 3023bis: liberal and conservative
... liberal: non-XPointer-syntax fragids or ones that don't resolve, aren't defined by +xml spec
... conservative: non-XPointer-syntax fragids or barenames that don't resolve, aren't defined by +xml spec
... non-identifying fragids that match the non-barename XPointer syntax are reserved
... consider application/schema+xml
... want to identify schema components via the elements that define them within a schema
... would appear to make sense to use XPointer syntax
... with semantics that identify schema components rather than elements
... but you could only do that if normal XPointer rules don't give you an answer

noah: doesn't XPointer always resolve to elements, by definition?

ht: yes, hence application/schema+xml should not use XPointer syntax to point to components

<masinter> IETF process is "rough consensus and running code" -- and this is (should be) standards track, so 3023bis needs to follow IETF rules

JeniT: but if you have a vocabulary that wants to use fragids to point to elements, then...

ht: ...then you define a new XPointer scheme

JeniT: you should include that in 3023bis

ht: yes

masinter: remember to check what implementations are currently doing

ht: yes
... I don't know any implementations aside from norm's and mine and Mozilla's that support schemes other than element()

timbl: where are the implementations?

ht: I don't know of any XInclude implementations (which are the main implementations of XPointer) that work on the client side
... possibly TEI

timbl: no one uses them with XHTML?

ht: no, people only use barenames with XHTML

<masinter> look at http://www.w3.org/2008/WebVideo/Fragments/WD-media-fragments-impl/

masinter: look at the media fragments URI implementations

<masinter> can you look at http://www.w3.org/2008/WebVideo/Fragments/WD-media-fragments-impl/#mf-ua-tc test cases too

ht: ok, good idea

Offline Web Applications

http://www.w3.org/2001/tag/2012/10/07-agenda#offline

noah: we have no actions on this

<noah> ACTION-717?

<trackbot> ACTION-717 -- Ashok Malhotra to propose outline for possible TAG document (finding or rec) on architectural issues relating to storage sync, linked data, etc. Due 2012-07-15 -- due 2012-10-10 -- OPEN

<trackbot> http://www.w3.org/2001/tag/group/track/actions/717

Ashok: they're on web storage

<noah> http://lists.w3.org/Archives/Public/www-tag/2012Sep/0009.html

noah: we need to decide whether we will proceed on something in this area
... we might define objectives in terms of the objective matrix

DKA: I want to highlight the work going on with FirefoxOS

noah: we'll let Ashok frame first

Ashok: we've been looking at web storage and appcache, and what these two technologies are helping us with
... and the target is offline apps
... Yves and darobin convinced me that AppCache is not quite right for offline apps

<Yves> ie: it's not to be used as 'local storage'

Ashok: so we don't have an offline app architecture
... and it's something that the web should be supporting
... so I started looking at what it would take
... there's very little user control on the offline app and the global data and how they work
... so I thought we should try to spell out an architecture for offline apps
... you could write a paper, but it would be better to be done in a WG, with a spec
... whether we can help make that happen is the question

noah: what's the state of other W3C work on offline apps?

Ashok: the only thing I know about is AppCache
... DKA tells me there are efforts to make it better

noah: and there's offline storage, which is also relevant
... web storage

timbl: you say there isn't enough to do an app? what about the financial times app?

DKA: I was going to highlight that as an example of the state-of-the-art

AM: Let's talk about those examples
... Local storage is a very simple mechanism
... prop/value pairs, no transactions, etc.
... The Web Database idea was meant to be more like SQL, ended up with what's now called SQLite
... One thing I've thought about doing was moving that to be the full SQL thing

DKA: Have people seen the FT WebApp -- it's a very good demonstration of what you can do in iOS (Android is coming)
... It brings up what amounts to a browser, but with a chrome-less view
... Once initialized, you can go offline, and it's still usable -- access to all the articles, with their text and images

NM: How is it working, with proxy cache or some kind of storage?

DKA: Local storage, v. simple

NM: Not proxy cache, not SQLite?

DKA: Neither -- it does use AppCache, which requires a prompt to enlarge the cache, which they prewarn users about

HST: AppCache?

DKA: Assets needed for an app

JT: Manifest gives a list of assets, browsers then downloading all those assets into a special cache (not their normal web cache)

NM: Could I send a link by copying to the clipboard?

LM: I posted an article for Adobe about it:

<masinter> http://www.adobe.com/inspire/2012/02/html5-next-disruptive-technology.html

DKA: Not easy to tell the answers to NM's question, given the UI

NM: I think it matters whether there's a real, reusable, URI there

DKA: We don't know

HST: AppCache vs. local storage?

DKA: AppCache is for the actual articles and pictures; local storage is for the app itself: javascript, icons, boilerplate

NM: Different from GMail 5 years ago? Updates?

DKA: This updates a lot

NM: I mean from the user up

DKA: that happens too

NM: Synchronisation sophisticated?

DKA: Don't know -- not really my point, I wanted to just show what SOA is, and underline that it isn't easy: FT had to buy the company that got AppCache to work well enough to make this viable

<masinter> http://blogs.adobe.com/digitalpublishing/2010/11/martha-stewart.html

<DKA> https://developer.mozilla.org/en-US/docs/Mozilla/Firefox_OS/Architecture

DKA: Firefox OS is for mobiles: a stripped down linux running the Gecko engine, everything the user sees is at the Gecko level

NM: File system?

DKA: It's there, but it's not user visible as such

<DKA> https://developer.mozilla.org/en-US/docs/Apps/Manifest

<masinter> http://www.w3.org/2012/05/sysapps-wg-charter.html

<masinter> http://www.w3.org/2012/05/mobile-web-app-state/

<Ashok> Robin, who is fixing Appcache? Is there a sketch of the fixes?

<darobin> http://www.w3.org/community/fixing-appcache/

<darobin> sketch of fixes: https://etherpad.mozilla.org/appcache-london and https://etherpad.mozilla.org/appcache

<Ashok> Thanks! Are you hopeful that this will make AppCache good for offline apps?

<darobin> Ashok, yes, I'm hopeful that we can make AppCache useful ... it's dearly needed. My plan is to bang heads together at TPAC and come out with a plan to write up the fix

<Ashok> Robin, Can I join you in the headbanging excercise?

<darobin> Ashok, of course :)

<darobin> Ashok, be warned though that stories of people using AppCache in real life are actually pretty damn scary ... it's not for the faint of heart ;)

<Ashok> Right! I've heard some of the stories :-)

DKA: It supports permission requests, application orientation, regular/privileged/certified wrt access to to APIs

JT: How do you get privilege?

DKA: Signing and an installation sequence which involves the user, or could be preinstalled
... Certified is restricted to Mozilla and mozilla partners

NM: So no misc. downloaded app could dial?

DKA: Right, but could request access to less crucial APIs

NM: Skype today gives me an installable plugin, which looks like it can dial out

NM: Not sure the user recognises the difference

DKA: The work Mozilla has done is in my view a descendant of the work we did on widgets -- Mozilla might not agree
... but it's definitely in the same spirit

DKA: None of this is a work item in any WG

AM: It goes further than widgets did wrt the execution/security model

<DKA> Sysapps: http://www.w3.org/2012/sysapps/

<darobin> SysApps is where the hosted app vs packaged app decision will happen ... this is important, I'd watch

<darobin> Hosted apps sit on the Web, they just have a manifest; packaged apps are wrapped in a container and downloaded (like widgets)

DKA: To clarify: the execution/security model has been submitted to the SysApps WG, but the packaging stuff has not been, at least not yet -- there is some disagreement about where it should go

LM: I thought the primary difference between the WebApps and SysApps remits were the security model
... SysApps has to go outside the sandbox, so WebApps said that was outside their scope

DKA: In Santa Clara there was a fundamental disagreement on whether a package format was needed
... Mozilla were opposed, but have since decided it is needed
... My understanding was whether packaging should end up in WebApps or SysApps

RB: SysApps isn't chartered to do anything about packaging
... But WebApps had something, which could be updated

<DKA> Jidgets!

AM: Coming back to next steps -- it looks like some agreement to 'fix' AppCache issues may emerge, so I think we need to wait until TPAC until we see how that goes ahead

NM: So do we need to decide at some point to engage with "Offline Apps" as a work item, or just keep an eye on it?
... I'm just worried we need to commit or move on
... I can't tell if this will deliver value or not

AM: Robin, what do you think?

RB: Two reasons to wait to decide: 1) AppCache changes -- will they be patches or a full-scale rewrite? 2) SysApps WG only just been launched, not yet really going on the security model and offline issue

NM: Yeah, but it's overlate to advance architectural guidance once those basic decisions have been made
... E.g the role of URIs

AM: TPAC is less than 4 weeks away

NM: OK, but I want to make a firm decision at our next f2f meeting

<noah> ACTION-717?

<trackbot> ACTION-717 -- Ashok Malhotra to propose outline for possible TAG document (finding or rec) on architectural issues relating to storage sync, linked data, etc. Due 2012-07-15 -- due 2012-10-10 -- OPEN

<trackbot> http://www.w3.org/2001/tag/group/track/actions/717

<jar> Are there any namespace or registry issues associated with storage?

<Ashok> JAR, to your question -- I don't think so

<jar> what about spec collision issues?

<noah> which specs colliding with what please?

<Zakim> timbl, you wanted to say that there seems to be an issue for authenticating, for access to some data, both the person and the app they are using. Currently we have systems which

TBL: Security model for phones is about privileges you give to an app
... Security model for computer is about privileges you give to a user

TBL: Are we seeing any convergence -- only a given set of people can get certain functionality from an app

DKA: Doesn't latest MacOS have something like this, as a result of influence from iOS? I can install an app on a multi-user system which may not have access to all files?

<masinter> i don't mind having "track" items that we revisit quarterly to see if we're ready to take on major work, and perhaps we should make that a TAG item

<noah> Larry, that's fine, as long as we agree that's what we're doing

<noah> These have always been framed as: let's see if we can spin up a major effort in storage/offline. We keep saying it but not doing it. That's not healthy IMO.

<noah> If we want to agree to revisit regularly without plan for major deliverable, that's OK, and would be managed accordingly.

HST: Windows has had the me/everyone question at app install for a while. . .

<darobin> on MacOS, the Unix permissions system always applies ... you can install for all users but if it's run as you, it will only have access to what you have access to

<masinter> we could have a quarterly deliverable? blog post, "what's happening with 'offline apps' " ?

<noah> Larry, sure, as long as we say that's what we're doing

<noah> The claims have been we're looking at something like a finding, and then we just get this far, rinse and repeat for > 1 year. You were the one who, in Lyons said: if we revisit something >3 times, we should have a goal.

<jar> I see the TAG's business being mainly 3 things: transparency, consistency, and redundancy. (of specs and/or interfaces.) If none of those are at issue here then I don't see that we have much reason to be involved.

<masinter> jar, what about 'completeness' ? if people are nibbling at the edges

<jar> masinter, I might add that to my list, although coverage/completeness seems to be a linear combination of the other three

TBL: I may want to allow DKA access to my contacts, w/o letting him use BadGuyApp to see them, because I know that BGA does bad things with contacts

NM: I don't mind keeping a watching brief, what I mind is not commiting to that vs. writing a finding

<Zakim> masinter, you wanted to talk about scheduling

LM: Couldn't there be a different kind of deliverable?
... For things we're watching, w/o committing to intervening at the architectural level

NM: We can do that, we don't need a process decision -- my concern here is on getting agreement on the level at which we take this forward

DKA: I think there is a specific issue, which is architectural, albeit a small one: should a packaged webapp implicitly hold greater privileges than one hosted on the web?
... Seems to me the TAG could give guidance on this
... Needn't be a finding, could just be email to the WG chair(s)

<masinter> registerProtocolHandler should have the same kind of 'system' priveleges

TBL: My inclination is that zipping should be architecturally indistinguishable from caching, and so should have no privilege implications

NM: There's an install step

DKA: And a signature

<masinter> Signature, virus scan, insure provenence

NM: zip itself doesn't seem like the crucial bit

LM: registerProtocolHandler, wrt its security model, needs to be treated as if it is an installation
... because you're making a permanent change

LM: There's no manifest, there's nothing to install, but it's really very similar

RB: Returning to hosted vs. packages -- I agree that the zip vs. cache point isn't central
... A packaged app isn't allowed to external resources w/o specific authorization
... It's 'frozen'
... So you're not e.g. loading jquery from elsewhere, where it might have suffered an injection attack
... So it's the frozen/sealed nature of the 'packaged' app that matters

HST: So that's why signing makes a difference, because all you're using is what was signed

NM: So next?

DKA: Should we send something back to SysApps?

NM: Decide now about rolling monitor, or eventual finding, or wait to decide after TPAC

AM: wait to decide after TPAC

<masinter> Dan, what would we send to SysApps that would be useful?

<noah> ACTION-717?

<trackbot> ACTION-717 -- Ashok Malhotra to propose outline for possible TAG document (finding or rec) on architectural issues relating to storage sync, linked data, etc. Due 2012-07-15 -- due 2012-10-10 -- OPEN

<trackbot> http://www.w3.org/2001/tag/group/track/actions/717

<noah> close ACTION-717

<trackbot> ACTION-717 Propose outline for possible TAG document (finding or rec) on architectural issues relating to storage sync, linked data, etc. Due 2012-07-15 closed

<noah> ACTION: Ashok based on TPAC feedback to propose either specific TAG goal/product page on offline apps, or propose future occasional discussion w/o deliverable, or propose to drop it - Due 2012-11-27 [recorded in http://www.w3.org/2001/tag/2012/10/09-minutes.html#action10]

RB Action review

<noah> ACTION-700?

<trackbot> ACTION-700 -- Robin Berjon to send note to tag@ that he will send later to the AB (as himself) proposing the changes to electoral proceedings -- due 2012-11-30 -- OPEN

<trackbot> http://www.w3.org/2001/tag/group/track/actions/700

RB: Overtaken by events, AB already discussing it

<noah> close ACTION-700

<trackbot> ACTION-700 Send note to tag@ that he will send later to the AB (as himself) proposing the changes to electoral proceedings closed

NM: Tactical voting in particular?

RB: Not sure

<noah> ACTION-686?

<trackbot> ACTION-686 -- Robin Berjon to try to find who is in charge of the current browser content sniffing clustermess, and see if there is a way of moving out of the quagmire -- due 2012-09-18 -- OPEN

<trackbot> http://www.w3.org/2001/tag/group/track/actions/686

RB: I did that -- it is Adam Barth

LM: I heard he had withdrawn

RB: I didn't detect any way out

LM: I proposed that HTTP2 should get us out, but they declined
... Reassign it to me, with a 3 month deadline

<noah> ACTION-686?

<trackbot> ACTION-686 -- Larry Masinter to try to find who is in charge of the current browser content sniffing clustermess, and see if there is a way of moving out of the quagmire -- due 2013-01-08 -- OPEN

<trackbot> http://www.w3.org/2001/tag/group/track/actions/686

<noah> ACTION-685?

<trackbot> ACTION-685 -- Robin Berjon to create a product page proposing the Task Force on Web Security/Privileges/Trust/etc. -- due 2012-09-30 -- OPEN

<trackbot> http://www.w3.org/2001/tag/group/track/actions/685

<noah> close ACTION-685

<trackbot> ACTION-685 create a product page proposing the Task Force on Web Security/Privileges/Trust/etc. closed

<noah> ACTION-713?

<trackbot> ACTION-713 -- Robin Berjon to incorporate into Privacy document "There are concerns wrt protocols and privacy, there are concerns wrt APIs and privacy. We're working on APIs, IETF are working on protocols, we will collaborate and try to converge at least on terminology going forward" -- due 2012-09-13 -- OPEN

<trackbot> http://www.w3.org/2001/tag/group/track/actions/713

<noah> ACTION-715?

<trackbot> ACTION-715 -- Robin Berjon to prepare draft FPWD of some patterns for privacy in APIs, to be reviewed by TAG and published -- due 2012-09-13 -- OPEN

<trackbot> http://www.w3.org/2001/tag/group/track/actions/715

NM: Anyone want to pick this up

<masinter> push to http://www.w3.org/Privacy/ or privacy interest group??

AM: This was narrowly focussed originally, then RB added a thread to it
... As it stands it needs more material to make sense of the addition

NM: Would you do work on it?

AM: To reduce the scope back to where it was, yes

LM: Just forward it to the Privacy Activity/Interest Group

NM: Is there a group that meets?

LM: Yes, they have telcons

AM: Privacy is this large amorphous area -- the good thing about this document originally was it had nice clear boundaries

DKA: Publish it as it stands as a note

NM: PL, can you help?

PL: Yes

NM: Please take a look and see if you can identify what we would / might publish as a Note?

LM: DKA, RB, could you draft a SotD you would be happy with?

HST: Could we answer NM's question first, please

NM: OK, PL, please include consultations with the original authors on taking this out

<noah> close ACTION-713

<trackbot> ACTION-713 Incorporate into Privacy document "There are concerns wrt protocols and privacy, there are concerns wrt APIs and privacy. We're working on APIs, IETF are working on protocols, we will collaborate and try to converge at least on terminology going forward" closed

<noah> close ACTION-715

<trackbot> ACTION-715 prepare draft FPWD of some patterns for privacy in APIs, to be reviewed by TAG and published closed

<noah> . ACTION Peter to prepare for review a draft of a TAG Note focusing on the minimization message in the privacy finding. PL to check w/Robin and Dan first. - Due 2013-11-27

<noah> ACTION Peter to prepare for review a draft of a TAG Note focusing on the minimization message in the privacy finding. PL to check w/Robin and Dan first. - Due 2013-11-27

NM: RB, you're now clear of actions, thanks for your help

RB: I'm still happy to host in Paris in spring 2013

NM: Thanks to BCS for providing facilities

[Adjourned]

Summary of Action Items

[NEW] ACTION: Ashok based on TPAC feedback to propose either specific TAG goal/product page on offline apps, or propose future occasional discussion w/o deliverable, or propose to drop it - Due 2012-11-27 [recorded in http://www.w3.org/2001/tag/2012/10/09-minutes.html#action10]
[NEW] ACTION: Ashok working with experts in security community, to suggest projects TAG might undertake relating to security - Due: 2012-11-20 [recorded in http://www.w3.org/2001/tag/2012/10/09-minutes.html#action07]
[NEW] ACTION: Ashok, working with experts in security community, to suggest projects TAG might undertake relating to security - Due: 2012-11-20 [recorded in http://www.w3.org/2001/tag/2012/10/09-minutes.html#action06]
[NEW] ACTION: Jeni to draft rough product page / briefing pape for "distributed web applications" [recorded in http://www.w3.org/2001/tag/2012/10/09-minutes.html#action04]
[NEW] ACTION: Jeni to with Larry work out what the exit criteria from CR for fragids best practices should be [recorded in http://www.w3.org/2001/tag/2012/10/09-minutes.html#action01]
[NEW] ACTION: Larry with help from Ashok (and TAG), 1) decide exit criteria on Publishing & Linking 2) Rec track vs. Finding 3) Update product page to match - Due 2012-10-16 [recorded in http://www.w3.org/2001/tag/2012/10/09-minutes.html#action03]
[NEW] ACTION: Larry to frame for telcon discussion possible TAG work relating to DWIM and Issue errorHandling-20 - Due 2013-11-13 [recorded in http://www.w3.org/2001/tag/2012/10/09-minutes.html#action08]
[NEW] ACTION: Larry, with help from Ashok (and TAG), 1) decide exit criteria on Publishing & Linking 2) Rec track vs. Finding 3) Update product page to match - Due 2012-10-16 [recorded in http://www.w3.org/2001/tag/2012/10/09-minutes.html#action02]
[NEW] ACTION: Noah to send Steve Zilles note saying Henry, Larry, Peter, Ashok, Tim and Jeni (Wed) could meet with AB [recorded in http://www.w3.org/2001/tag/2012/10/09-minutes.html#action09]
[NEW] ACTION: Noah to think about how to evaluate results vs. success criteria on closed work - Due 2013-01-01 [recorded in http://www.w3.org/2001/tag/2012/10/09-minutes.html#action05]

Minutes formatted by David Booth's scribe.perl version 1.135 (CVS log)
$Date: 2012-11-27 01:57:06 $