See also: IRC log
<scribe> scribenick: jar_
<scribe> scribe: Jonathan Rees
<noah> RESOLUTION: minutes of 31 March 2011 http://www.w3.org/2001/tag/2011/03/31-minutes are approved
Question of whether to meet next week
<ht> My participation on 28 April is uncertain
Noah: So what were all the high level topics that were taken up?
Henry's final slide deck is here: http://www.w3.org/2001/tag/doc/IAB_Prague_2011_slides.html
These are what he presented from in Prague, except that he ended up skipping slides 16-20.
ht: I need to remind people... at our F2F when I took this on, I didn't know how big this was.
... There were 800 or so people in the room, not the 50 I expected.
... There are intense undercurrents around what IETF means by "application"
... different opinions regarding relative importance of 'real time protocols' (e.g. SIP)
... I learned a lot in the 24 hours prior to the talk and significantly revised the slides
... IETF have been around for 30? years. 1000 people at meeting instead of 3000 formerly
... some people have been with IETF through their entire careers
... W3C doesn't necessarily loom large in the view of those involved in IETF
... For IETF, the main concern is physical architecture. Very hard edged, engineering viewpoint
<noah> Yes, as I said often at the time the TAG was formed, the W3C's use of the word "architecture" confused me for many years. I've learned to get on with it, and what the TAG does is useful, but it's somewhat more informal than what I "grew up" learning as architecture.
ht: not about best practices & guidelines (for the most part).
... For much of IETF's history, IAB ran the IETF - had authority.
... Mine was the first talk at the plenary event. I cut most of the examples, did talk about resource loading
... Added to the talk at request of chair: slides on HTML5 and the W3C Web platform idea
... The big question there was, is there no need for standards and protocols (sensu IETF) any more?
... Debate. Middle ground: yes, there will be more use of HTTP, but there will also be new protocols.
<Yves> See http://tools.ietf.org/search/bcp56
<Yves> "On the use of HTTP as a Substrate"
ht: Afterwards, a number of people from IAB said, yes our two groups should meet together some time
... calendaring nightmare of course.
... There's an IAB retreat coming up, potential for interaction there, maybe
ht: Jim Gettys says hi
... talked about buffer bloat
masinter: Architecture in IETF is nuanced.
... The IAB was an inspiration when the TAG charter was written, with some talk about need to liaise
<Larry> Jon Peterson
<Larry> I'd encourage Noah to discuss with IAB chair possibility of a joint meeting
<Larry> and also consider ongoing liaison
<Larry> including reviewing of documents of mutual interest
<Larry> e.g. IAB has a document on protocol extensibility, which reflects on versioning
<Larry> Major topic [at meeting]: registries and IANA
larry: There was a discussion including Mark Nottingham, Thomas Roessler, and IANA and application area directors re registries
<Larry> Now there is a new mailing list, "Happy IANA"
<Larry> and a wiki with problem & proposed solutions
larry: hope we can resolve registries, mime type, etc. I need to remove administrative / process issues from my draft
... they can be fixed separately from technical content
<Larry> this from draft-masinter-mime-web-info IETF internet draft i will remove administrative issues
larry: e.g. fragment ids and magic numbers for sniffing are different from registration process tracking
... Some other webarch-related topics were discussed at the IETF meeting
... e.g. IRI
... short discussion of sniffing
... do not track header. Will anyone from the TAG be at the upcoming Princeton workshop?
<DKA> I won't be there.
<timbl> I won't -- have been in previous ones but can't make that one
<Larry> what i'm looking for and haven't seen is the threat analysis
larry: I haven't seen a threat analysis of the do-not-track idea
... E.g. look at Paypal position paper - what are we trying to prevent? Use that as a way to evaluate any given proposal
timbl: Privacy workshops have been going on for some time, [would be surprised if this hasn't been done]
<Yves> DNT can't solve the issue of people not respecting rules, just limit the privacy leakage form sites with a good reputation
noah: Are you concerned that this effort isn't well grounded in use cases?
larry: Yes, afraid people are jumping into solutions without a clear understanding of the problem.
timbl: Privacy has come to prominence recently, lots of papers and analysis, threat models... no comprehensive solution yet, but lots of prior work... doesn't look like a rush job to me
noah: I assumed threat analysis was implicit, established
<Yves> do-not-track is trusting good sites, api minimization is not trusting bad sites, so yes completely different approach and kind of solution
<Larry> would like some way for TAG review Princeton workshop to see if there are TAG issues to address
noah: Here's something we don't quite understanding. How important to find out?
larry: Workshop is in Princeton in a couple of weeks. Would be good if someone went.
<ht> Background doc. for the IAB panel at IETF-Prague: http://tools.ietf.org/html/draft-tschofenig-post-standardization-00
Ashok wrote in email on March 4: "It's easy for me to drive to Princeton and those dates look open so far.
So, I'll be happy to cover this for the TAG as long as I don't have to write a position paper"
<Larry> I think the mandate would be just to observe and come back with any architectural issues we might want to take up
noah: Let's ask Ashok to attend.
<Larry> look at long-term issues and maybe a post-mortem on P3P
<ht> One of the authors is Jon Peterson, who a) chaired the panel and b) is on the IAB
<ht> IAB membership: http://www.iab.org/about/members.html
lm: [ask Ashok to] observe and come back with any architectural issues that come up
<noah> . ACTION Ashok To report to TAG, after privacy workshop, regarding architecture issue on privacy and especially degree to which "do not track" meets real use cases
<noah> ACTION Ashok To report to TAG, after privacy workshop, regarding architecture issue on privacy and especially degree to which "do not track" meets real use cases
<trackbot> Created ACTION-545 - Report to TAG, after privacy workshop, regarding architecture issue on privacy and especially degree to which "do not track" meets real use cases [on Ashok Malhotra - due 2011-04-21].
<Larry> rather than negative, look at positive: what other real use cases need additional privacy work?
<Larry> "especially degree in which other work beyond DNT is needed"
<noah> Report to TAG, after privacy workshop, regarding architecture issue on privacy and especially degree to which use cases beyond those addressed by "Do Not Track" need attention
<trackbot> ACTION-545 -- Ashok Malhotra to report to TAG, after privacy workshop, regarding architecture issue on privacy and especially degree to which use cases beyond those addressed by "Do Not Track" need attention -- due 2011-04-21 -- OPEN
<noah> ACTION-545 Due 2011-05-03
<trackbot> ACTION-545 Report to TAG, after privacy workshop, regarding architecture issue on privacy and especially degree to which use cases beyond those addressed by "Do Not Track" need attention due date now 2011-05-03
larry: Jim G has been tracking source of latency in net
... much of it is around excessive buffering
<ht> IAB Chair situation: "Bernard Aboba replaced Olaf Kolkman, who stepped down"
<timbl> [this is] below our level, but interesting in that it is one of the things we depend on.
<Larry> "Buffer Bloat"
noah: Sun WSs way back when had great [i.e. low] latency, due to fewer copy operations
<noah> Excessive buffer copying has been a recurring implementation issue for 25 years. It's an important implementation issue.
ht: Noah, this has nothing to do with Jim's work. Let's take it up over lunch
timbl: It's about misconfiguring their routers...
<timbl> Brian Carpenter used to say that network systems perf is just a question of round trips and times the data is copied.
larry: and lots of other things too.
<noah> Proposal: we're having a F2F in June. Should invite JG for lunch + a session?
<timbl> That is is different.
<timbl> It is below our level
<noah> HT: At least lunch. May not be a TAG issue.
<timbl> Somebody Else's Problem (tm)
<noah> ACTION: Noah to ask Jim Gettys about joining us for lunch at June F2F Due 2011-05-15 [recorded in http://www.w3.org/2011/04/14-tagmem-irc]
<trackbot> Created ACTION-546 - Ask Jim Gettys about joining us for lunch at June F2F Due 2011-05-15 [on Noah Mendelsohn - due 2011-04-21].
<ht> Noah, please note: Bernard Aboba IAB Chair bernard_aboba AT hotmail.com
larry: (missed) Thomas has taken it up. If other staff members etc. involved in IANA etc then ...
<noah> thanks, ht
larry: Another major topic at meeting: HTTPbis. Testing can start soon
... URNbis is also proceeding, has repeated that http: URIs aren't enough.
<trackbot> ACTION-542 -- Larry Masinter to and Henry to report on IETF Meeting Due 2011-04-07 -- due 2011-03-31 -- OPEN
<noah> close ACTION-542
<trackbot> ACTION-542 And Henry to report on IETF Meeting Due 2011-04-07 closed
<trackbot> ACTION-540 -- Larry Masinter to try to arrange for Thomas Roessler to participate in the meeting about Registries at the IETF meeting in Prague -- due 2011-03-24 -- CLOSED
<trackbot> ACTION-539 -- Larry Masinter to liaise with Thomas Roessler about the registries issue background -- due 2011-03-24 -- PENDINGREVIEW
<noah> LM: 539 is effectively with Thomas now.
<noah> close ACTION-539?
<Larry> yes, think it doesn't need TAG follow up
<trackbot> ACTION-519 -- Peter Linss to frame architectural opportunities relating to scalability of resource access -- due 2011-04-05 -- OPEN
<Larry> that was a topic that Henry didn't have a chance to really go into it in depth, and maybe we should talk about with IAB
<timbl> URNP 1.1
ht: From the floor, Dave Crocker suggested, put them [the overfetched resources] in the DNS.
<Larry> suggest Peter talk to Dave Crocker about Dave's idea
<noah> ACTION-519 Due 2011-04-26
<trackbot> ACTION-519 Frame architectural opportunities relating to scalability of resource access due date now 2011-04-26
larry: In summary - I was pleased - there are great opportunities for coordination between IETF and W3C at architectural level
<DKA> Current draft here: http://www.w3.org/2001/tag/doc/publishingAndLinkingOnTheWeb.html
<noah> Maybe we want to say roughly what Larry is saying: I.e. give relatively narrow definitions of "cache", "archive", etc. then say: in actual Web systems, software makes copies sometimes for cacheing, sometimes for archiving, or sometimes for combinations of such reasons.
<JeniT> I agree
dka: Want to clarify re URIs and public identifiers even outside the context of linking...
... maybe Rigo's point is outside the scope of the document?
<Larry> Caching and archiving are two examples of making copies of stuff you got from somewhere else, common elements of distributed system design. The web's contribution (around "404 not found")
dka: Rigo said, if you use a namespace with a controversial name[space?], that's another use of a URI that could be a problem...
Jeni: We're talking about the actual consequences of the URIs as links, not URIs as names in namespaces
noah: We talked about what the TAG's role might be in this... settled on a terminology document as starting point?
dka: Yes, that's what this is. Need some more iterations on the document. Aiming to say something that is not legal but might be relevant in legal (or whatever) discussions... not there yet
<JeniT> no problem
noah: Pls give some guidance for agenda preparation & scheduling
... Re German high court decision, do we need to talk about this?
<noah> In particular, do I need to track any followup activity relating to it?
<DKA> Jar: it would be useful to have a set of legal questions in the document.
<noah> JAR: Thinking about this over the past week or two, how do we finesse legal/technical issues. It may be useful to have a set of legal questions that we have in mind. We can't speak to the questions themselves, but they are sort of use cases for us.
<noah> JAR: Some of these are things that have yet to be decided by courts.
<noah> NM: Though in the German case, Rigo pointed to a high court decision
<Larry> are there "expert witness" technical statements from previous court cases over linking?
<trackbot> ACTION-541 -- Jeni Tennison to helped by DKA to produce a first draft of terminology about (deep-)linking etc. -- due 2011-04-12 -- OPEN
<noah> ACTION-541 Due 2011-04-26
<trackbot> ACTION-541 Helped by DKA to produce a first draft of terminology about (deep-)linking etc. due date now 2011-04-26
larry: mailing list discussion has been productive
<trackbot> ACTION-480 -- Daniel Appelquist to draft overview document framing Web applications as opposed to traditional Web of documents -- due 2011-04-12 -- OPEN
<noah> ACTION-480 Due 2011-04-26
<trackbot> ACTION-480 Draft overview document framing Web applications as opposed to traditional Web of documents due date now 2011-04-26
<trackbot> ACTION-537 -- Daniel Appelquist to reach out to Web apps chair to solicit help on framing architecture (incluing terminology, good practice) relating to interaction -- due 2011-04-05 -- OPEN
<noah> ACTION-537 Due 2011-04-26
<trackbot> ACTION-537 Reach out to Web apps chair to solicit help on framing architecture (incluing terminology, good practice) relating to interaction due date now 2011-04-26
<trackbot> ACTION-514 -- Daniel Appelquist to draft finding on API minimization Due: 2011-02-01 -- due 2011-04-12 -- OPEN
dka: Re 514, not clear on what process to use to progress this document
dka: need more material, review
... at F2F, I suggested this be a small document. Is it time to circulate it more widely?
<noah> ACTION Dan to ask on www-tag and Web Apps mailing lists for advice on moving forward with http://www.w3.org/2001/tag/doc/APIMinimization.html
<trackbot> Created ACTION-547 - Ask on www-tag and Web Apps mailing lists for advice on moving forward with http://www.w3.org/2001/tag/doc/APIMinimization.html [on Daniel Appelquist - due 2011-04-21].
<trackbot> ACTION-510 -- Tim Berners-Lee to write a note conveying the TAG's concerns re: the microdata -> RDF URI mappings in the HTML5 microdata draft Due: 2011-01-20 -- due 2011-03-09 -- OPEN
<noah> ACTION-510 Due 2011-04-26
<trackbot> ACTION-510 Write a note conveying the TAG's concerns re: the microdata -> RDF URI mappings in the HTML5 microdata draft Due: 2011-01-20 due date now 2011-04-26
<trackbot> ACTION-404 -- Yves Lafon to track HTML WG ISSUE-27 rel-ownership -- due 2011-04-08 -- OPEN
<noah> ACTION-404 Due 2011-04-26
<trackbot> ACTION-404 Track HTML WG ISSUE-27 rel-ownership due date now 2011-04-26
<Larry> this action is tied to the IANA discussions
<trackbot> ACTION-341 -- Yves Lafon to follow up with Thomas about security review activities for HTML5 -- due 2011-04-05 -- OPEN
larry: Re the rel-ownership issue, check with Thomas and Mark N
yves: need to move fast
<trackbot> ACTION-341 -- Yves Lafon to follow up with Thomas about security review activities for HTML5 -- due 2011-04-05 -- OPEN
<noah> ACTION-341 Due 2011-05-10
<trackbot> ACTION-341 Follow up with Thomas about security review activities for HTML5 due date now 2011-05-10
<timbl> The URNP protocol may be used to dereference any URN. URNs are here the set of identifiers for which there exists no dereferencing protocol.
<DKA> Larry can you post something on it to www-tag? I think it's worth generating some discussion on this issue.
<DKA> I think "browser-centricity" is relevant to the question around packaging of web applications...