======================================================================== AppsArea Meeting Minutes IETF 79 Monday, November 8, 2010, 0900-1130 (Morning Session I -- Garden Ballroom 1) Chairs: Alexey Melnikov and Peter Saint-Andre Scribe: Ted Hardie Agenda: Slides: Audio: Chat Log: ======================================================================== MINUTES 1. Administrivia Note well reviewed Logistics reviewed Agenda bashed Alain Durand's presentation was cancelled because his co-chair was not available to cover for him. 2. First talk: MIME and the Web Document by Larry Masinter, presentation by Alexey Issues reviewed, then recommendations discussed. John Klensin's capsule review: thumbs down Ted Hardie commits to providing a review to apps-wg Tobias Gondrom asks if Larry Masinter knew about the MIME-sniffing draft; Alexey believes the answer is yes. 3. Second talk: Referral Objects (draft-carpenter-referral-ps-01) Brian Carpenter speaks on the problem statement for referral objects. Scope and referrals discussed. Harald Alvestrand suggests that referrals are the wrong question: connections are the correction. The other way to look at is the connection and the referral are two separate problems and ICE is about connections, not referrals. Pete Resnick notes that the footnote that path selection and referral aren't the same problem is a layer violation. Applications don't want to worry about reachability -- but that implies that we need to fix the problem at the network layer, rather than fixing in application layer. Dave Crocker notes that this may be IRTF work in that the problem may not yet be an engineering task. 3. Chairs mention Eliot Lear's draft on possibly migrating the time zone database to the IANA (draft-lear-iana-timezone-database) and ask for reviews. Barry Leiba asks whether it should go into APPSAWG, but Alexey says it is a bit early to discuss that, since it was not clear the work was coming to the IETF (Unicode is also being considering, for example). 4. Third talk: John Klensin on IDNA and Unicode 6. He is giving the "hostile" version, as it is funnier. Provision for backward compatibility, with presumption would be used rarely. But, as it turns out, there are already 3 incompatible changes. Two choices: a. pretend it never happened. b. create exceptions using available mechanism to freeze these in 5.2, maintain exception table. General agreement that writing an I-D to document what happened would be valuable, modulo objections from Pete Resnick. Harald suggests using the APPS Area Wg for further discussion. Marc Blanchet observes that these Unicode changes are normal, whatever they guarantee. Therefore we should plan on managing changes instead of dreaming that it won't happen. John notes that this may cause the need for working groups on PRECIS-like topics to note that this will happen. 5. Miscellaneous topics Ted Hardie asks for folks to review draft-hardie-mdtls. Hannes Tschofenig reiterates his pointer to the OAUTH drafts and tutorial (after the plenary on Wednesday in Jade). Harald Alvestrand mentions RTCWeb, which is trying to do the same thing using ICE and RCP. RTCWeb did not get lined up for this meeting, but it will be emerging soon. It is noted that on Wednesday after the Plenary, HTTP streaming discussion will be held in the Emerald room. 6. APPSAREA Working Group (APPSAWG) Barry Leiba and Jiankang Yao to chair. Barry outlines the purpose of the APPSAWG: ensuring reviews of the documents that might not other get sufficient review. First order of business discussing adoption strategy. Jon Peterson asks what our immune system is for avoiding problem spaces that cannot achieve consensus. Under discussion. Much additional discussion of pushing to a new WG first, rather than assuming APPSAWG. Humm taking on participation. A handful of folks raised their hands. Barry mentions the three drafts already suggested: draft-nottingham-http-portal, Patrik Faltstrom's IDNAbis I-D (already discussed), and draft-merrit-jms-uri. Discussion of how to focus continued for some time. Paul Hoffman raises issues about what adoption will mean for documents that are accepted by whose authors go missing. Jon Peterson describes the slippery slope to the DISPATCH WG, on the gliding paths of requirements documents. Marc Blanchet says that this WG should not be working like DISPATCH in the RAI area. Jon suggests that folks not be constrained in their thinking of what a working group is, and to note that some structures may have another purpose than the usual constituency-completes-milestones view of a WG. END ========================================================================