See also: IRC log
Present: davidb dbooth hugo katia mike mikec suresh yinleng
Scribe: Any corrections on minutes?
<Scribe> ACTION: Katia to review Policy and Message models and update OWL by next week. [DONE]
<Scribe> ACTION: dbooth to clarify term "service provider" and "service requester" and expand glossary [DONE]
<Scribe> ACTION: Mike to add and wordsmith text in 3.11 choreography [PENDING]
<Scribe> ACTION: Mike to propose changes to WS Reliability section in stakeholders perspective [PENDING]
<Scribe> ACTION: MikeC to ping DavidOrchard for his input on the Resource [DONE]
<Scribe> ACTION: Suresh to review doc and propose text on EBXML [DROPPED]
<Scribe> ACTION: dbooth to remove the Management Model from section 2 [DONE]
<Scribe> ACTION: YinLeng to verify that the Management Stakeholder section is using the correct text [DONE]
<Scribe> ACTION: YinLeng to verify that management terms are ok in Glossary [DONE]
<Scribe> ACTION: MikeC to add Abbie's security text to next week's agenda [DONE]
Scribe: 3 questions at teh editors call:
... any existing text from security needs to be incorporated?
... what about the requirements, duplication with requirements doc
... should the securities spec be in the arch or in roger's list
Roger: needs to be preserved
... abbie's list is a good section. Needs to be preserved.
hugo: we had a section on trust and discovery added to abbies section
Scribe: duplication of requirements less obvious now, so it is now good glue for the security section
... the technologies: only place in the arch where technologies discussed. taken out for now.
Roger: I think the security specs deserve attention. The section is useful.
Abbie: i'm relaxed about this
... I added it because it was requested
Gerald: I want to see security addressed
Scribe: scribe votes for annex
... Mike propose an annex
Roger: I like the idea. The text has a higher level of consensus than my list.
<mchampion> dbooth, you're on IRC but not able to call in?
<mchampion> ACTION: hugo will incorporate Abbies list of security specs as an appendix
Scribe: ACTION: Hugo will incoporate Abibie's security as an annex
... The initial version of the OWL models are in http://www-2.cs.cmu.edu/~softagents/WSA_Ontologies/
Roger: hugo's version reads more smoothly
... SOA has two arch constraints: small set of simple interfaces
... descriptive messages
<mitrepauld> rat hole
Scribe: Roger suggests using mike wording: Messages sent in a platform-neutral, standardized format delivered through the interfaces.
<mitrepauld> Avoid "self-descriptive"
... Description is part of SOA.
... Third SOA constraint: describe everythign to death
<mitrepauld> 1.6.3 WWW is a SOA?
Scribe: Third SOA constraint: Services, messages are described in a machine processable format
Roger: section of REST is paltry
... we can be more aggressive about what we mean
... I proposed some text
oger: what should we say about REST?
Dbooth: Our arch permits both RESTful and non-RESTful Web services
Roger: No critical analysis of REST
<mitrepauld> soap 1.2 provides the guidelines
<hugo> Note that I have replaced in the document my #2 constraint by Mike's
Mario: REST does not include intermediaries security etc.
Roger: We should write a critical analysis of REST, SOAP etc
Mike: Will draft something
Roger: The world needs something on the meaning of REST
Scribe: ACTION: mike will draft text on REST
<yinleng> Frank, I've just checked CVS, the current version of section 3.8 is old, not my latest
<mitrepauld> Yin Leng: set of mgmt "operations" since wsdl 1.2 has single "interface"
Roger: we need a common reference version of the arch
pauld: The WWW is a SOA? Maybe not ...
<dbooth> We should all print the document at 4pm US Pacific time on Friday so that we are all looking at the same version.
... US pacific time is UTC-8.
Scribe: Any other comments on Abbie's security text?
... Liberty Alliance to go into security annex list
Scribe: Time proportional to pages
Scribe: Introduction needs a disproportionate amount of time
... Reliability needs time
... Remaining issues
... The future and what we don't do
<mitrepauld> MEP in 188.8.131.52
Scribe: MEP text needs special thinking
... Editorial comment: how about dropping 3.6? It seems vestigial
paulD: registry model seems blurry
davidB: authoratative applies to registry not to its contents
pauld: registries some are authoratative others are not.
Scribe: how do you spell authorotative?
Mike: registries are auth...ive because people tend to go there
Scribe: everyone is in agreement
<Scribe> ACTION: DavidB to revise wording on autho...ive
<yinleng> Should we say "recognized" place rather than "authoritative" place to look?
davidB: please ask your AC rep to voice your opinion
Katia: what happens to document, editing etc.
Scribe: Only mechanical changes can be sanctioned
... W3C can only promote the doc if there is an obvious continuation of some kind
... private individuals and companies can mention and reference it
... SWSL intends to reference it
Mike: plan to use it for evangelical purposes
Roger: also intends to pretend that its authorotative
<yinleng> Got to go now, have a good f2f!
<mitrepauld> got to go too, bye