See also: IRC log
<monica> rama will not be able to join today
<bob> zakim [microsoft is mrgoodner
<scribe> scribe: Paul
minutes of March 19 approved
Bob: Several new issues added to mail list
<bob> http://lists.w3.org/Archives/Public/public-ws-addressing-comments/2007Mar/0000.html
Bob: this one describes a pretty
clear typo
... This one is an editorial item, resolved as proposed.
<bob> http://lists.w3.org/Archives/Public/public-ws-addressing/2007Apr/0003.html
Bob: Next: attaching policy to an EPR.
TRutt: At WS-Policy, the feeling
was that WS-Addressing should address it
... It should be pretty fast, about 3 pages
MrGoodner: Skeptical of how fast it will be
gpilz: Can it be in metadata spec?
David: Don't think it should be in metadata spec.
Bob: Earlier in WSA we did not
feel we could address it well, and returned it to WS-Policy
group
... NOt sure we have enough domain knowledge in the group to do
it. It is more in the purview of WS-Policy.
<anish> i would hope the ws-policy wg would review our spec
TRutt: It is not rocket
science.
... We just need to come up with a way to reference policy
EPR.
Anish: It should be done either here or in WS-Policy. We could potentially do it if we have the expertise. Perhaps we could form a joint task force with WS-Policy.
Bob: record this as a new item
<scribe> ACTION: Bob to coordinate with WS-Policy about some way to address it jointly. [recorded in http://www.w3.org/2007/04/02-ws-addr-minutes.html#action01]
<bob> http://lists.w3.org/Archives/Public/public-ws-addressing/2007Apr/0008.html
Bob: Next Item from Anish
... We will accept this as a new issue.
... Now for the main event: F vs. G
... Consensus forming around G or G'; little support for F
TRutt: G is less verbose in most normal cases
Bob: Any objectin to using G as
the base for further wordsmithing?
... No objection, we will use G going forward.
TRutt: Monica had some important points
<monica> http://lists.w3.org/Archives/Public/public-ws-addressing/2007Apr/0007.html
Monica: Describes her items as addressed in her message.
<anish> i should point out that this ties in with one of my comments too
TRutt: How do we position these
additional assertions?
... It's a matter of how we word it, to imply we are adding
limitations.
Monica: When you have an empty policy, there is no claim. It does not infer behavior.
MrG: The nested addressing assertions imply further processing like the mixed mode.
TRutt: It is a user refinement of the capabilities.
Monica: Why not put a section in WSA that all these types are supported?
Tony: We have not written such a section yet.
Monica: Would like to see such a section.
Anish: It seems counter-intuitive that both anon and non-anon would be supported.
MrGo: Don't see why you need to explicitly state the ability.
<monica> Explicitly: An alternative with zero assertions indicates no behaviors. An alternative with one or more assertions indicates behaviors implied by those, and only those assertions.
<monica> Can't infer from empty.
<monica> When an assertion whose type is part of the policy's vocabulary is not included in a policy alternative, the policy alternative without the assertion type indicates that the assertion will not be applied in the context of the attached policy subject.
Anish: Would like to be able to state that you can have ReplyTo could be anon and FaultTo can be non-anon
<David_Illsley> monica, the alternative would include an assertion, the wsam:Anonymous one for which we define the semantics
MrGo: Not sure what you want here.
<bob> axk gp
gpilz: Anish is talking about granularity.
<anish> i agree with gil, that this is tangential to the current discussion
<anish> that is what i was trying to estabilish, that split usecase is not supported by any of the proposal (including mine)
gpilz: It is tangential.
<David_Illsley> anish, I agree with your statement that we don't need to support that detailed usecase
TRutt: It is one of our use cases
we need to support. It's enough that the implementation
supports either anon or non-anon.
... Hope we can come up with wording to get around empty
problem.
Monica: Empty does not tell you anything. See text pasted above.
Bob: Would an empty nested
assertion satisfy the need? I've heard it described in various
ways - asserts no influence, etc.
... We're grappling with the nuances.
MrGo: The proposal I sent in was vetted with our (Microsoft) policy guys.
David: Wouldn't outer non-nested assertions indicate the behavior?
Monica: If it's not there, you can't imply anything from it - null behavior. It is policy alternative vocabulary.
TRutt: The algorithm treats empty as having some implied restriction. If we can come up with some text to deal with it...
Anish: It is counterintuitive.
I'd like an assertion that says "I support addressing."
... If I don't put it in, it should not imply that I support
all protocols.
<dhull> s/find/finds/
Anish: What if you make a WSA claim on an endpoint that only supports one-way operations?
Bob: Even one-way MEP has a way to support faults.
Ram: The absence of top-level or the absence of the nested assertion should be handled differently.
Monica: The primer does not address the nested assertion explicitly.
Ram: The nested assertions are to qualify the top-level assertion.
Bob: Should we construct a well-worded question to send to the WS-Policy group?
Anish: The spec does not say that if you support WSA then you must support anon or non-anon.
Bob: the WSDL spec and now the Metadata spec elaborate on this.
<dhull> +1 to WSA is more than response endpoints
Anish: We are focusing on anon and non-anon, but there are other cases we need to consider.
Mrgo: This is a new use case we have not considered before.
Anish: We should have an assertion that simply expresses support for WSA.
MrGo: I thought that was what proposal G was for.
Anish: I suggest adding some explanation disclaiming support for either anon or non-anon. I think we actually agree.
TRutt: That capability was
considered several proposals ago. F & G are simple use
cases.
... The intersection algorithm precludes some approaches.
Anish: If we go with the clarification I am suggesting, would a client wanting non-anon response EPRs, would it not find a match?
TRutt: We need to understand the use cases.
David: My concern is with letting the empty case imply support for specific response types.
MrGo: If you require mixed use case, you need nested policies to express it.
Monica: Is that domain-specific processing?
MrGo: No, it doesn't have to introduce domain-specific processing.
Anish: The top-level assertion does not make claims, without the nested assertions.
gpilz: I'm confused by this
discussion.
... WS-Policy defines vocabularies in the context of an
assertion, which make it more complicated.
<TRutt_> qa+
Anish: yes, with nothing inside, it has different implications.
<anish> +1 to getting clarification from ws-policy wg
gpilz: Back to Bob's proposal to put pointed questions to WS-Policy.
TRutt: The problem stems from Chris' inability to prevail on the meaining of empty in the WS-Policy group. That group is split on this issue. We should problably write what we want to support our use cases.
MrGO: We discussed the negation issue. We don't think it's an issue here.
Anish: Somebody should work with WS-Policy on this.
Monica: Work is still going on in the Policy WG.
Bob: Can we package alternative "G" in a question for WS-Policy? Is that a reasonable thing to do now?
MrGo: Can we begin by adopting proposal G?
TRutt: G with my proposals?
Anish: No problem with that.
Bob: We seem to be resolved to incorporate G with Tom's subsequent wordsmithing.
<monica> does that include anish's comments?
TRutt: I think Tony can do the wordsmithing.
<anish> ping
Bob: We have resolved to do what we just discussed, insert the text of G with Tom Rutt's additional text.
<monica> http://lists.w3.org/Archives/Public/public-ws-addressing/2007Apr/0010.html
Monica: that is a link to Anish's comments.
Anish: Why should we have this restriction described in my comments?
Bob: the WS-Policy group indicated it could not be done, due to the level of abstraction. It was in Chris Ferris' message to WSA.
Anish: I will look at that.
... Not clear if we have defined this assertion closely enough.
What about SOAP vs. non-SOAP binding?
<bob> chris' response http://lists.w3.org/Archives/Public/public-ws-addressing/2007Feb/0006.html
TRutt: This sounds like a new issue.
Bob: Anish, can you submit
something on this topic?
... Please extract the issue.
<monica> yes, that is the one
Anish: It is in comment number 1.
Bob: It is in point E in Chris's message.
Anish: It does not really explain
why.
... There were two issues in my comment... the second is what
it is claiming conformance to.
Bob: I will take the contents of the email and make it two issues.
Anish: okay
<scribe> ACTION: Bob split Anish's email to make two issues. [recorded in http://www.w3.org/2007/04/02-ws-addr-minutes.html#action02]
Bob: I hope we can preserve
proposal G, with clarifications, to send back to WS-Policy. It
should be ready before the next last call round.
... It would be good to finish before the last call , so we can
get feedback early from WS-Policy instead of having it raised
as an issue at last call.
... Once we think we have a resolution, we should send it to
WS-Policy, according to the W3C policy.
Trutt: We can send it as a reference to the new editor's draft.
Bob: Better to send it explicitly
to WS-Policy.
... Tom, can you send me the wordsmithed text to modify G.
Anish: Gil's question about negation, with interpretation of the absence of a nested assertiion implying negation, is that resolved?
MrGo: Don't need to deal with that now, with WS-Policy.
Anish: That seems to take us back to proposal F, not G.
Bob: The nuance is the point at which the vocabulary is defined.
<monica> there is no negation in current ws-p
Bob: Maybe we should excise the assertions if they are so difficult to understand.
<anish> if we all agree that there is no negation, then i don't have a problem. But like Gil, I remember Chris ferris vehemently stating the negation part
<monica> which chair?
Bob: I have solicited one of the WS-Policy co-chairs' opinion on this, and it appeared that either F or G is possible; G was a bit better.
<anish> it would be easier if the policy wg resolved this issue once and for all
Bob: Suggest a call next
Monday... It is a bank holiday in UK... ?
... Aim for April 16th for next meeting. We need as much
attendance as possible.
... Meeting adjourned due to lack of linear arguments.
... As of today, we have resolved on G. A through F are dead.
There were no objections when the question was called.
Anish: Initiating circular discussion among WS-Policy members on the call. Minutes not taken.
<bob> Paul, thanks for scribing
This is scribe.perl Revision: 1.128 of Date: 2007/02/23 21:38:13 Check for newer version at http://dev.w3.org/cvsweb/~checkout~/2002/scribe/ Guessing input format: RRSAgent_Text_Format (score 1.00) Succeeded: s/i would/i would hope/ Succeeded: s/objectin/objection/ Succeeded: s/limitatins/limitations/ Succeeded: s/ike/like/ FAILED: s/find/finds/ No ScribeNick specified. Guessing ScribeNick: PaulKnight Found Scribe: Paul WARNING: No "Topic:" lines found. Default Present: Bob_Freund, Gilbert_Pilz, monica, David_Illsley, Anish_Karmarkar, TonyR, Tom_Rutt, ram, Paul_Knight, yinleng, David_Hull, mrgoodner Present: Bob_Freund Gilbert_Pilz monica David_Illsley Anish_Karmarkar TonyR Tom_Rutt ram Paul_Knight yinleng David_Hull mrgoodner Got date from IRC log name: 2 Apr 2007 Guessing minutes URL: http://www.w3.org/2007/04/02-ws-addr-minutes.html People with action items: anish bob email s split WARNING: Input appears to use implicit continuation lines. You may need the "-implicitContinuations" option. WARNING: No "Topic: ..." lines found! Resulting HTML may have an empty (invalid) <ol>...</ol>. Explanation: "Topic: ..." lines are used to indicate the start of new discussion topics or agenda items, such as: <dbooth> Topic: Review of Amy's report[End of scribe.perl diagnostic output]