See also: IRC log
Chair: Any corrections?
... Any objection to accepting the minutes for 4 April 2005
tlecon
<scribe> DONE: 2005-04-04: Anish Karmarkar to propose issue about relationship between destination property and WSDL
<scribe> DONE: 2005-04-04: Hugo Haas to review IPR details on test case documents
Expect today: 2005-04-04: Glen Daniels to prose an issue that captures the async requirements in the charter.
Expect for FTF: 2005-04-04: Paul Downey to create initial test cases document for discussion
<scribe> UNKNOWN: 2005-04-04: issue 021 - Francisco Curbera to revise proposal
<scribe> ONGOING: 2005-04-04: Hugo Haas to discuss issues surrounding WSDL document split with W3T/W3M, report back
Chair: Will get a revised agenda
out
... main topics: WSDL Binding, issue 21 and async work
... other work regards Last Call comments and test
document.
... My plan is to not spend the whole FTF on the WSDL doc,
depending on how many LC issues we have as of early next
week.
... Both process, and start talking about issues.
Tom: Is there a start time recorded?
Chair: Yes, in the mailed agenda.
<mnot> http://www.w3.org/mid/[email protected]
Tom: There is a reception area,
wait there for escort. On-site cafeteria and nearby
restaurants.
... Internet access, wired and wireless. Speakerphone to dial
into Zakim.
Hugo: Will check on the dial-in number.
Chair: June FTF registration up,
please register soon.
... Some pushback on the dates in WSDL.
... WSDL WG will meet T/W to avoid the holiday.
... WS-A Wed afternoon session cancelled, just meet Th/F.
... June 2-3.
Winkler: , cancel room for Monday?
Glen: Don't know the scope.
Winkler: Won't cancel till after this next FTF.
Chair: Our latest public spec for the WSDL Binding spec isn't consistent with the LC specs.
<bob> good idea
Chair: SHould we repub?
MarcH: Has it changed that much? Not that much value.
Hugo: We moved the metadata stuff.
Paul: What is the implication of
the TF on this document?
... Might be nice to have a cutoff before the TF impacts kick
in.
... Useful to publish to show where we were at this point in
time.
(nobody thinks it's a bad idea.)
Chair: Only downside is we might make a lot of changes next week and repub shortly after.
Hugo: Or we could publish right
after next week. I have more time to publish after next
week.
... Only problem is namespaces if we publish twice within a
calendar month.
... If we use the same dates as we did for the other LC docs it
makes it easy to pub.
Chair: Any objections? Editors
would have license to deal with namespaces and so forth.
... No objections
RESOLUTION: to publish the WSDL Binding spec as a WD.
Chair: Use RDDL at namespace
URIs
... plans to add this to the issues list.
<scribe> ACTION: Jonathan to draft a sample RDDL doc to further the proposal. [recorded in http://www.w3.org/2005/04/11-ws-addr-minutes.html#action01]
Chair: Determining the value of [destination] from WSDL.
Anish: Last week we discussed how
to determine the value of [destination] by looking at the
WSDL.
... The issue is that there is no mechanism in WSDL.
... Our intro to the WSDL binding says that this specs binds
the abstract properties defined in the core.
... destination is a required property.
... Five proposals:
... 1) use wsdl20:endpoint/@address attribute
... 2) wsaw:UsingAddressing/@destination
... 3) (2) but default to (1)
... 4) wsdl20:endpoint/wsa:EndpointReference
... 5) (4) but default to (1)
Prasad: Are these alternate proposals?
Anish: Yes.
... 3 and 5 represent the possible combinations.
Hugo: You say in (1) "in the absence of additional information". What is that?
Anish: There is no clear indication that the WSDL writer has put in something special. You may be using WSDL doc in conjunction with something else. Didn't want to force you to use this value in that case.
Hugo: So, "in case you have an extension"?
Anish: But in proposal 1 there is no extension.
Chair: Anish also proposed
another issue
... How to specify [reference parameters] in WSDL.
Anish: Similar problem, similar
options as for the previous issue.
... 1) wsdl20:endpoint/wsa:ReferenceParameters
... 2) wsdl20:endpoint/wsa:EndpointReference
... Gives you the power of EPRs in WSDL.
MarcH: Sounds like a winner to me
?: If you have EPR and address, which one would has priority?
Anish: If they are different, use the one in the EPR.
MarcH: A non-WSA aware client would use the @address.
Anish: We also get back to the difference between logical and network address.
DaveH: You want to be able to say this service is located at this endpoint, which may include refparams.
Anish: Yes.
... There is a use case where you may want to include ref
params in your WSDL.
... Second, you may have multiple ways to get to the same
destination, potentially you could use the same destination in
different endpoints.
DaveH: How does this work downlevel?
Anish: You'd use wsdl:required.
[Missing Ugo's point...[
Chair: We'll be addressing this
next week.
... We've talked about [ref params], but our charter says we
need to consider other properties.
... Seems like people thought we've already dealt with the rest
or decided we didn't need to.
... I'll probably raise an issue about this to track
it.
<scribe> ACTION: Chair to raise an issues checking that we've looked at knobs for each property in WSDL. [recorded in http://www.w3.org/2005/04/11-ws-addr-minutes.html#action02]
Chair: Anish has a new proposal.
<mnot> http://www.w3.org/mid/[email protected]
Chair: Discussed this last week.
People thought it was reasonable now that we've dropped the
difference between logical and network addresses.
... Do we need more time to look at this?
Jonathan: Yes.
Prasad: Yes
Chair: We'll try to close this next week.
Chair: Still waiting for Paco's
Action.
... Anything else we can do today?>
Prasad/Jonathan: Related issue on splitting the wsaw namespace between specs.
<scribe> ACTION: Prasad to raise and issue by next week on splitting namespaces across specs and it's effects on versioning strategies. [recorded in http://www.w3.org/2005/04/11-ws-addr-minutes.html#action03]
Chair: Anything we need to deal with from Async?
Glen: Ongoing discussion about
what happens at the various layers.
... Some discussion about how to think about what flows upwards
from transport and downward from application.
... Such as faults.
... We could discuss that but we don't really have clear
questions to bring back to the group at large.
... We're trying to see whether there is something fundamental
we're not closing on or if it's just different viewpoints.
Chair: AOB?
... I'll get revised agenda out in the next day or two. See you
in Sunnyvale!