W3C

WS Policy WG Teleconference

30 Aug 2006

See also: IRC log

Attendees

Present
Abbie Barbir, Charlton Barreto, Chris Ferris, Felix, SergeyBeryozkin, JongLee, Paul, Ruchith, JCrump, dmoberg, Fabian, bijan, maryann, prasad, asir, Mark Temple-Raston, Whenry, Ashok Malhotra, Monica Martin, Nadalin, Toufic Boubez, GlenDaniels, Dave Orchard, Umit, Yakov
Regrets
Frederick
Chair
Chris
Scribe
Maryann

Contents


 

 

<trackbot> Tracking ISSUEs and ACTIONs from http://www.w3.org/2005/06/tracker/wspolicy/

<paulc> test

<cferris> pong

<paulc> Chris: Your new agenda at http://lists.w3.org/Archives/Public/public-ws-policy/2006Aug/0168.html

<paulc> does not mention DavidO's email for Action 28 (http://lists.w3.org/Archives/Public/public-ws-policy/2006Aug/0159.html)

Roll call and selection of secretary, Chair

chris, tony nadalin is with me on the chat

<cferris> scribe: maryann

scribe + nadalin

<asir> Scribe: Maryann Hondo

<asir> ScribeNick: maryann

Review and approval of telcon minutes, Chair

agenda item #2

approval of minutes from 23rd

<paulc> Minutes for Aug 16 and Aug 23 are both approved unanimously.

Future WG meetings, Chair

editor update

<paulc> Discount room rate is available until Sept 1.

asir is providing update on editors action items

<paulc> All editorial work except 3605 is done.

<paulc> fsasaki(fsasaki@128.30.52.28)] tracker for the editors is ready now, except trackbot, that will be ready tomorrow. See http://www.w3.org/2005/06/tracker/wspolicyeds/

editors request an update on tracker

<asir> We completed all the editorial actions from last week conference call except issue 3605

<asir> We are still waiting for the terminology extraction XSLT (DaveO is on point

<asir> We are planning to provide a drop (editors' drafts) for the F2F meeting. Our target is Tuesday Sept 5 at Noon

<asir> We spend a lot of time tracking our actions. We'll appreciate an ETA for our tracker tool

asir: there is a link to general tracker but not specific tracker

<fsasaki> I'll make the HP update

asir: 3605 is in progress

toufic: was working on 3605, there were 2 corrections
... only one has been updated
... will complete today

paul: what is eta for next editors draft

<paulc> ETA for editor's draft is Tue Sep 5 for F2F meeting

asir: targetted for 5th

review of open actions

chris: logistics from f2f

dave: working on logistics update

chris: new date?

paul: by the f2f
... new date for update sept 12th

chris: action 48 update?
... action 59 no update
... there is another action about new members which has been fixed
... paul push to next week to complete
... item 63 this belongs on the editors list

Resolution: close 63

chris: action 67

glen: is the issue done?

paul: this is to update the issue

glen: yes

RESOLUTION: glen said its done, to send update to paul

paul: close this action with a pointer to the item

<GlenD> http://www.w3.org/Bugs/Public/show_bug.cgi?id=3577

chris: done with action items

new issues

chris: opportunitiy to bring up new issues
... so that when we reach the F2f we can resolve them

<scribe> topic : issue 3590

daveO: 3590 as an editor, added attribute extensibility
... did a rewrite in bugzilla for all extensibility places
... numbered the extension points 1 to 8
... children are treated as assertions
... 6,7,8 more interesting, started looking at policy reference
... looked at some of the other inclusion mechanisms
... all allow element extensibility and listed types
... we can''t predict, but there are cases where others have chosen to have element extensibility
... add element extensibility to policy reference
... normalization rules would need to be updated if we did attribute extensibility
... not sure what the treatment would be

<bijan> +1

daveO: unless we can come up with a reasonable solution i propose no change on all and exactly one

<bijan> I think think exactlyone and all shouldn't be touched

daveO: propose text be made consistent with the schema

<Ashok> i agree with Bijan ... what are semantics of allowing extensibility on them

daveO: number 4 is the significant item
... slight change to the notation section

chris: questions on dave's issue?
... hopefully we can address this in time for the f2f

paul: or next week
... trying to get an idea of what to put on next week's agenda

daveO: there is a firm proposal
... add eleement extensibility
... asir may have some pushback

asir: agree this is ready for next week
... agree with 1,2, 3
... some pushback on 4

daveO: people use other namespace....

chris: can we respond on the list?

daveO: ok we'll keep emailing

<paulc> David's proposal is in http://www.w3.org/Bugs/Public/show_bug.cgi?id=3590#c2

paul: proposal is in comment 2 in bugzilla

<asir> Response from Asir is at http://lists.w3.org/Archives/Public/public-ws-policy/2006Aug/0156.html

daveO: its also in the email archive

3599

<cferris> http://www.w3.org/Bugs/Public/show_bug.cgi?id=3599

<paulc> Update proposal for 3590 is in http://lists.w3.org/Archives/Public/public-ws-policy/2006Aug/0155.html

ashok: if we agree people want to attach policies to components then we have to specify how wsdl 1.1 is referred to
... this is in the email sent

<cferris> http://lists.w3.org/Archives/Public/public-ws-policy/2006Jul/0019.html

ashok: do we agree that we refer to wsdl 1.1 components using external attachment?

chris: does everyone understand the issue?

<asir> I am not aware of any discussion thread

paul: is there a email thread?

ashok: yes

umit: why are we discussing the issue?

<asir> Response from Asir is at http://lists.w3.org/Archives/Public/public-ws-policy/2006Jul/0085.html

umit: do we need a proposal? maybe ashok should publish it again so people can review
... with proposed changes

chris: ashok, reply?

ashok: does the WG want this written up?

daveO: i'd like to see something that shows capabilities that can't be expressed with what we have now

ashok: there is no way to refer to WSDL 1.1 definition

daveO: not sure why you attach policy to a definition

ashok: component?

daveO: the root element?
... the definition in not a policy subject

asir: 2 points,
... what is the interop issue?
... the uri and fragment identifiers, the domain expression is an XML element

<umit> I was not aware that we are limiting our selves to interop only. WS-Policy is a general framework.

chris: lets take this to the list
... ashok, we need a more concrete proposal

pauL; action for F2F

<scribe> ACTION: ashok to present proposal at F2F [recorded in http://www.w3.org/2006/08/30-ws-policy-minutes.html#action01]

<trackbot> Created ACTION-75 - Present proposal at F2F [on Ashok Malhotra - due 2006-09-06].

chris: action 3602

dan: an assertion whose type is part of vocabulary.........(reading from the bug)

<cferris> http://www.w3.org/Bugs/Public/show_bug.cgi?id=3602

dan: one interpretation is that I'm explicitly prohibited
... outlined an example where that breaks down
... there could be legitimate cases that can be prohibitied by the framework
... we proposed to remove the text

chris: any questions?

<cferris> http://www.w3.org/Bugs/Public/show_bug.cgi?id=3604

chris: item 3604

bijan: haven't heard from any one else on list
... can re-iterate but doesn't like the section

chris: proposal is to nuke goals section?

bijan: yes

<asir> +1 to Bijan

3607

chris: yakov??
... anyone object?

<scribe> ACTION: editors to remove [recorded in http://www.w3.org/2006/08/30-ws-policy-minutes.html#action02]

<asir> ACTION: Editors to implement the resolution for issue 3604 [recorded in http://www.w3.org/2006/08/30-ws-policy-minutes.html#action03]

<trackbot> Created ACTION-76 - Remove [on Editors - due 2006-09-06].

<trackbot> Created ACTION-77 - Implement the resolution for issue 3604 [on Editors - due 2006-09-06].

paul: trying to find the thread in archive

chris: bijan is there a thread?

asir: there is email in the archive

<asir> it is at http://lists.w3.org/Archives/Public/public-ws-policy/2006Aug/0132.html

RESOLUTION: resolved 3604 closed as proposed

<bijan> Head of the thread: http://lists.w3.org/Archives/Public/public-ws-policy/2006Aug/0100.html

3613

chris: frederick?

paul: frederick is on holiday

<paulc> http://lists.w3.org/Archives/Public/public-ws-policy/2006Aug/0141.html

<scribe> ACTION: chris to ping frederick about proposal [recorded in http://www.w3.org/2006/08/30-ws-policy-minutes.html#action04]

<trackbot> Sorry, couldn't find user - chris

oops

3619

<bijan> maryann: updated bugzilla with ref to mail to WSA group

<bijan> maryann: One response to the WSA list, but unsure how to followup and coordinate

<bijan> cferris: do we take it to the CG?

<GlenD> Marc's mail: http://www.w3.org/mid/5B3DE3A6-D12D-4F67-B0CC-6EA968805F33@Sun.COM

<bijan> <some chatter to try to find the message; glend saves the day>

<bijan> <bugzilla woes>

paul: maryann will have to do the change again

<scribe> ACTION: maryann update bug again [recorded in http://www.w3.org/2006/08/30-ws-policy-minutes.html#action05]

<trackbot> Created ACTION-78 - Update bug again [on Maryann Hondo - due 2006-09-06].

asir: why does this have to go to the coordination group?

<scribe> ACTION: paul & chris to bring up to CG [recorded in http://www.w3.org/2006/08/30-ws-policy-minutes.html#action06]

<trackbot> Created ACTION-79 - & chris to bring up to CG [on Paul Cotton - due 2006-09-06].

ACTION 6= paul & chris to bring up 3619

<cferris> ACTION 6= paul & chris to bring up 3619 on next wek's WSCG call

3620

glen: we don't explain how to put policy into EPRs

chris: is this for this group?

glen: yes

chris: glen can you propose language?

glen: sure
... we should have discussion first

<Ashok> I agree it shd be in the spec

paul: this is out of scope
... this is V next

glen: i don't get that from the charter

paul: there is an explicit list in the charter

daveO: notwithstanding the charter discussion, the risk of not doing this is that others may do it in a different way
... another can put it somewhere else, and then you don't have interoperability
... might be able to do something simple
... and might prevent potential interop issues

<GlenD> I do believe a solution to this can actually be pretty simple (esp. in that EPRs :: Endpoint Subjects), and agree with Dave's interop concerns.

daveO: since addressing got done before us it seems we should do somthing here

ashok: i remember that ws-addressing spec lets you attach policies to an epr
... how can they speak about it and that's out of scope for us?

chris: not agreement on out of scope
... we take this discussion to list and discuss in time for F2F
... might involve some technical work and we don't want to wait

umit: in favor of daveO
... lets move on and we need to figure out if metatdata is about epr or target and the use cases need to be clearly explained

asir: interesting work, but think its a major piece of work and is out of scope in the charter

<danroth> http://specs.xmlsoap.org/ws/2004/09/mex/WS-MetadataExchange.pdf

asir: metadata exchange is the correct technical solution

<GlenD> Even if MEX was in-scope for us, I don't think it actually does a good enough job of specing this either.

<GlenD> But that's another discussion :)

daveO: where metadata (policy is one type) is defined, there are several places
... 3 types or places where specification of metadata is done, pros and cons in each
... wsdl- can't go back, schema- can't go back

chris: varying opinions on this issue

<asir> charter says, 'If some function, mechanism or feature is not mentioned here, and it is not mentioned in the Scope of Work section either, then it will be deemed to be out of scope.

chris: suggest we take this to the list

paul: is there a thread?

glen: no thread, bug in bugzilla
... i will start a thread

paul: point thread back to bugzilla

chris: thread on in scope

<scribe> ACTION: paul to start in scope for 3620 [recorded in http://www.w3.org/2006/08/30-ws-policy-minutes.html#action07]

<trackbot> Created ACTION-80 - Start in scope for 3620 [on Paul Cotton - due 2006-09-06].

<scribe> ACTION: glen to provide proposal for 3620 by next call [recorded in http://www.w3.org/2006/08/30-ws-policy-minutes.html#action08]

<trackbot> Created ACTION-81 - Provide proposal for 3620 by next call [on Glen Daniels - due 2006-09-06].

3621

<cferris> http://www.w3.org/Bugs/Public/show_bug.cgi?id=3621

bijan: formal semantics would help resolve some of the issues we are discussing

chris: questions?
... is there a thread?

bijan: will take it to the list

paul: it is an action

chris: dig up thread?

<paulc> http://lists.w3.org/Archives/Public/public-ws-policy/2006Aug/0145.html

<bijan> Thanks paulc

3622

<cferris> http://www.w3.org/Bugs/Public/show_bug.cgi?id=3623

bijan: extensions to wsdl

paul: took action on this
... this is on the agenda of the next call
... i'm going to ask that group to review our working draft

bijan: i thought we were referencing their spec

paul: not in the charter

bijan: thought it was an oversight

paul: what do we need to do?

bijan: attachment allows for inline and out of band for wsdl

chris: we will discuss this at the CG group

paul: we can attach policies to wsdl, they can attach to wsdl, why does this require coordination?

bijan: there is overlap

paul: many flowers bloom
... its not in the liasson section of the charter

<scribe> ACTION: chris and paul to track & report on WG CG for 3623 [recorded in http://www.w3.org/2006/08/30-ws-policy-minutes.html#action09]

<trackbot> Sorry, couldn't find user - chris

message 158

paul: 3639

ashok: we'd like to figure out which of the possible alternatives is being followed

<paulc> http://lists.w3.org/Archives/Public/public-ws-policy/2006Aug/0158.html

ashok: no way to do this, have to check all possible alternative
... should be able to tell from a message

<paulc> and http://www.w3.org/Bugs/Public/show_bug.cgi?id=3639

ashok: we need to work on this

<paulc> 1. An algorithm to select a single alternative if more than one alternative in the two policies matches

paul: both bug and the mail include 3 questions

<paulc> 2. A mechanism to indicate the selected alternative

<paulc> 3. An ability for the message to indicate the policy alternative it is following

<paulc> 3 is the same as 2

paul: are 2 & 3 different?

<paulc> Just 2 questions.

paul: need clarification?
... there is not a concrete proposal

dan: when you says it seems wrong?

ashok: its extra work

dan: do you wan the optimization?

ashok: yes
... its fundamental to how we use policy

bijan: when we say alternative, do you mean the normal form?

ashok: yes

bijan: isn't only one alternative allowed?

ashok: taking two policies and matching to find one complete alternative in common
... you can have one or more match

chris: might have A & B as provider, consumer has A
... consumer needs to indicate A selected
... no formal proposal
... ashok, can you produce a formal proposal?

<bijan> I wonder if just sending a policy that contains the selected alternative would do the job

ashok: is this an item we will pursue?

chris: is there email?

paul: yes

chris: respond to thread
... look for consensus next week

messag 157, issue 3638

ashok: want a policy assertion that adds something to message
... we then want the message logged, and you want timestamp then log
... no way to order in the framework, although WS-Security policy does this for signing/encryption
... some mechanism is often useful

asir: security policy does not order assertion
... it has assertions that order the runtime behavior

daveO: lots of comments.....
... toward the end of the meeting, i've been trying to address one of the editor items and i'm confused about what is to be done

chris: we'll add to the agenda
... ashok?

ashok: didn't understand the point about runtime
... many of the assertions are applied at runtime, asir was trying to make a fine distinction that eludes me

<bijan> I think asir and I were making a similar point; asir on the call, me in the email thread

chris: "l" and "m" are probably major issues, encourage getting this on the list

paul: it would help, people are referring to another spec.

<Ashok> bijan, I guess I didn't understand either of you ...

paul: ashok should show where this inference is

editorial issues

daveO: working on this
... making progress, hopefully have it for the f2f
... if can't get a resolution, will manually insert

chris: back to 3604

paul: bijan suggested replacing the entire section

daveO: except that there's a term defined there

bijan: i don't recall this

daveO: doc has changed since you first wrote this up
... we can remove this section and do the termdef later
... but that means more work

chris: open a new issue

<bijan> I'm looking for the definition

<scribe> ACTION: editors to attempt to make doc consistent and need to deal with "cascading delete" [recorded in http://www.w3.org/2006/08/30-ws-policy-minutes.html#action10]

<bijan> Oh, I see, in the editors draft

<trackbot> Created ACTION-82 - Attempt to make doc consistent and need to deal with \"cascading delete\" [on Editors - due 2006-09-06].

<bijan> Isn't it also defined in section 4?

chris: out of scope issues

<whenry> YEs, closed

paul: bug is closed

<whenry> http://www.w3.org/2005/06/tracker/wspolicy/actions/67

RESOL

<whenry> Woops, sorry

RESOLUTION: 3592 closed with no action\

minor technical issues

3549 - discuss next week

3557

<cferris> http://lists.w3.org/Archives/Public/public-ws-policy/2006Aug/0164.html

<cferris> ashok's email from this morning

ashok: want one paragraph to say something like domain expressions can be used to refer to wsdl 1.1 components, whatever and followup with one non-trivial example
... there seems to be a lot of pushback on this

paul: is there missing text?

ashok: I thought editors could wordsmith

paul: message 164

ashok: just making spec clearer for reader

asir: i had an issue to raise issues on mailing list
... but i don't see answers

paul: you two are talking past each other

ashok: what is the problem with putting in this paragraph?

paul: you are not participating in the dialog

chris: we need to have continued discussion and ask that ashok address the quetions asir raised and asir address ashok's question?

ashok: its a small thing

paul: the chairs are trying to get both sides to answer the questions

dan: i have a question, i'm confused
... we had a section with a domain expression and we removed it
... why are we going back?

ashok: it would be nice to add a quasi-real example

umit: this came from the eprs
... what are other possible examples? that are illlustrative?
... question to asir, what is your concern?
... i'm not following your question
... jms is domain specific, and it can be a policy subject, so is the objection to the jms domain?
... that was the spirit of the example

chris: it could be that the example is contentious not because of the policy example
... use an example other than jms

ashok: that isn't part of the example

chris: asir's email is asking about jms
... it could be that no one wants a jms example in there, might we come up with a less controversial example

ashok: that's not what we're asking for

chris: the idea is to engage in a dialog to address the questions in the previous mail

<GlenD> /me I think it's http://www.w3.org/mid/4DF3D07B9910264B9470DA1F811D1A950B58E898@RED-MSG-43.redmond.corp.microsoft.com

major technical issues

3577

chris: semantics of specific interaction

<cferris> http://www.w3.org/Bugs/Public/show_bug.cgi?id=3577

glen: spec states that the semantics of intersection can be determined by extension specific extensions
... if your processor does not understand, you can't know if the intersection will work and this is a problem for generic tooling
... proposed a number of solutions

<bijan> +1 to option 2

glen: there has been some discussion on list

dan: clarification, interop issues, because of lack of metadata, can you elaborate?

chris: please do this in email

umit: this could eliminate a custom response

glen: yes this would be at a qname level it could affect other things

umit: no opt out in #2

3564

<cferris> http://www.w3.org/Bugs/Public/show_bug.cgi?id=3564

umit: was sick last week, not much progress
... its about chosing alternatives
... we don't have a deadline to produce this, and won't be able to deliver it quickly because going on vacation

paul: worded how?

umit: will write this as part of guidelines for handling optionality
... identify the things to watch out for, and that's the minimum

<scribe> ACTION: umit to draft proposal for primer to address 3577 [recorded in http://www.w3.org/2006/08/30-ws-policy-minutes.html#action11]

<trackbot> Created ACTION-83 - Draft proposal for primer to address 3577 [on Umit Yalcinalp - due 2006-09-06].

paul: need to change due date

<bijan> Hmm. I just noticed we skipped one of my new issues :)

<bijan> J

<bijan> j)

<cferris> which one? I thought we did both

<bijan> there were 3 :()

<bijan> ':)

<cferris> lol

<bijan> j) Policy assertion equivalence and generality, Bijan Parsia

<bijan> http://www.w3.org/Bugs/Public/show_bug.cgi?id=3622

<cferris> sorry

<bijan> S'ok

<bijan> Three ina row is too much :)

Summary of Action Items

[NEW] ACTION: ashok to present proposal at F2F [recorded in http://www.w3.org/2006/08/30-ws-policy-minutes.html#action01]
[NEW] ACTION: chris and paul to track & report on WG CG for 3623 [recorded in http://www.w3.org/2006/08/30-ws-policy-minutes.html#action09]
[NEW] ACTION: chris to ping frederick about proposal [recorded in http://www.w3.org/2006/08/30-ws-policy-minutes.html#action04]
[NEW] ACTION: editors to attempt to make doc consistent and need to deal with "cascading delete" [recorded in http://www.w3.org/2006/08/30-ws-policy-minutes.html#action10]
[NEW] ACTION: Editors to implement the resolution for issue 3604 [recorded in http://www.w3.org/2006/08/30-ws-policy-minutes.html#action03]
[NEW] ACTION: editors to remove [recorded in http://www.w3.org/2006/08/30-ws-policy-minutes.html#action02]
[NEW] ACTION: glen to provide proposal for 3620 by next call [recorded in http://www.w3.org/2006/08/30-ws-policy-minutes.html#action08]
[NEW] ACTION: maryann update bug again [recorded in http://www.w3.org/2006/08/30-ws-policy-minutes.html#action05]
[NEW] ACTION: paul & chris to bring up to CG [recorded in http://www.w3.org/2006/08/30-ws-policy-minutes.html#action06]
[NEW] ACTION: paul to start in scope for 3620 [recorded in http://www.w3.org/2006/08/30-ws-policy-minutes.html#action07]
[NEW] ACTION: umit to draft proposal for primer to address 3577 [recorded in http://www.w3.org/2006/08/30-ws-policy-minutes.html#action11]
 
[End of minutes]

Minutes formatted by David Booth's scribe.perl version 1.127 (CVS log)
$Date: 2006/09/10 21:06:03 $