WS Policy Working Group
4 Oct 2006

<cferris> http://lists.w3.org/Archives/Public/public-ws-policy/2006Oct/0032.html

Chris: version3 of the agenda

Roll call and selection of secretary for the next meeting, Chair

Chris: updated the scribe faq

<cferris> http://lists.w3.org/Archives/Public/public-ws-policy/2006Oct/0032.html

<cferris> http://lists.w3.org/Archives/Public/public-ws-policy/2006Oct/0032.html

<PaulC> Thanks to Systems Team and Felix for helping MSFT get connected.

Review and approval of telcon minutes

RESOLUTION: no objections
... minutes adopted unaminously

Future WG meetings, Chair

Chris: Felix provided the logistics page

Editorial team report

Maryann: we have 10 open issues

Chris: lets defer the decision on primer publication untill we reach item 8
... lets target 'terms definitions'

* sorry for confusing the names...

PaulC: Is David speaking about Glens issue ?

David'suggestion 3 is the closest to what Glen proposed

Mail dated 13th of October

<PaulC> http://www.w3.org/Bugs/Public/show_bug.cgi?id=3720

PaulC: target should be defined in the aprropriate section
... they should not be anchored in section 1

Chris: not all can live with option 1

<PaulC> David's email: http://lists.w3.org/Archives/Public/public-ws-policy/2006Oct/0013.html

Chris: its about a marker, definition is moved to the end...
... is option2 in line with option1 ?

Glen: option3 is the way to go

PaulC: is it possible Dave did not present all the info in his message

<toufic> option 3 has hrefs

<toufic> +1

PaulC: agree with Glen on option 3

<Yakov> +1

Chris: is everyone ok with option3 ?

<abbie> +1

RESOLUTION: instruct David to go with option3

<prasad> May be we should capture in the mintes a clarification regarding including hyper links?

<toufic> the resolution was not captured

Prasad: we need to capture in the minutes a clarification regarding including hyper links

Felix (added after the call while editing the minutes): the usage of the terms in the introduction should be hyperlinked to the actual term definitions in their sections

RESOLUTION: instruct David to go with option3

<cferris> from his email http://lists.w3.org/Archives/Public/public-ws-policy/2006Oct/0013.html

<toufic> JUST KIDDING!

Review action items, Chair

<PaulC> http://www.w3.org/Bugs/Public/show_bug.cgi?id=3706

PaulC: I saw Maryann updated it, lets go to the text

<PaulC> closed by editors....http://lists.w3.org/Archives/Public/public-ws-policy-eds/2006Sep/0098.html

<PaulC> removed unqualified use of the term "domain" in the framework document, there

PaulC: removed unqualified use of the term "domain" in the framework document, there

<PaulC> are now uses of "domain-specific" which is necessary to distinguish between

<PaulC> domain-independent and domain-specific processing of the framework, but "domain

Mariann: there;re several places where domain-only assertions were used

<PaulC> authors" was replaced by "authors" and "domain assertions" by "assertions.

04 01Maryann : should domain-specific or domain-independendt

RESOLUTION: the text http://lists.w3.org/Archives/Public/public-ws-policy-eds/2006Sep/0098.html is agreed upon by the group as resolution for 3706

Maryann: still workin in the first item

Chris: can we finish it next week ?

Maryann: it is possible

<toufic> it must be some localized effects - I'm having no issues with the W3C pages ...

Chris: Glen, can we get a resolution for your item next week ?
... 48 and 96 items are due by Monday

Ashok: Spoke to Felix on Action 106, he recommends using XPath instead of XPointer

PaulC: With XPointer you need to point to a wsdl element

Ashok: Felix says wsdl1.1 component is not defined

Chris: element(...) but it isn't clear to me that that form of xptr is widely supported

<cgi-irc> you are right Chris, it is an issue

Ashok : please give me some more time

PaulC : Xpointer may be the easiest approach

PaulC: Ashok, Felix, can you move the discussion on the list

<cferris> paul suggests that Ashok, PLH and Felix have their dialog on the list

Felix agrees to do it

<fsasaki> ACTION: Felix to start a mail thread on the discussion between Ashok, PLH and Felix on issue 3599 [recorded in http://www.w3.org/2006/10/04-ws-policy-minutes.html#action01]

<trackbot> Created ACTION-121 - Start a mail thread on the discussion between Ashok, PLH and Felix on issue 3599 [on Felix Sasaki - due 2006-10-11].

Maryann: would like more time on Action 108, a week or so

Prasad: Action 110 was aasigned to editors

Prasad: it's about security considerations

PaulC: action 109 seems to be related but it was not

Action 110 is pending

Glen: did Action 113 today

<GlenD> http://www.w3.org/mid/80A43FC052CE3949A327527DCD5D6B2701ECEBFA@MAIL01.bedford.progress.com

PaulC: I did reply as per Action 117, done

Chris: Paul, please send the link

Ashok: Action 119 is closed

<PaulC> Ashok's email re 3730: http://lists.w3.org/Archives/Public/public-ws-policy/2006Sep/0218.html

dan: Action 120, hasn't heard from Paul yet

Chris: it was about indirect references

PaulC: lets close it

Action 96 is deferred

Liaison items

<fsasaki> Response mail from Paul at http://lists.w3.org/Archives/Member/member-ws-policy/2006Sep/0022.html

Chris: Paul, please give the overview of the response on xml:id

<cferris> apologies for not including this link in the agenda

PaulC: I quoted their text and we don't want to remoce xml:id
... because we want be backward compatible
... without xml:id we;d have problems with CR
... third, it's premature to depend on C14N usage note
... there might be time constraints
... recommend to leave the spec as it is
... add a reference to a material describing possible problems with canocalization

<umit> +1 to Paul's proposal

RESOLUTION: group unaminously agrees with PaulC's response on xml:id

<prasad> original mail from Jose Kahan to which Paul will reply: http://lists.w3.org/Archives/Public/public-ws-policy/2006Sep/0130.html

Chris: WS-Addressing group agrees with our response

Open Issues with pending actions

<cferris> Ashok's proposal: http://lists.w3.org/Archives/Public/public-ws-policy/2006Sep/att-0218/Proposal_for_Bug_3730.pdf

Ashok: you can create URI refs to reference WSDL 2.0 components
... we can use these URIs for policy attachments
... would like to have a section 3.4.1
... about attaching policies to wsdl components

Umit: +1 to this proposal
... second page is empty in your proposal

Ashok: no example was intended
... I think the first page gives enough context

PaulC: Ashok refers to a fictitious WSDL

<Frederick> suggestion - use example.com for examples

PaulC: wanted to try a fragment id and it didn't work too
... can we have a well-known WSDL doc we can refer to ?

<cferris> paul suggests that we host a wsdl20 example document at: http://samples.otn.com.LoanFlow.

PaulC: we can use just a browser

Ashok: will try to find one

PaulC: is willing to help here

<cferris> I would actually make two observations... first, sample.otn.com should really be a domain that is intended for examples such as example.com or a w3c domain

<PaulC> Paul would like the example to work if it implies it should work.

<cferris> secondly, I agree that a working example would be a "Good Thing (tm)"

<PaulC> If the example was a WSDL for http://example.com

PaulC: would like the example to work if it implies it should work.

<PaulC> then we know via the RFC that it won't work.

<PaulC> I am also okay with pointing to a "well-known" WSDL that is publically available and using a frag id in the text that actually works.

Paul: I am also okay with pointing to a "well-known" WSDL that is publically available and using a frag id in the text that actually works.

Dan_Roth: Ashok, you need to define an element for your domain expression

<cferris> dan suggests that ashok needs element content for the attachment

Chris: 10 01dan suggests that ashok needs element content for the attachment

<umit> the schema requires a wrapper

Ashok: no probs

Dan: lets use the language of the attachment spec

<cferris> paulc: you need to put the element somewhere... need to update the schema

PaulC: is there a schema for the attachment

<cferris> paulc: suggests that the proposal needs more work

Ashok: I need to have a look

Dan: Schema of the domain expression element is important

Chris: we need to include a schema if it's a normative section, need to be aware of UPA constraints

<cferris> if we add the element before the xs:any, then there will be UPA issues

PaulC: AppliesTo requires an element content

Element name should tell what kind of URI processor to invoke

by referring to the right WSDL version

PaulC: this proposal needs more work
... need to define a proper namespace

Umit: for WSDL 1.1 and WSDL 2.0 we need two diff elements

Ashok: it;s just a suggestion

<Ashok> Can someone put the URL of the WS-Policy Schema on IRC ?

PaulC: at f2f I created an element WSDL Ref and the question was how do you figure out wnat the WSDL version was ?
... elemeny name itself should tell what WSDL ref is in place

Chris: seperate element per WSDL is not the best idea

<PaulC> Chris does not think having the version of WSDL in the element name is bad.

<cferris> actually, I think that having a separate element is unnecessary

<PaulC> Further work on the proposal:

Chris: not very excited about an element per WSDL version

<PaulC> a) better example that works

<PaulC> b) define element name for use in AppliesTo

<PaulC> c) deal with question about whether element name should or should not be strongly typed with version of WSDL

<PaulC> d) update Policy namespace/schemd (http://www.w3.org/2006/07/ws-policy.xsd)

<PaulC> Note on d) the element name might be in another namespace.

Chris: 7.b is pending

<GlenD> http://www.w3.org/mid/80A43FC052CE3949A327527DCD5D6B2701ECEBFA@MAIL01.bedford.progress.com

Chris: we need a link to Glen's note on 7.c

Glen: as far as I remember if we can write down how to do EPR with policy attachment
... it's a critical usecase for us

Chris: one way is to have a small task force

Glen: one way is to have a small task force
... the group should reach the consensus with respect to overriding and ask WS-Addr group to review it

Chris: prefers a small task force

<Maryann> i think thats a good idea

Chris: not confident cross-posting emails will be of benefit

PaulC: prefer not to take any action on this item today

<Maryann> to start a discussion with the wsa group to form a consensus

PaulC: need more time, please

<Fabian> sorry, pressed wrong key

<Fabian> got disconnected, will dial in again

Chris: defer the action until next week

<Fabian> just wanted to mention that I had a question logged on the issue

Fabian: could please someone follow up on the related issue I put into Bugzilla

<monica> http://lists.w3.org/Archives/Public/public-ws-policy/2006Sep/0072.html

<monica> in http://www.w3.org/Bugs/Public/show_bug.cgi?id=3619

<cferris> email link above is fabian's comments to the bugzilla entry

Fabian: don't undertand why external policy attachments are not merged with policies attached to WSDL elements

Dan_Roth: they're different

Chris: lets move on to item 7.d

Levels of indirection for policy references

Dan: my response in email describes how the required level of indirection can be achieved

RESOLUTION: we take no action but we close issue 3719 with the reference to Dan's email. PaulC has an action to reply to PaulD

<PaulC> http://www.w3.org/Bugs/Public/show_bug.cgi?id=3719

<PaulC> rssagent, where am i?

<PaulC> rssagent: where am i?

Chris: moving on to 7.e : Which policy alternative was selected?

Ashok: two different solutions to the problem

<cferris> I would note that while private email exchanges are not disallowed, that technical discussion of issues probably SHOULD be conducted on the public mailing list

Ashok: would it be appropriate to add related text/explantion to the primer ?

Dan: we can invent a new protocol, split polices across endpoints, etc...

Chris: encourages technical discussions on the public list

<umit> +1 to Chris

<GlenD> Almost by definition this ("do it your own way") means non-interoperable solutions. I think the issue is about an interoperable way of solving the problem.

Umit: +1 to moving tech discussion to the public list

<umit> this discussion is about self describing nature of msgs and should be covered in the guidelines document

<Yakov> +1 to Dan in advance. This is implementation specific. External policy engine, etc

<danroth2> http://lists.w3.org/Archives/Public/public-ws-policy/2006Sep/0203.html

Umit: I'd like to suggest there're various ideas there

<fsasaki> ACTION: Daniel to reinitiate a discussion on item 7.e (Which policy alternative was selected?) [recorded in http://www.w3.org/2006/10/04-ws-policy-minutes.html#action04]

<trackbot> Created ACTION-123 - reinitiate a discussion on item 7.e (Which policy alternative was selected?) [on Daniel Roth - due 2006-10-11].

Umit: we need to enumerate known approaches which can help us create self-described messages

Fabian: lets defer this action

Chris: Fabian is to start a new thread and explain why

<Ashok> SB: class of assertions of interest primarily to provider but maybe to the requester

New Issues since last meeting

<cferris> ashok: what do you want the WG to do here>

<Ashok> SB: But knowledgeable requester may be able to use these assertions for processing

<cferris> sergey: wants examples in the guidelines

<Ashok> AM: Sb, what do want WG to do here?

<cferris> sergey: how to properly express such assertions

<whenry> +q

<Ashok> SB: Needs some words about how such asssertions are used -- some clarification and examples

<Ashok> FR: We are onto something bigger here ... class of assertions called obligations in XACML

<Ashok> FR: These assertions are processed differently by the 2 sides

<whenry> Is this not similar to the wsp:local http://www.w3.org/Bugs/Public/show_bug.cgi?id=3721

<umit> I tend to think this is similar...Trying tounderstand the diff

<Ashok> YS: Is this related to wsp:local issue?

<Ashok> SB: I think this is a separate issue although not unrelated

<Ashok> SB: wsp:local is an assertion that can be ignored

<whenry> Actually Askot I think that SB said wsp:local MUST be ignored - which I'm not sure I agree with

<Ashok> FH: When do these assertions get involved in intersection algorithm?

<Yakov> +1 to Chris

<Ashok> DR: Not clear what issue is with optionality ... what needs to be clarified

<Ashok> SB: Explains why wsp:optional is not appropriate to indicate assertions that do not influence wire format

<Ashok> SergeyB can you take over as scribe?

Thanks Ashok

new issues on "relation between primer and guidelines"

<Yakov> I didn't agree on two documents :)

Maryann: section 4 overlaps with the guidelines
... we should focus on the implementers
... we have a primer in CVS
... we submitted another with Umit a doc to Felix

<PaulC> WS-Policy Guidelines for Policy Assertion Authors

Umit: we have WS-Policy Guidelines for Policy Assertion Authors doc but we can come up with a diff name

<Zakim> umit, you wanted to ask a related question about assigning a uri for guidelines document

PaulC: would like to create a doc with Felix which captures all issues

PaulC @ text will point to bugzilla

Umit: would like a stronger statement

PaulC: how about the text saying the final location of the text will depend on the resolution of this issue

Umit: why are we in a hurry to publish a primer doc ?

PaulC: so that we can get feedback sooner

Umit: lets do our homework first

<Maryann> it is possible that if we don't separate the content, that people will be confused not helped by the documents

Umit: we should put more work into preparing a primer upfront

Frederick: agrees with PaulC

<umit> no one has addressed my question

<umit> what is the hurry

Chris: lets defer a decision for a week

<danroth2> The faster we publish the faster we get feedback

Felix: I need to check it all into CVS

<cferris> ACTION: Felix Paul and Chris to draft status section and proposed ednotes for Primer referencing pending work on guidelines and possible issue resolutions resulting in content being moved. etc [recorded in http://www.w3.org/2006/10/04-ws-policy-minutes.html#action06]

<trackbot> Created ACTION-124 - Paul and Chris to draft status section and proposed ednotes for Primer referencing pending work on guidelines and possible issue resolutions resulting in content being moved. etc [on Felix Sasaki - due 2006-10-11].

<cferris> ACTION: Felix to input Guidelines doc in CVS [recorded in http://www.w3.org/2006/10/04-ws-policy-minutes.html#action07]

<trackbot> Created ACTION-125 - Input Guidelines doc in CVS [on Felix Sasaki - due 2006-10-11].

Chris: thanks to everyone !
... optionality tar ball has a lot of diff issues interleaved
... recommends think more on what all people are saying, try to imagine how it all works