WS Policy WG
12 Sep 2007

 

 

Agenda - http://lists.w3.org/Archives/Public/public-ws-policy/2007Sep/0010.html

<monica> good morning

<PaulC> Agenda: http://lists.w3.org/Archives/Public/public-ws-policy/2007Sep/0010.html

<maryann> scribenick: maryann

scribe for next week is Yakov

<fjh> isn't next week call cancelled?

frederick: the meeting next week is marked cancelled on the admin page

<PaulC> Meeting schedule; http://lists.w3.org/Archives/Member/member-ws-policy/2007Jul/0022.html

paul: the intent behind this was based on giving the authors time to work on final edits assuming we were at that point but it doesn't seem we are ready

chris: so we will have the meeting next week and yakov will scribe

approval of minutes from last call

chris: any objections?

none heard

<scribe> ACTION: editors report [recorded in http://www.w3.org/2007/09/12-ws-policy-minutes.html#action01]

<trackbot> Created ACTION-348 - Report [on Editors - due 2007-09-19].

<toufic> thanks

editors report

<fjh> only one action is outstanding - action 353

<fjh> http://www.w3.org/2005/06/tracker/wspolicyeds/actions/353

frederick: the editors need to communicate better about who is doing what

toufic: this action will be done this week

review of action items

AI 333- Phillipe will be back next week, changing target date

AI 339- Maryann sent a proposal and Felix sent editorial items

AI 339 is done

AI 340 - Chris to work with Ken, and open new issues and this is done

Chris to follow up with Maryann and Ken

AI 342 - Editors report back, due date changed to 9/19

AI 343 - Ashok and Maryann to develop proposal ---Ashok spoke with Monica and would like to leave this open

Ashok would like to discuss today

its on the agenda

AI 344 - chris and paul to respond to phillipe, move due date to next week

AI 346 done (actually a dup)

AI 347- close the AI with note from Asir

<cferris> http://lists.w3.org/Archives/Public/public-ws-policy/2007Sep/0006.html

recap of AI

open 333, 342, 343

recommendations issues and errata

<fsasaki> RECs in record time!

chris: commend group and editors on acheiving rec status

<fsasaki> +1 to Chris

paul & chris: thanks and congrats to felix for his ongoing support and contributions

http://lists.w3.org/Archives/Public/public-ws-policy/2007Sep/0000.html

<prasad> Thanks to Chairs for the leadership

asir: thanks to the chairs as well

<fsasaki> +1 to asir too

paul: assume there is an empty errata page

felix: yes

paul: item for editors to discuss, what to do for errata page and make recommendation

<cferris> ACTION: Editors to work on format of eventual errata page [recorded in http://www.w3.org/2007/09/12-ws-policy-minutes.html#action02]

<trackbot> Created ACTION-349 - Work on format of eventual errata page [on Editors - due 2007-09-19].

<scribe> TOPIC : Primer document

<monica> http://lists.w3.org/Archives/Public/public-ws-policy/2007Aug/0021.html

item C- Issue 4943

<cferris> http://lists.w3.org/Archives/Public/public-ws-policy/2007Aug/att-0018/wspolicy-action338-section3.4-primer-context-082207.pdf

response from Asir ...http://lists.w3.org/Archives/Public/public-ws-policy/2007Sep/0005.html

monica: shows 2 companies with different parent assertions for same nested assertions

asir: feed back was editorial only

monica: in the examples, should we take out the description (request from asir) because that was asked for before?

asir: it just makes it easier to read without all the references

<monica> "and different QNames."

<monica> c/different QNames/different QNames and different parent policy assertions.

<monica> c/are different QNames/are different QNames and parent policy assertions.

<monica> that's it

monica: any other questions?

chris: gramatically this doesn't make sense

<cferris> This is because the parent policy assertions EndorsingSupportingTokens and SignedSupportingTokens have different QNames.

<monica> http://lists.w3.org/Archives/Public/public-ws-policy/2007Sep/0005.html

<cferris> so proposal http://lists.w3.org/Archives/Public/public-ws-policy/2007Sep/0005.html as amended by above sentence

<cferris> RESOLUTION: Issue 4943 closed with proposal from Monica as amended by Asir in http://lists.w3.org/Archives/Public/public-ws-policy/2007Sep/0005.html as amended by above sentence from Chris: This is because the parent policy assertions EndorsingSupportingTokens and SignedSupportingTokens have different QNames.

<monica> thank you

<cferris> http://lists.w3.org/Archives/Public/public-ws-policy/2007Sep/0009.html

issue 5036

<fsasaki> Sec. 2.1: change

<fsasaki> "Policy is a simple language that has four elements - Policy, All, ExactlyOne and PolicyReference - and one attribute - wsp:Optional." to

<fsasaki> "Policy is a simple language that has four elements - Policy, All, ExactlyOne and PolicyReference - and two attributes - wsp:Optional and wsp:Ignorable."

<fsasaki> Sec. 2.2: change

<fsasaki> "Child elements of the Policy element are policy assertions." to

<fsasaki> "Child elements of the Policy element that are not from the Policy namespace are policy assertions."

<asir> These two look good

chris: any questions?
... curious about 2.2, policy namespace is extensible, right?
... so if we add foobar, that's supposed to be added as an assertion

asir: there is text on extensibility that dave o put in

paul: that sounds like a last call issue

chris: any other comments on proposal?
... any objections to proposal?

<cferris> RESOLUTION: Issue 5036 closed by adopting felix's proposed edits

felix: there were also edits for the guidelines doc

<cferris> eesagent, where am i?

<fsasaki> related guidelines issue is at http://lists.w3.org/Archives/Public/public-ws-policy/2007Sep/0011.html

<fsasaki> http://www.w3.org/Bugs/Public/show_bug.cgi?id=5041

felix: issue 5041 is for editorial changes to guidelines doc

<cferris> http://www.w3.org/Bugs/Public/show_bug.cgi?id=5041

5041

any objections to the proposal from felix?

<cferris> RESOLUTION: Issue 5041 closed by adopting felix's proposed edits

5038, 5039 need more input, will do on next call

Agenda item 8 -- guidelines document

ashok: we had one call abou this
... we need to speak about how to take this forward, Maryann had thought it might be able to be addressed by a community process, but conversation with Monica raised some new questions
... if i need to specify logging before encryption, there are questions about how to do this, and in talking to my product folks they have said they have a solution to this

paul: why isn't this part of the semantics of the assertion?

ashok: text of the assertion might indicate there could be several patterns of ordering for the assertion

paul: why not specify the rules for the ordering as part of the assertion semantics?

ashok: the text says the order can vary

asir: you could specify the ordering as part of the assertion

ashok: where in the spec does it say this?

asir: it should be in the logging spec, not the policy spec

ashok: but our spec says there is no ordering of assertions

paul: maybe having something in the primer would help

ashok: that may be what you have in mind, but that's not what i heard
... what are we agreeing on?

paul: we don't have to make any changes to framework or attachment document
... and that we could put something in guidelines or primer

ashok: i think we said guidelines

paul: that's the minimum to declare victory

ashok: monica and I were not convinced that we could do this just through text in the guidelines

paul: we need a specific proposal

monica: i said it was difficult to separate ordering from preference
... at this point in time, we might need something that is out of scope for now, because its a more complicated issue

tom: i have sympathy for ashoks use case, but i think the framework says this is out of scope

<PaulC> My apologies to Tom for making him wait on the queue.

tom: the framework says that assertion authors can specify semantics regarding ordering behaviour, and this might not be the best, but this is where we are and this is out of scope for now

chris: we have a concrete example in security policy
... of ordering from an author semantics perspective
... policy doesn't tell you what the SOAP message processing is

<PaulC> The behaviour defines any ordering. The policy assertion just indicates the behaviour to be used.

chris: i thought we talked about using the security policy example

ahok: thats only for one domain

ashok: it doesn't show where you do ordering in a combination, as in reliable messaging and security

chris: in that example the security assertion tells you what to do

ashok: it tells you to encrypt before adding the reliable messaging header?

chris: there is a security considerations section in ther reliable messageing spec

paul: ashok, you and I had an offline discussion and maybe you can help me understand this
... we have assertion a & b and we have a new one c and we need to specify what to do with c
... the semantics for c are relative to a& b, but if we now add d, c does not know about d

ashok: yes

tom: i'm not familiar with security, but RM does not talk about ordering, if you are concerned about ambiguitiy it might be a problem for profiling not policy

ashok: its not an ambiguity its about how the specs are used together
... suppose i figure out how to use them and then there's a new spec, this kind of problem is what's behind the question

chris: i'm not hearing consensus

<asir> http://www.w3.org/TR/2007/WD-ws-policy-guidelines-20070810/#bp-specify-composition

asir: i agree with tom and we have a guideline
... it says you should specify how to compose your assertions with others

<Zakim> asir, you wanted to talk about http://www.w3.org/TR/2007/WD-ws-policy-guidelines-20070810/#bp-specify-composition

asir: you have to be aware

paul: ashok, i'm curious how you think we should process this
... we need to come up with a plan

ashok: if someone could write the example we could try to look at how to address it, i would have written an example, it has turned out to be difficult

paul: are you willing to give up the AI?

ashok: yes

paul: is anyone willing to take the AI and produce an example?
... or provide a reasoning for why its not an issue

asir: i have not encountered an assertion to use to illustrate

ashok: i've given an example of RM and security, and logging and security

asir: i'll take an action to follow up

<monica> We also have v.next issue for preference: http://www.w3.org/Bugs/Public/show_bug.cgi?id=4179

paul: we've heard two examples, the concrete ones above and the more abstract ones i stated above
... we need someone to elaborate on these

chris: asir are you taking this one?

asir: yes

ashok: i have one question related to this..can i ask it?

<cferris> ACTION: Asir to follow up to AI 343 relating to issue 4951 to either provide an example of ordering, or point out where in the specs issues related to Ashok's concerns are addressed [recorded in http://www.w3.org/2007/09/12-ws-policy-minutes.html#action03]

<trackbot> Created ACTION-350 - Follow up to AI 343 relating to issue 4951 to either provide an example of ordering, or point out where in the specs issues related to Ashok\'s concerns are addressed [on Asir Vedamuthu - due 2007-09-19].

ashok: one of the things we thought is that this could be a v next requirement, if it were to be a v next requirement where do we speak of such items? its unclear what the forum is to bring up issues and future requirements?

paul: why not the public working group and mark it v-next?

ashok: just not sure what the process is?

paul: because we're a public group we are an IG ( interest group)
... suggest something like annotating mail with "v-next ordering assertions"

<cferris> http://lists.w3.org/Archives/Public/public-ws-policy/2007Sep/0006.html

<whenry> Do you want to mute the scribe?

<abbie> yes

9a Need volunteers - Guidelines section 5.7.1

chris: we were looking for volunteers and asir has volunteered to do this for the WSDL section

paul: this is really 9b on the agenda

asir: no changes are required to make the guidelines in the WSDL section more specific to WSDL
... should the best practices address differences between wsdl 1.1 and wsdl 2.0

<prasad> +1

asir: determine if terminology is used consistently

<prasad> Asir's mail: http://lists.w3.org/Archives/Public/public-ws-policy/2007Sep/0006.html

<cferris> RESOLUTION: Close issue 5043 with change proposals in Asir's email http://lists.w3.org/Archives/Public/public-ws-policy/2007Sep/0006.html... leave the UDDI concerns to AI 339

<cferris> Status: See email from Maryann:

<cferris> http://lists.w3.org/Archives/Public/public-ws-policy/2007Sep/0007.html

<cferris> and follow-up from Felix:

<cferris> http://lists.w3.org/Archives/Public/public-ws-policy/2007Sep/0008.html

<cferris> scribenick: cferris

mah: discussions with monica, luc clement, toufic (others?)
... some of the discussion may overlap with Asir's proposal we just addressed
... has certain items called out as BP, but formatting needs to be worked on
... reusability of assertions
... how to leverage defined attachment mechanisms
... how to reuse defined policy subjects
... moving some text that existed into this new intro section (now 5.7.1)
... in existing text, there was something that should have been called out as a BP, we should consider calling it out as its own BP

<prasad> They are attachments to the mail Chris

chris: were there any changes to the WSDL section?

mah: yes... called out a BP as discussed above
... added WSDL qualifier

chris: consistent with Asir's changes?

mah: yes

asir: proposal looks good
... don't see a BP statement

mah: what is implied is that words below

asir: we just need two statements

mah: willing to take another pass

paulc: willing to let editors take a pass?

mah: yes

chris: do we have general agreement on the overall proposal?

monica: are we going to discuss felix's changes?

chris: yes, but need general concensus first, do you agree?

monica: yes

chris: everyne in general agreement?

no objections

chris: felix, please take us through your edits

felix:

repitirion of asserions has same meaning

scribe: spelling error in second sentence

<monica> suggest reference to 3.2 and 4.5.

monica: would like to add a specific references

felix: ok by me

RESOLUTION: Issue 5044 closed with maryann's proposal in http://lists.w3.org/Archives/Public/public-ws-policy/2007Sep/0007.html as amended by felix's edits in http://lists.w3.org/Archives/Public/public-ws-policy/2007Sep/0008.html and adding explicit references to sections 3.2 and 4.5 as suggested by monica

<Ashok> I will not be able to attend the call next week