OASIS Mailing List ArchivesView the OASIS mailing list archive below
or browse/search using MarkMail.

 


Help: OASIS Mailing Lists Help | MarkMail Help

ws-rx message

[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]


Subject: RE: [ws-rx] CD8


A couple of comments.

1. Precendent on referencing WS-Addressing shows that people think CR is far enough a standards route
2. I think changing from a specific version to a general model is a substantive change and would require another public review.


Martin.

>-----Original Message-----
>From: Pete Wenzel [mailto:pete.wenzel@Sun.COM] 
>Sent: Thursday, March 15, 2007 5:32 PM
>To: Tom Rutt
>Cc: ws-rx@lists.oasis-open.org
>Subject: Re: [ws-rx] CD8
>
>
>Dug's CD08.1 proposal accomplishes the same.  I agree that's 
>the best course of action at this point, and in the spirit of 
>our charter.
>
>--Pere
>
>Thus spoke Tom Rutt (tom@coastin.com) on Thu, Mar 15, 2007 at 
>01:02:35PM -0400:
>> Paul Fremantle wrote:
>> >Dug
>> >
>> >Thanks very much for all your hard work on this.
>> >
>> >Anish - what are your thoughts on the CD8.01 proposal 
>outlined below?
>> This is doing what our charter says "If a referenced standard is not 
>> far
>> enough along on
>> standards track .... abstract its use"
>> 
>> With this, all ws policy references should be put in non normative
>> references section.
>> 
>> Also, I suggest useingthe wording
>> 
>> "wsp" is an abstraction pertaining to any ws policy 
>namespace which is
>> compatible
>> with the policy assertion types defined in this specification.
>> 
>> With this abstraction approach, we do not need to wait for ws policy 
>> to
>> complete.
>> 
>> Tom Rutt
>> >
>> >Paul
>> >
>> >Doug Davis wrote:
>> >>
>> >>All,
>> >>A new folder called "CD08" has been created in the TC's document
>> >>section [1].  In there you'll find 3 pdf files that shows 
>the latest 
>> >>version of the specs with the WS-Policy namespace URI fixed:
>> >>   WS-RM: 
>> 
>>>http://www.oasis-open.org/apps/org/workgroup/ws-rx/download.p
hp/22965/wsrm-1.1-spec-cd-07.pdf 
>> >>
>> >>   WS-RMP:
>> 
>>>http://www.oasis-open.org/apps/org/workgroup/ws-rx/download.p
hp/22966/wsrmp-1.1-spec-cd-08.pdf 
>> >>
>> >>   WS-MC:
>> 
>>>http://www.oasis-open.org/apps/org/workgroup/ws-rx/download.p
hp/22967/wsmc-1.0-spec-cd-04.pdf 
>> >>
>> >>
>> >>and there's also a zip file:
>> >>   ZIP:
>> >> 
>http://www.oasis-open.org/apps/org/workgroup/ws-rx/download.php
/22964/WSRX%20CD8.zip 
>> >>
>> >>that contains everything, including 2 *-diff.pdf files showing what
>> >>was changed from the old CD (for just MC and RMP - RM did not 
>> >>change). The only changes you should see in there are:
>> >>1 - increase in the CD version # (on title page and footers)
>> >>2 - update to the "previous version" names/URL - needed since we 
>> >>bumped up the CD version #
>> >>3 - fix to the WS-Policy namespace URI
>> >>
>> >>I encourage everyone to look these over to ensure that not only is
>> >>the URI correct but also that no other unintended changes 
>were made.
>> >>
>> 
>>>--------------------------------------------------------------------
>> >>----------------------------------------------
>> >>
>> >>
>> >>Odds are the issue of pointing to the CR version of WS-Policy will
>> >>come up on today's conf call, so to see if we can move 
>things along 
>> >>there's also a new "CD08.1" folder. In there, again, you'll find 3 
>> >>pdf files:
>> >>   WS-RM: 
>> 
>>>http://www.oasis-open.org/apps/org/workgroup/ws-rx/download.p
hp/22969/wsrm-1.1-spec-cd-07.pdf 
>> >>
>> >>   WS-RMP:
>> 
>>>http://www.oasis-open.org/apps/org/workgroup/ws-rx/download.p
hp/22970/wsrmp-1.1-spec-cd-08.pdf 
>> >>
>> >>   WS-MC:
>> 
>>>http://www.oasis-open.org/apps/org/workgroup/ws-rx/download.p
hp/22971/wsmc-1.0-spec-cd-04.pdf 
>> >>
>> >>and a zip file:
>> >>   ZIP:
>> 
>>>http://www.oasis-open.org/apps/org/workgroup/ws-rx/download.p
hp/22968/WSRX%20CD8.1.zip 
>> >>
>> >>The files in this folder contain all of the changes 
>mentioned above.
>> >> However, WS-RMP and WS-MC also contain the following additional 
>> >>changes (which you'll see in the *-diff pdfs):
>> >>1 - removed "wsp" from the WS-RMP namespace table (not 
>change needed 
>> >>for WS-MC)
>> >>2 - added the following sentence to the "Namespace" 
>section of WS-RMP 
>> >>and WS-MC:
>> >>The assertions defined within this specification have been 
>designed 
>> >>to work independently of a specific version of WS-Policy and 
>> >>WS-Policy Attachment. Within this specification the use of the 
>> >>namespace prefix "wsp" refers generically to the WS-Policy 
>namespace, 
>> >>not a specific version.
>> >>3 - In WS-RMP, moved the references for WS-Policy and 
>> >>WS-PolicyAttachments to be non-normative (no change needed 
>for WS-MC)
>> >>
>> >>Note that the WS-Policy/Attachments references themselves still 
>> >>point
>> >>to the CR versions of the specs.
>> >>
>> >>It seems we need to balance our desire to reference the latest
>> >>version of WS-Policy with our immediate need to also be able to 
>> >>support existing customers. Maybe the above set of changes could 
>> >>allow both camps to squint a little and be able to move on.  With 
>> >>these changes our specs clearly push people towards the CR 
>version of 
>> >>WS-Policy but still acknowledges that these assertions 
>we're defining 
>> >>are really independent of any version of WS-Policy.  *Note 
>that none 
>> >>of the schema files actually reference WS-Policy at all*.  
>So, given 
>> >>that all of these WS-* specs are supposed to be 
>composible, it seems 
>> >>reasonable to create our assertions (and specs) in such a way that 
>> >>they not only work with the CR version but with any subsequence 
>> >>versions of WS-Policy that the W3C may create. If the WS-Policy WG 
>> >>were to rev the WS-Policy specs it would be a shame to 
>have to update 
>> >>WS-RMP and WS-MC just to update our WS-Policy 
>namespaces/references.  
>> >>(Of course, if WS-Policy were to change in such a way that our 
>> >>assertions themselves needed to be changed to fit into some new 
>> >>framework then we'd have to reopen our specs.)   Anyway, 
>please look 
>> >>over this proposal and see what you think....
>
>-- 
>Pete Wenzel <pete.wenzel@sun.com>
>Open ESB Community <http://open-esb.org/>
>



[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]