W3C home > Mailing lists > Public > public-ws-policy@w3.org > August 2006

RE: Action item 9: "information about cascading changes" with URI -> IRI

From: Asir Vedamuthu <asirveda@microsoft.com>
Date: Fri, 11 Aug 2006 16:54:04 -0700
Message-ID: <4DF3D07B9910264B9470DA1F811D1A950B260FAE@RED-MSG-43.redmond.corp.microsoft.com>
To: "Felix Sasaki" <fsasaki@w3.org>, <public-ws-policy@w3.org>

Thank you for carefully enumerating these changes. This e-mail helps the
WG to make an informed decision. For consistency, we suggest one minor
addition to this list: change 'URI of their own' to 'IRI of their own'
in Section 3.1 WS-PolicyAttachment.

Asir S Vedamuthu
Microsoft Corporation

-----Original Message-----
From: public-ws-policy-request@w3.org
[mailto:public-ws-policy-request@w3.org] On Behalf Of Felix Sasaki
Sent: Wednesday, August 09, 2006 9:28 PM
To: public-ws-policy@w3.org
Subject: Action item 9: "information about cascading changes" with URI
-> IRI


Regarding my Action item [1]
"Felix to update the issue
with more information about cascading changes."

and Paul's request from today's call [2]
"(I) want to see where URIs are used and how change would impact"

I took Jonathan Marsh's mail [3] about the IRI adoption in addressing
(which Chris pointed to at [4]) as an input.


- I don't see a need to change "All information items defined by this
specification are identified by the XML namespace URI [XML Namespaces]",
since this is an URI. I don't expect a possibly necessary new XML
namespace URI to contain non-ascii characters.

- sec. 4.2 : I propose to change the reference from RFC 3986 to RFC
3987, here and in the appendix: "associated with a IRI [IETF RFC
3987].", "The identity of the policy expression as an absolute IRI [IETF
RFC 3987]. ", and all occurrences of "URI with "IRI" in that section.
Impact of the change: /wsp:Policy/@Name is defined in terms of
xs:anyURI. As Jonathan stated for addressing in his mail, the data type
IRIs maps well to xs:anyURI, so my change proposal has no effect on the
data type of @Name.
About /wsp:Policy/@wsu:Id , the section says: "to refer to this policy
expression [identified by @wsu:Id], a URI-reference  MAY be formed using
this value per Section 4.2 of WS-Security [WS-Security 2004]." I would
propose to say "IRI-reference". WS-Security relies on WS-Policy, which
itself operates on the information set level, and not on the
serialization of XML. On the info set level, there is no difference
between URI and IRI.

- sec 4.3.4 :
About /wsp:Policy/.../wsp:PolicyReference/@URI : I propose to say "This
attribute references a policy expression by an IRI" and all occurrences
of URI with IRI in that paragraph.
Impact of the change: see my explanation on /wsp:Policy/@Name and
/wsp:Policy/@wsu:Id above.
About /wsp:Policy/.../wsp:PolicyReference/@Digest and
/wsp:Policy/.../wsp:PolicyReference/@DigestAlgorithm : I don't expect
use of IRIs here, so I propose no change.


sec. 3.3 XML Element Attachment :
I propose to write "The wsp:PolicyURIs attribute contains a white
space-separated list of one or more IRIs [IETF RFC 3987].", and change
the reference in the appendix, and change all occurrences of URI to IRI
in that section.
impact of the change: see my explanation on /wsp:Policy/@Name and
/wsp:Policy/@wsu:Id above.

sec. 3.4  :
I propose to change "typically using URIs." to "typically using IRIs.".
sec. 5.2, 5.3 : I propose to change "URI" to "IRI".
These changes have no impact, since they rely on the definition of a
policy expression, which "IRIness" is given by the change in the
framework document.


[1] http://www.w3.org/2005/06/tracker/wspolicy/actions/9
[2] http://www.w3.org/2006/08/09-ws-policy-minutes.html#item07
Received on Friday, 11 August 2006 23:54:10 UTC

This archive was generated by hypermail 2.4.0 : Friday, 17 January 2020 19:33:13 UTC