W3C home > Mailing lists > Public > xml-dist-app@w3.org > October 2001

Re: FW: Proposed positive text for XML Base

From: Christopher Ferris <chris.ferris@sun.com>
Date: Thu, 18 Oct 2001 12:21:49 -0400
Message-ID: <3BCF019D.1000502@sun.com>
To: xml-dist-app@w3.org
CC: Jonathan Marsh <jmarsh@microsoft.com>
Hugo,

I agree with the bit about the encodingStyle attribute,
but it doesn't end there. We can place a constraint on
the encodingStyle attribute restricting it to be absolute
but requiring a scheme (at least I believe that we can do this
in the schema through restriction). I don't think that
encodingStyle URI value is meant to be used as a URI Reference.

The actor attribute is quite different. I believe that the
actor attribute MAY be used as a URI Reference. From section 2.2
in the SOAP1.2 Part 1 spec:

      "For example, SOAP Actors MAY be named with a URI useable
      to route SOAP messages to an appropriate SOAP node."

This implies that the resource identified by an actor
attribute URI is resolvable, hence it is a URI Reference.
I see no reason to restrict this to be an absolute URI
as the resource that is the SOAP node to which the SOAP
message is to be routed might be relative to the base URI
of the SOAP message.

In addition, the href attribute used in the encoding of multi-ref
items is also a URI Reference because its intended purpose is
to provide a URI reference that is resolvable. It
may also (need to) be a (relative or absolute) URI reference to
some resource external to the SOAP envelope document if I understand
correctly the recent discussion surrounding resolution of
issue #30.

I think that it would be imprudent to restrict the href value-space
to absolute URIs.

Cheers,

Chris

Hugo Haas wrote:

 > * Jonathan Marsh <jmarsh@microsoft.com> [2001-10-17 16:22-0700]
 >
 >>Henrik and Paul consulted with me on this issue, in my role as
 >>editor of the XML Base Recommendation, resulting in an updated
 >>proposal based on David Orchard's text.  Although I am a member of
 >>the XML Core WG, and believe the text below is consistent with the
 >>feedback the Core WG provided on this issue, I do not supply this
 >>text under the aegis of the Core WG.  I hope that this proposal is
 >>helpful to the XP WG in considering this issue.
 >>
 >>
 >>BASE URI's and Relative URI Resolution
 >>--------------------------------------
 >>This version of the SOAP specification supports the W3C XML Base
 >>Recommendation. An xml:base attribute information item MAY appear on
 >>any of the element information items defined in this specification.
 >>This specification does not make use of URI References, only
 >>absolute URIs, so xml:base attributes will not affect the behavior
 >>of the SOAP envelope itself.  Relative URI values (such as for the
 >>SOAP-ENV:EncodingStyle attribute) are not recognized as matching the
 >>absolute URI values mandated by this specification.
 >>
 > [..]
 >
 >>Rationale:
 >>----------
 >>The simplest reading of SOAP 1.2 does not allow relative URIs
 >>(specifically URI References) in any of its attributes.  Making such
 >>a change requires positive motivation.
 >>
 >
 > "Absolute" and "relative" are two words that do not appear in the SOAP
 > Version 1.2 Part 1: Messaging Framework Working Draft[1].
 >
 > The encodingStyle attribute[4] is of type anyURI[3] as defined by XML
 > Schema Part 2: Datatypes[2], i.e. a URI as defined by RFC2396 as
 > amended by RFC2732.
 >
 > The simplest reading tells me that it can by any URI, absolute or
 > relative. I understand that, in SOAP Version 1.2's current form, use
 > of relative URIs leads to undefined results since there is no clear
 > base URI, but it does _not_ disallow them.
 >
 > Therefore, the change here is to explicitely disallow relative URIs.
 >
 > Regards,
 >
 > Hugo
 >
 >   1. http://www.w3.org/TR/2001/WD-soap12-part1-20011002/
 >   2. http://www.w3.org/TR/2001/REC-xmlschema-2-20010502/
 >   3. http://www.w3.org/TR/2001/REC-xmlschema-2-20010502/#anyURI
 >   4. http://www.w3.org/TR/2001/WD-soap12-part1-20011002/#soapencattr
 >
Received on Thursday, 18 October 2001 12:25:27 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 7 December 2009 10:59:04 GMT