RE: Issue 30: Action item 2001/06/20

I concur with this position.  XML Base ought not to, nor did it, 
retroactively change every other XML specification.  That's not my point 
though.

Infoset does make a normative reference to XML Base, so Infoset behavior 
relative XML Base behaviour is defined.

A version of SOAP that normatively refers to Infoset then inherits the 
normative reference to XML Base.  This may be a reason to not do a SOAP 
infoset version, but that's a different issue.  It means that SOAP Infoset 
does have defined behaviour for xml:base, yet SOAP 1.2 non-infoset doesn't 
have behaviour defined.  Which seems extremely bizarre to me.

I think that Chris and I are proposing that SOAP should define behaviour 
for xml:base, rather than leave it undefined.  This solves any ambiguity 
between infoset/non-infoset versions of SOAP, as well as being seemingly 
sensible for the non-infoset version of SOAP.  I'm very curious as to 
reasons not to define behaviour for xml:base in SOAP messages.  Wouldn't 
defining this preclude some interoperability problems, which seems like a 
good thing?

As an aside, it seems to me that this dependency of specifications is yet 
another justification for the TAG.

Cheers,
Dave

On Thursday, August 16, 2001 11:59 AM, Paul Cotton 
[SMTP:pcotton@microsoft.com] wrote:
> When the XML Base specification was completed, it was clearly stated
> that it did NOT create a normative affect on other specifications.  But
> in affect that other specifications had to state explicitly that they
> supported XML Base.  Note the following text from [1]:
>
> "The behavior of xml:base attributes in applications based on
> specifications that do not have direct or indirect normative reference
> to XML Base is undefined."
>
> /paulc
>
> [1] http://www.w3.org/TR/xmlbase/
>
> Paul Cotton, Microsoft Canada
> 17 Eleanor Drive, Nepean, Ontario K2E 6A3
> Tel: (613) 225-5445 Fax: (425) 936-7329
> <mailto:pcotton@microsoft.com>
>
>
>
> > -----Original Message-----
> > From: christopher ferris [mailto:chris.ferris@east.sun.com]
> > Sent: Thursday, August 16, 2001 1:10 PM
> > To: Paul Cotton
> > Cc: W3C XML Protocol IG (E-mail)
> > Subject: Re: Issue 30: Action item 2001/06/20
> >
> >
> > If SOAP is expressed as an XML syntax, then how can it
> > be ignored? Are we saying that XMLBase cannot be used in
> > the context of a SOAP message?
> >
> > Cheers,
> >
> > Chris
> > Paul Cotton wrote:
> > >
> > > We only have to deal with XML Base if we think SOAP 1.2
> > should support
> > > it.  Personally, I do not think this is mandatory for SOAP 1.2
> > > especially since SOAP 1.1 did fine without refering to XML Base.
> > >
> > > I suggest you open a new issue about XML Base support if
> > you think its
> > > support is mandatory.  It is really orthogonal to Issue 30.
> > >
> > > /paulc
> > >
> > > Paul Cotton, Microsoft Canada
> > > 17 Eleanor Drive, Nepean, Ontario K2E 6A3
> > > Tel: (613) 225-5445 Fax: (425) 936-7329
> > > <mailto:pcotton@microsoft.com>
> > >
> > > > -----Original Message-----
> > > > From: christopher ferris [mailto:chris.ferris@east.sun.com]
> > > > Sent: Wednesday, August 15, 2001 3:28 PM
> > > > To: Paul Cotton
> > > > Cc: W3C XML Protocol IG (E-mail)
> > > > Subject: Re: Issue 30: Action item 2001/06/20
> > > >
> > > >
> > > > Paul,
> > > >
> > > > Just a nit.
> > > >
> > > > The href attribute should be of type "anyURI" as defined in
> > > > XML schema datatypes.
> > > >
> > > > It seems to me that we need to address any implications
> > of XML Base
> > > > on the value of the href attribute if it isn't expressed as an
> > > > absolute URI.
> > > >
> > > > Cheers,
> > > >
> > > > Chris
> > > >
> > > > Paul Cotton wrote:
> > > > >
> > > > > Action item 2001/06/20 asked me to clarify Issue 30
> > [1].  This issue
> > > > > originated in my email [2] that outlined how SOAP 1.1
> > meet the XML
> > > > > Protocol R4xxx Requirements:
> > > > >
> > > > > "R403
> > > > > ----
> > > > > Requirement: Data serialized according to the XML Protocol data
> > > > > representation may contain references to data outside the
> > > > serialization.
> > > > > These references must be Uniform
> > > > > Resource Identifiers (URIs).
> > > > >
> > > > > Comment: The SOAP/1.1 encoding uses the "id" and "href"
> > > > attributes to
> > > > > name
> > > > > and refer to resources or sub-parts of resources. The
> > > > format of the href
> > > > > attribute is of type "uri-reference" as defined by XML
> > > > schema. The "id"
> > > > > attribute is of type "ID" as defined by XML/1.0. There are no
> > > > > restrictions
> > > > > on the value of a URI used as value in a href attribute.
> > > > >
> > > > > Judgement: SOAP/1.1 covers this requirement although it is not
> > > > > explicitly
> > > > > stated that URIs can in fact point to anything."
> > > > >
> > > > > Issue 30 Clarification:
> > > > > In my opinion, the only point that we want to clarify (and
> > > > it is only a
> > > > > clarification) is that a consequence of using URIs is
> > that they can
> > > > > point to anything and not only within the same document (of
> > > > the style
> > > > > #foo). Some implementers may be surprised that the
> > value of the href
> > > > > attribute could be something like
> > > > "http://www.foo.com/some.doc" if we do
> > > > > not point this out in a clarification. In addition we
> > might want to
> > > > > indicate that they can point to an attachment to the SOAP
> > > > message [3].
> > > > > In both of the latter cases we want to be sure to indicate
> > > > that these
> > > > > URI's point outside of the current SOAP message.
> > > > >
> > > > > [1] http://www.w3.org/2000/xp/Group/xmlp-issues#x30
> > > > > [2]
> > > > http://lists.w3.org/Archives/Public/xml-dist-app/2001Feb/0045.html
> > > > > [3] http://www.w3.org/TR/SOAP-attachments
> > > > >
> > > > > Paul Cotton, Microsoft Canada
> > > > > 17 Eleanor Drive, Nepean, Ontario K2E 6A3
> > > > > Tel: (613) 225-5445 Fax: (425) 936-7329
> > > > > <mailto:pcotton@microsoft.com>
> > > >
> >

Received on Thursday, 16 August 2001 15:30:19 UTC