Re: xmlsch issues discussion

Hi Dave,

Thanks for the speedy analysis.

As a general point,  I hope that the WG will respond warmly to the comments 
we have received from the xml schema folks.  To my mind, there has in the 
past been a gulf in our respective points of view, and I believe that with 
a little care we may be able to build some bridges across that gap.

At 11:49 12/03/2003 +0000, Dave Beckett wrote:


>The chairs will have to rule on whether my claims below that some of
>these comments are out of scope/charter are on target.
>
>
>ISSUE http://www.w3.org/2001/sw/RDFCore/20030123-issues/#xmlsch-08
>
> >    We believe there are probably good reasons for using an rdf:datatype
> >    attribute, instead of re-using the existing xsi:type attribute which
> >    has (when the type is defined in a schema defined by XML Schema 1.0)
> >    the same semantics.In particular, rdf:datatype does not assume or
> >    assert the existence of the type named as a type in a schema defined
> >    by XML Schema, so it would be problematic to use xsi:type.
>
>Just commentary, no discussion or changes required.  We also agree
>there are good reasons that xsi:type as designed is not appropriate
>for RDF/XML which uses with URIs for datatypes, not qnames.  Maybe
>the XML schema WG wants to hear more reasons why xsi:type wasn't
>used?
>
> >    We do fear, however, that users are likely to find this
> >    near-duplication of the meaning and function of xsi:type confusing. It
> >    is not clear to us what, if anything, can or should be done to
> >    minimize this danger.
>
>The Primer is the place that users can already look for to see how to
>write RDF datatypes in the RDF/XML syntax:
>   http://www.w3.org/TR/2003/WD-rdf-primer-20030123/#typedliterals
>   http://www.w3.org/TR/2003/WD-rdf-primer-20030123/#example7
>   http://www.w3.org/TR/2003/WD-rdf-primer-20030123/#example22

So my reading of this is that they wondered why didn't just use xsi:type, 
and figured out part of the answer.  I suspect there is a spread of opinion 
within that group and we can help by giving a clear explanation of why we 
didn't use xsi:type.

So we can say that we didn't introduce rdf:datatype capriciously, but because:

   - we use URIREF's not qnames as identifiers
   - to allow qnames in just one attribute would have caused confusion and 
we felt that introducing
     the use of qnames throughout the syntax was beyond our charter
   - and, as they point out, we allow datatypes not defined by xml schema
   - we recognise the potential for confusion, we have explained as best we 
can in the primer, but if they have other editorial suggestions, we'd be 
happy to hear them.


>There is already discussion near figure 9
>   http://www.w3.org/TR/2003/WD-rdf-primer-20030123/#figure9
>
>Summary: closed. explanation, no action
>
>
>
>ISSUE http://www.w3.org/2001/sw/RDFCore/20030123-issues/#xmlsch-09
>
> > "We were unable, on a first reading, to determine whether the default
> >    namespace declaration, and thus unprefixed names, were or were not
> >    allowed in order to encode 'RDF URI References'."

Personaly, I'm not sure what is being said here.


>That XML detail is discussed in "The Syntax Data Model Section"
>where it already states in the fifth paragraph:
>
>   "The syntax does not support non-well-formed XML documents, nor
>   documents that otherwise do not have an XML Information Set; for
>   example, that do not conform to XML Namespaces [XML-NS]."
>   -- 
> http://www.w3.org/TR/2003/WD-rdf-syntax-grammar-20030123/#section-Data-Model
>
>So if there is no namespace name, it isn't RDF/XML.

I think what they are getting at is whether or not QNAMES are required, for 
example

<RDF xmlns="http://www.w3.org/1999/02/22-rdf-syntax-ns#">
   <Description>
     <value>10</value>
   </Description>
</RDF>

is legal RDF but there are no qnames in it.  The text they quote:

   http://www.w3.org/TR/rdf-syntax-grammar/#section-Syntax-intro

[[In order to encode the graph in XML, the nodes and arcs have to be 
represented by XML element names, attribute names, element content and 
attribute content. RDF/XML uses XML QNames to represent RDF URI References. 
The namespace prefix part of all QNames is associated with a URI Reference 
as defined in XML Namespaces [XML-NS]. The RDF URI Reference represented by 
a QName is determined by appending the local name part of the QName to the 
URI Reference associated with the namespace prefix part of the QName. This 
is used to shorten the RDF URI References of all property arcs labels and 
some nodes. RDF URI References identifying subject and object nodes can 
also be stored as XML attribute values or XML element names via QNames. RDF 
Literals (which are only object nodes) become either XML element text 
content or XML attribute values.]]

left them unsure at first reading whether the above example was legal.  At 
least that is my guess at what is being said, but it might be good to 
confirm that.

If that is the concern, then a minor editorial change to the quoted 
paragraph would clear it up.


>Summary: closed. explanation, no action
>
>
>ISSUE http://www.w3.org/2001/sw/RDFCore/20030123-issues/#xmlsch-10
>"cannonical syntax"
>
>Discussion of out of what I believe is out of RDF Core's charter,
>creating a new XML syntax or modifying RDF/XML (created 1998)
>to work with XML technology designed later.
>
>   "The RDF Core WG is neither chartered to develop a new RDF syntax,"
>   -- http://www.w3.org/2001/sw/RDFCoreWGCharter
>
>Summary: closed.  we don't have to do this, out of charter ??

I suggest we can agree that it is hard to process RDF with general XML 
tools such as XSLT for the reasons they point out but that RDFCore decided 
that this issue was outside our charter.  None the less we recognise that 
there is some disquiet about the current RDF/XML syntax, we have tabled an 
issue for consideration by a future WG and would hope that WG would welcome 
involvement of the xml schema wg.  We could also note that the XHTML WG are 
already considering a DTD friendly syntax for inclusion in XHTML documents 
with the support and encouragement of the RDFCore WG.




>ISSUE http://www.w3.org/2001/sw/RDFCore/20030123-issues/#xmlsch-11
>"layering on xml"
>
>There is some explanation to do since they seem to have missed
>reading the Syntax Data model section in defining RDF/XML on top of
>the pile of XML technologies.
>
> >    Not only does this confuse levels and thus readers, it also runs the
> >    risk of inadvertently defining an XML subset. It also appears, on a
> >    strict reading, to rule out XML documents not derived from the parsing
> >    of character streams as possible RDF/XML (so that it would be
> >    illegitimate to regard a data structure created using a DOM interface,
> >    for example, as RDF/XML).
>
>It neither creates a subset nor rules out using other XML APIs as
>long as an XML Infoset is available.
>
>   "This model illustrates one way to create a representation of an
>   RDF Graph from an RDF/XML document. It does not mandate any
>   implementation method - any other method that results in a
>   representation of the same RDF Graph may be used.
>
>   In particular:
>   ...
>    * This specification does not require the use of [XPATH] or [SAX2]
>    * This specification places no constraints on the order in which
>   software transforming RDF/XML into a representation of a graph,
>   constructs the representation of the graph.
>   "
>   -- 
> http://www.w3.org/TR/2003/WD-rdf-syntax-grammar-20030123/#section-Data-Model
>
>Action: closed. explanation, no changes

I found myself wondering more about this comment.  The suggestion that the 
syntax is defined at the character level seems just wrong.  It is also 
noteworthy that the approach they suggest of considering a transform into a 
canonical form is not dissimilar to the snail approach we considered, and 
rejected.

This comment belongs in the category of "we would have done it 
differently", which may be true, but there you go.  I think the response 
might also include:

   - agreement that layering of specs is important
   - emphasis that unlike M&S the grammar is not character level, but is 
built on top of an infoset
   - acceptance that are other ways to do it, some of which we considered 
(snail) and why we decided on this approach (the grammar is clearer)
>
>
>ISSUE http://www.w3.org/2001/sw/RDFCore/20030123-issues/#xmlsch-12
>"capricious syntax"
>
>Discussion and proposal for a better XML syntax (RDF/XML 1998) to
>better match the RDF concepts and abstract syntax (2002).  Out of
>charter:
>
>   "The RDF Core WG is neither chartered to develop a new RDF syntax,"
>   -- http://www.w3.org/2001/sw/RDFCoreWGCharter
>
>Comments on the problems W3C XML schema (designed 1999/2000), XSLT
>and XQuery has dealing with RDF/XML and its many variations.
>
>Long comments on difference between XML and RDF data models meaning
>different tool sets.  Tree (sort of) and graph.  We knew that.

Right, and we should say we recognize and share these concerns.


>Proposal to work together to create new XML serializations.

How much of their arm would they like us to bite off?


>Action: closed
>   out of charter/scope??,
>   coordination group work??
>   joint work / postponed work / feedback to new activities??

Right - postponed work.

Brian

Received on Wednesday, 12 March 2003 10:57:27 UTC