W3C home > Mailing lists > Public > www-rdf-interest@w3.org > September 2000

off the top issues list

From: McBride, Brian <bwm@hplb.hpl.hp.com>
Date: Thu, 7 Sep 2000 07:27:57 +0100
Message-ID: <5E13A1874524D411A876006008CD059F2393AB@0-mail-1.hpl.hp.com>
To: "RDF Interest (E-mail)" <www-rdf-interest@w3.org>
Off the top of my head, some issues that I can remember.
 
M&S makes conflicting statements about anonymous resources.  In 2.1 it
states "Resources are always named by URI's plus optional anchor id's".  in
2.1.1 it states "The sentence above does not give a name to that resource;
it is anonymous, so in the
diagram below we represent it with an empty oval".  Do anonymous resources
exist in the data model?
 
An API could enable an application to create RDF models which cannot be
represented by the XML serialization syntax.  For example, two web pages
each with a creator property to the same anonymous resource.  Are such
models legal RDF models.
 
Relationship between Properites, namespaces and schemas needs clarification.
Does every Property have an associated namespace and an associated schema?
Does the URI name the namespace or the schema or both?  Can an RDF processor
determine the namespace of a property from the URI?  How does an RDF
processor find the schema for a Property?
 
It is unfortunate that resources of type rdf:alt and rdf:bag are linked to
their members by numbered properties.  Since membership of Alt and Bag is
unordered, would it be more appropriate to have a memberOf property.  The
current scheme requires that implementations  renumber existing members when
a member of a bag or alt is deleted.
 
What is the data model associated with an aboutEach description?  What is
the data model associated with an aboutEach description where the bag is
defined using an aboutEachPrefix?
 
Are parsers required to output bags for all description elements?
 
Request for a standard format for representing triples in a character
stream.
 
Concern over relative naming: use of the ID= attribute e.g. in class
definitions results in class names which are relative to the URI of the
document containing them.  Thus if the document is accessed via a different
URI, e.g. an
HTML redirect, or the text is cut and pasted to another document, then its
meaning changes.  Wow!
 
Request for an algorithm to digitally sign a model.
 
The text of section 6 makes several statements about Description elements
which should also apply to typedNodes.
 
Need for a definition of a schema validation algorithm.
 
The grammer in 6.0 is ambiguous - there are two paths through which
containers can be parsed.  One path does not allow a container to have an
about attribute, the other does.  Which is correct?
 
From our recent email exchanges, request for a transformation grammar (must
look up what that is - but it sounds like the sort of thing I'd like) to
define translation from syntax to model.
 
Hope most of these make sense.
 
Brian
 
 
 
 
 
 
 
Received on Thursday, 7 September 2000 02:28:01 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 7 December 2009 10:51:44 GMT