Re: refactoring RDF/XML Syntax

Hi Shelley,

Shelley Powers wrote:

 > I've chatted with Brian McBride on this issue, and also checked out the
 > clarification on containers using the formal grammar (container vs typed
 > nodes). I have no problem with understanding that containers are redundant
 > and that the same data can be recorded using a typed node.
 >
 > Still, if the concept of a "container" is eliminated, wouldn't this be a
 > change to the specification? Even as a clarification? I realize that the
 > syntax is, itself, backwards compatible -- but a change in understanding is
 > still a change to the specification.


The working group has made the following claricications so far with respect to
containers:

    o it has removed redundancy in the grammar, as you commented above.

    o it has decided that partial descriptions of containers are legal

    o it has removed containers from the formal model section of the RDF model


Further issues related to containers remain to be addressed by the WG.  The WG
also decided to publish the model theory working draft before it had a full
treatment of containers.

Much criticism has beem aimed at containers.  The WG has discussed a strategy of
   removing the current definition of containers from core model of RDF but
retaining them as a vocabulary defined using RDF.  This would enable the WG to
stay within its charter with respect to clarifying the current specification,
whilst at the same time creating space for other container mechanisms, such as
the daml:collection mechanism of DAML+OIL, to be defined on an equal basis.  The
steps taken so far are consistent with that strategy.


 >
 > Will the clarification of the concept of "containers" be included in a new
 > release of the specification? Or as some form of an addendum, or something
 > along these lines?


The WG has not completed considering containers.  There are no plans at present
to drop them from the specification.  Note that other specifications make use of
them (CC/PP), which would make removing them, shall we say, controversial.


 >
 > Do you see a M & S 1.1 in the next year?


Depends on what you mean by M&S 1.1.  The RDFCore WG is behind schedule and the
schedule is currently being updated.  I hope to have an updated schedule within
a month.

Brian

Received on Tuesday, 16 October 2001 08:13:38 UTC