Core terminology?

At 11:06 AM 6/12/01 -0400, Dan Brickley wrote:
>I suspect we tend to talk past ourselves a little, with phrases such as
>"move out of the core", "distinguished status" etc. We have various things
>"in the core" in the sense of being specified in one of the specs owned by
>the RDF Core WG. Beyond that, we have not really defined a more precise
>technical sense of "being in the rdf core".

I think I have a fairly clear idea (in my head) about what it means to be 
"in the core".  If I'm right, I should be able to express it, so here's a 
shot...

I think the RDF core is that which every RDF application must process 
according to some common set of rules.  A feature is "in the core" if it is 
part of the RDF core.

In practical terms, this means those things in the RDF M&S specification 
that have the normative force of MUST.

So this is different from your "definition", at least in the sense that I 
do not count RDF schema as part of the core.

...

I used the term "distinguished status" as a slightly emotive phrase for "in 
the core", suggestive of an idea that does not inherently need to be 
universally understood, but that the RDF specification requires to be so.

...

>Containers are also, as you point out, syntactically privileged in the RDF
>syntax specification. But then, so is the rdf:type construct: we can write
><wn:Person foaf:name="dan"/> instead of a more verbose piece of XML,
>because the RDF syntax provides sugar for common idioms. The container
>machinery in the syntax is in the same category...

I think one could see "syntactically privileged" as a distinct case of "in 
the core" that applies only to the XML-serialized form of RDF, not to the 
abstract graph syntax.  (I have said before that I'd like to see these 
aspects of RDF separated.)

Some cases of "syntactic privilege" require support from the graph syntax 
core:  the 'typedNode' XML syntax needs rdf:type to be a specified property 
in the core.

In my opinion, treating rdf:li as syntactically privileged, per 
Brian/Dave's proposal, requires some  support from the graphical core to 
the extent that rdf:_n are recognized as an ordered family of 
properties.  But I don't think that the definitions of rdf:Bag, rdf:Seq, 
rdf:Alt need to be in the core for this to work.

#g


------------------------------------------------------------
Graham Klyne                    Baltimore Technologies
Strategic Research              Content Security Group
<Graham.Klyne@Baltimore.com>    <http://www.mimesweeper.com>
                                 <http://www.baltimore.com>
------------------------------------------------------------

Received on Wednesday, 13 June 2001 04:35:11 UTC