RE: Why RDF was a good choice

Hi all,


I think that we are discussing an interesting subject that is very important in 
the life of CC/PP.


Well, I can summarize discussed points in the following: 


1/ RDF use in CC/PP
2/ The used XML serialization of RDF (i.e. the W3C XML serialization)
3/ CC/PP validation, and 
4/ Namespaces use 


Point 1/ and 2/: 
Nowadays, many people use CC/PP and have adopted it for their applications, 
and I think that CC/PP implementers are those who feel more some particular 
limitations of the current CC/PP. This is why, in my opinion, any future CC/PP 
directions must have a large consensus between CC/PP designers and developers. 
'RDF semantic', 'RDF logic', 'if RDF can offer semantic advantages?' are some 
points that represent a complex and deep debate that can take a lot of time. 
In our context, the question is if CC/PP should use RDF or not? and I think 
that a particular attention must be paid to this point if we want avoiding 'bad 
decisions'. 

Mark Butler> Unfortunately the W3C does not seem to be following 
Mark Butler> that rule - it wants to use RDF for everything even though it 
Mark Butler> not finished!

Here, we are two choices:
- Leave RDF, and thus adopting a new model
- Use RDF properly and ameliorate its use (serialization, etc.)
The first choice is hard to do and requires a lot of efforts that will, in my opinion, meet 
certainly some efforts done and doing by the RDF group 
(http://lists.w3.org/Archives/Public/www-rdf-interest/) 




Point 4/: 

Mark Butler> ... adopting a brand new namespace for each vocabulary every 
Mark Butler> time they wanted to add new attributes. When they did this, 
Mark Butler> they copied all the existing attributes to this namespace .. 


I think that adopting a new namespace for a vocabulary depends to the nature 
of this vocabulary, keeping the same namespace is feasible only if the vocabulary 
is incremental. By incremental, I mean that the vocabulary remains consistent 
even if we add new attributes. Otherwise, creating a new namespace is obligatory 
for compatibility reasons and to avoid confusion, for example in the case of types change. 
A simple example that has already made a confusion: The UAProf version 20-Oct-2001 defines the 
"SecuritySupport", in the RDF User Agent Profile Schema (having the namespace:
http://www.wapforum.org/profiles/UAPROF/ccppschema-20010430# ),
as a 'literal bag', while the same attributes is defined as 'Literal' in a previous schema
having the namespace http://www.wapforum.org/UAPROF/ccppschema-20010330#.


A last point: 

I agree with Graham K. that immediate focus should be the logical 
constructs and expressing in CC/PP, and this to ensure a 'clear data model' 
as Mark mentions. In the expressing context, Graham gives an efficient syntax 
for describing media feature sets (RFC 2533: http://www.ietf.org/rfc/rfc2533.txt, 
algebra: http://ftp.ics.uci.edu/pub/ietf/http/draft-ietf-conneg-feature-algebra-03.txt). 
Such syntax and expressing way can be of high benefit in the CC/PP context. 


Regards 

Tayeb* 



----------
Tayeb Lemlouma
http://www.inrialpes.fr/opera/people/Tayeb.Lemlouma/index.html
Opera project
National Research Institute in Computer Science and Control (INRIA Rhône-Alpes, France )
Office B213, phone (+33) 04 76 61 52 81, Fax (+33) 04 76 61 52 07.