Re: A certain difficulty

I have only been half paying attention to this discussion, but wanted
to put forward another example of a group that considered using RDF
and decided not to and why. The W3C P3P group
(see http://www.w3.org/P3P/) had originally planned to use RDF
to encode website privacy policies. Two years into the project at
W3C, the group decided to abandon RDF and just use XML
(although we tried to maintain a syntax that could be modeled easily
with RDF and thus could be translated to RDF if somebody wanted to).
We had lengthy discussions about this, but basically what it came
down to was a feeling that an RDF encoding would be useful to
us IFF P3P developers were likely to already have RDF parsers 
available to them in the tools in which they were embedding P3P
user agents. If RDF parsers were not available, implementers would
either have to implement them first (making implementation more
difficult), or write a special-purpose parser that pretty much ignores
all the advantages of using RDF. Given that not a single one of
our perspective implementers had any plans to use RDF (other
than the W3C itself), we could not in good conscience put this
extra burden on implementers. We do, however, include an
RDF data model diagram in the P3P spec (see 
http://www.w3.org/TR/2000/WD-P3P-20000211/#Appendix_RDF )
which gives some idea of how our syntax maps onto RDF.

Also note that during the two years we were planning on using
RDF we went through many syntax rewrites, in part due to the
changes made in RDF over that time and in part due to our
changing understanding (or lack of understanding) of the RDF
spec. I believe we finally understood it in the end, and we did finally
figure out how to use it correctly. (W3C members see
http://www.w3.org/P3P/Group/Specification/rdf.html ). But by that 
point we decided it was not worth the effort.

Lorrie Cranor
P3P Specification Working Group Chair

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Lorrie Faith Cranor <lorrie@research.att.com>
AT&T Labs-Research, Shannon Laboratory 
180 Park Ave. Room A241, Florham Park, NJ 07932 
Phone: 973-360-8607  FAX: 973-360-8970
http://www.research.att.com/~lorrie/

Received on Monday, 28 February 2000 16:47:13 UTC