Re: [waldinge@kestrel.edu: Re: Yet another proposal]

   I took a brief look;
     ftp://ftp.cs.yale.edu/pub/mcdermott/papers/noworry.ps.gz
     http://www.cs.yale.edu/homes/dvm/daml/
   it looks pretty good; I'd like
   to spend some time implementing it, but my previous
   commitments don't leave me much spare bandwidth,
   so it doesn't look likely.

We're implementing a parser/generator for it.

   One detail: it seems to assume that RDF sequence
   syntax provides the "and that's all" property,
   when as far as I can tell, they don't.

Does it assume that?  There are probably other subtleties we've
missed.  However, can't we just stipulate that any input RDF document
that violates certain obvious conditions (e.g., no container has two
_1 properties, or a _1 and a _3 but no _2) is ill-formed?  I realize
that nothing prevents someone from the other side of the planet adding
elements to a container I created, but our parser doesn't have to take
such extraneous elements seriously.

                                             -- Drew