>
> h. Qualifiers that violate some aspects of these principles
> but that nevertheless are widely used (arguably providing
> thereby some measure of interoperability) should be
> considered on a case-by-case basis.
>
> [Giant excape clause. Does not seem meet the earlier requirements for
> rigor.]
>
I have to agree with Erik here. We seem to be trying to straddle the fence
between rigor and consensus and its not clear whether that is possible. In
the end, our goal should be to prevent non-interoperable dialects of the
DC15 to evolve. It seems that if a qualifier is non-dumb-downable and does
not refine but extends the semantics of an element that it is simply
non-interoperable and its use will break any search system that relies on
such rules. What we end up with over time if we allow such an escape clause
is a qualifier regime that "generally" only refines semantics of elements,
but "ocassionally" extends the semantics of elements. I don't know how to
write search or indexing systems or inferencing engines that understand that
fuzzy rule (AI experts may have a different opinion).
So, our other choice is to modify "escape clause h" with:
h. Qualifiers that violate some aspects of these principles
but that nevertheless are widely used (arguably providing
thereby some measure of interoperability) should be
considered on a case-by-case basis. In which case the definition of
the
respective DC element will need to be changed to cover the semantics
of the
violating qualifier.
Sorry to make this sound like getting stuck between a rock and a hard place
but this also gets problematic since the element semantics could then grow
over time into even more amorphous blobs then they already are.
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|