[All: Please trim the CC list if you feel that's appropriate.]
> Comments appreciated.
I think that it would be useful to let the "SSR language" be used on a
wider scale than RSS 2.0 (for example, in RDF, and perhaps RSS 1.0 as
a module). I use transformations often, and I'd find it useful to be
able to state which XSLT sheet provides which output for which
This would involve slightly--but only slightly--increasing the
complexity; in particular, allowing more output formats than just RDF,
and rewording the document to make it more general. The RSS 2.0
example would only increase in complexity by a single typing
attribute. I envision something like:-
RSS 2.0/arbitrary XML:
(note: only one attribute more is required; doesn't even have to be a
MIME type; you could have a central repository of string to MIME type
mappings if "application/rdf+xml"'s registration state scares you).
same in Notation3, for those who appreciate something readable
[ ssr:type "text/html"; ssr:transform
[ ssr:type "text/plain"; ssr:transform
> ] .
Of course, this proposal runs counter to the KISS principle. With
these sorts of applications, however, it's very difficult to
implement-when-needed (YAGNI?), since that often involves a
model/vocabulary change, and evolution-friendly XML applications
aren't widespread yet...
Idle musing: perhaps the xml:style discussions need to be raised
again. The ?xml-stylesheet PI hack seems to have taken its hold,
Sean B. Palmer, <http://purl.org/net/sbp/
"phenomicity by the bucketful" - http://miscoranda.com/