[
Lists Home |
Date Index |
Thread Index
]
David Megginson wrote:
> So this is more of a proposal than a final specification, then --
> you're making it public to get comments and give the rss-dev people
> something to work with?
Not just for that, but also for general implementation. The "DRAFT"
status is merely a caveat stating that there may be minor undiscovered
bugs in the specification that wider review will uncover. For example,
we've collected three so far and they're all trifling from the point of
view of a feed producer:
1) Cardinality values in OWL should be datatyped with xsd:int, though
the person who raised this bug never does that in his own schemata.
2) Charmod's C018 recommendation is actually worded so that a single
character encoding set should be used, so we'll probably drop down to
specifying that only UTF-8 is allowed.
3) Cardinality and allValuesFrom constraints mightn't be allowed
together on an OWL restriction in OWL Full, though the specification is
unclear about this. We'll be separating these out anyway to ensure that
those particular constraints are usable in OWL DL.
So two OWL concerns and one i18n.
At the moment, the biggest concern that's being levelled (as we rather
anticipated) at us is that RSS version proliferation isn't a good thing;
but since this is intended as a slow bugfix replacement rather than a
completely new addition or fork, that bothers us not.
Cheers,
--
Sean B. Palmer, http://inamidst.com/sbp/
|
