- From: Sandro Hawke <sandro@w3.org>
- Date: Mon, 16 Feb 2009 20:53:38 -0500
- To: Jim Hendler <hendler@cs.rpi.edu>
- cc: Christine Golbreich <cgolbrei@gmail.com>, Ivan Herman <ivan@w3.org>, public-owl-wg@w3.org
Jim Hendler writes:
> Frankly, in practice I'm finding it extremely hard to explain why
> three profiles are needed, and I would again propose that we consider
> moving the other documents to CR, but hold back the profiles document
> for further work on explanations and to avoid confusion that could
> lead to less adoption of DL/Full, which still seems to me to be where
> the most important OWL 2 extensions currently live.
FWIW, I think it's easier (and more useful) to explain the difference
between QL and RL [1] than the difference between CR and Rec [2]. It
seems to me like this "hold back" strategy would leave people (rightly)
confused about W3C process instead of trying to understand the
differences between the profiles.
-- Sandro
[1] They are both ways to make queries of a database and have some OWL
inferencing done to give you additional query results. With QL, the
data is left as-is, and the queries are re-written to also return
OWL inferences. With RL, the *queries* are left as-is, but a
process is run to add more data (the OWL inferences) to the
database. The choice between the two depends on which inferences
you care about (some can only be implemented with one approach, some
with the other), and the resource/performance demands of your
application. [This is off the top of my head, trying to be clear
and simple.]
[After writing that, I (as a user) want a clear and simple table of
which OWL features are in each, making it easy to see which are in
both. That doesn't have to come from OWL-WG.]
[2] A "Candidate Recommendation" (CR) is a W3C specification that is
mature enough that people should try to implement it, but the W3C
has not yet determined whether it has been sufficiently implemented
to demonstrate that it *can* be implemented. [This is off the top of
my head, trying to be simple and clear. Since I expect QL and RL to
be easy to implement, I don't really see how they could legitimately
be stuck in CR.]
Received on Tuesday, 17 February 2009 01:53:47 UTC