- From: Ed Barkmeyer <edbark@nist.gov>
- Date: Thu, 09 Feb 2006 11:24:29 -0500
- To: Dave Reynolds <der@HPLB.HPL.HP.COM>
- CC: public-rif-wg@w3.org
Dave Reynolds wrote: > I am perfectly happy that phase 2 might well include extensions which > are close to the bleeding edge. However, the task of phase 1 is to > define a simple common, but extensible, core and even when we get to > phase 2 it might be reasonable to prioritize the extensions which are > most widely implemented and needed first. Fully agree. And I apologize for accusing you of a literal "least common denominator" philosophy. I also meant it in the extended sense: a simple, well-established and widely implemented common core. I agree with Piero that we may well be standardizing a "dead language", but I think that is the job of standards committees -- to specify clearly and consistently that which is commonly implemented, and in our case, to provide a way to convey unambiguously concepts that are well understood. If we stray in the direction of concepts that are less well understood generally, or not generally practiced/supported, we increase expressiveness at some cost to interoperability. While OWL was ground-breaking in a certain sense, the principal breakthrough was getting wide agreement on what all was *common* to tractable inferencing systems, and what set of extensions (all of which had been relatively well tested in academic research) was possible for most of those systems. -Ed P.S. In the "because I have the grey hair" tradition: COBOL and FORTRAN succeeded because they met a user/market need and had core standards early in that market that were widely implemented. What made them valuable in the marketplace was that they created relatively transferable programs, and more importantly, transferable programming skills. We think of them as dinosaurs because their computational models and language design predated widespread academic understanding of computational models and language design. And for 12 years, there was *no* other standard language! I don't think the situation in rules languages is at all similar -- we have 20 years experience with the technologies, but no common language, and no transferable skills. OWL and HTML are much better analogies. -- Edward J. Barkmeyer Email: edbark@nist.gov National Institute of Standards & Technology Manufacturing Systems Integration Division 100 Bureau Drive, Stop 8263 Tel: +1 301-975-3528 Gaithersburg, MD 20899-8263 FAX: +1 301-975-4482 "The opinions expressed above do not reflect consensus of NIST, and have not been reviewed by any Government authority."
Received on Thursday, 9 February 2006 16:24:43 UTC