- From: Danny Ayers <danny.ayers@gmail.com>
- Date: Fri, 31 Mar 2006 14:33:10 +0200
- To: "tim.glover@bt.com" <tim.glover@bt.com>
- Cc: sowa@bestweb.net, semantic-web@w3.org
On 3/30/06, tim.glover@bt.com <tim.glover@bt.com> wrote: > I agree that language design requires taste and judgment, and that RDF > and OWL conspicuously lack both. I beg to differ, but these are subjective qualities, not easily quantified. More generally, the obsession with > using XML and URIs for everything, including programming language > syntax, has been a disaster for the Semantic Web, and the Web in > general, and must have cost billions. URIs are the identifiers of the Web, without them it couldn't function. They also enable global disambiguation of names, rather a prerequisite for a Semantic Web. Whether something other than XML might have been better for many of the places on the Web in which it's used is debatable, but having a common syntax has shown significant benefits on and off the Web. It's support for Unicode is a huge plus. > It seems to me that in order to make progress, we have to look backwards > for a while. When the SW community is familiar with PROLOG perhaps they > will be ready to move forwards, to pragmatics :) You don't need to look back to Usenet to identify a troll ;-) [John] > Apple's OS X is built on top of a version of Unix, which is > much more flexible and modular than Windows. Therefore, they > have already implemented most of the features that are planned > for Vista at a fraction of the cost in time, money, and human > effort. HTTP+(X)HTML+URIs, are effectively the OS of the Web. For basic doc/hypertext work on the Web these are more than adequate. Semantic Web technologies build on these. When I'm working with logic-thin syndication data I generally use these plus RDF plus inverse functional properties (from OWL, to allow by-description person identification). I've mostly been using the Python binding of Redland on Linux. The SPARQL query language has simplified a lot of interfacing with the data, an XML projection of results is easy to convert into any target format. One of my dayjob contracts requires a kind of validation outside of RDF/OWL, this I'm implementing using custom (mostly hard-coded) logic on top of the RDF/OWL representation. There's too much data for complete OWL DL reasoning in the large, but for sanity-testing and local data checks Pellet has been useful. (The dayjob stuff is based on Jena/Java, the deployment target is again Linux but I've been developing on this, Win32 and more recently OS X). Throughout this work I'm mixing and matching numerous RDFS vocabularies/OWL ontologies as demanded by the domain. The layering of the framework allows me to do all this in a Web-friendly, consistent fashion. If this isn't modular, I don't know what is. > Moral: I suggest that the SemWebbers either think more like > Steve Jobs than Bill Gates or that they do more design > competitions and evaluation of alternatives. I seem to remember Bill Gates had little to say about the Internet in the first edition of "The Road Ahead". In the last year or two both Apple and MS have been leaping on the lowest rungs of the SemWeb layer cake, notably profitably "embracing and extending" the XML applications RSS and Podcasting. Believe me, they're not good role models for logic on the Web ;-) Cheers, Danny. -- http://dannyayers.com
Received on Friday, 31 March 2006 12:33:24 UTC