- From: Dan Connolly <connolly@w3.org>
- Date: Thu, 15 Jul 2004 15:06:34 -0500
- To: Jim Hendler <hendler@cs.umd.edu>
- Cc: Rob Shearer <Rob.Shearer@networkinference.com>, public-rdf-dawg@w3.org
On Thu, 2004-07-15 at 13:30, Jim Hendler wrote: [...] > I assume Dan that you were aiming for easy to implement as the main > point actually, no. I think the main point is... "the less powerful the language, the more you can do with the data stored in that language . If you write it in a simple declarative from, anyone can write a program to analyze it in many ways." It's sorta subtle. The reason HTML is better for the web than TeX isn't that it's easier to parse/implement; it's that it's more reusable accross devices, etc. > (i.e. implementational complexity over computational > complexity) and particularly easy to evaluate whether the system did > the right thing? I'm still okay with your wording (again, assuming > we drop the last sentence) but wanted to be clear you mean the > complexity of languge implementation over the complexity of the > computation allowed -- i.e. relational databases are limited in > expressivity to keep things polynmial in most cases, even though SQL > is a bear to implement in full, in part because of this. We would > aim for the opposite - ease in design and implementation, as opposed > to guarantee of efficient computation (which, given the expressivity > of RDF we provably could not gaurantee without putting new > restrictions on the RDF graphs) The "expressivity" of RDF is totally separate from query evaluation. Inference and query are separate in the design we're persuing. The complexity I'm talking about is the complexity of the algorithm that computes the results from a query and an input graph. -- Dan Connolly, W3C http://www.w3.org/People/Connolly/
Received on Thursday, 15 July 2004 16:06:57 UTC