Re: Nodes and Arcs 1989-1999: WWW history and RDF (revisited...)

Hi

Thanks for the suggestions, I think you're right, it's kindof a jumble
right now. Tempting also to bloat it out with all sorts of things that
have happened in the last 10 years, SVG, XSLT etc... and how they might be
part of a bigger picture. I'll try to hold back from a 'grand unified
theory of everything' in next version at least. That's more the job of 
http://www.w3.org/DesignIssues/ ->
http://www.w3.org/DesignIssues/Architecture.html


> If would be great to see a miniature data set that fit the CERN diagram
> and which could be used for demo and testbed purposes, but that's
> probably asking for a bit much. 

Oh, I can help you out there: the version from last year included 
a little RDF example based on the CERN diagram, with the RDF/XML syntax
annotated with comments to help folk understand the way RDF folds this
stuff into XML:

http://lists.w3.org/Archives/Public/www-rdf-interest/1999Nov/0039.html
-> http://www.w3.org/1999/11/11-WWWProposal/
-> http://www.w3.org/1999/11/11-WWWProposal/proposal.rdf

as well as a demo/testbed that queries the RDF statements in there:

-> http://www.w3.org/1999/11/11-WWWProposal/rdfqdemo.html

That demo uses an early version of Jan Grant's 'prolog in javascript'
system (which I'm trying to persuade him to rename 'Pipsqueak', for
obscure reasons), and may only work on some Javascript/ECMAScript
browsers. See http://tribble.ilrt.bris.ac.uk/~cmjg/logic/ for more
recent versions. It nearly includes an SQLish ("Squish"? :)
query interface similar to the one Guha's using in
http://www.guha.com/rdfdb/ and that sketched in TimBL's WebIzing doc, 
http://www.w3.org/DesignIssues/Webize.html


Regarding the testbed version of CERN example, I decided not to spend too
long puzzling over how best to represent it faithfully in an RDF model,
since (a) the original was quirky and evocative rather than a serious
piece of modelling, (b) many of the constructs  can be replaced with RDF
builtins (eg. type/class) or vocabularies like Dublin Core. I also didn't
write a schema for the vocab I used, which is a shame since I think we can
query class and property hierarchies in the JS tool now (not to mention
simple inferencing examples...).

Anyway it would be much more fun to build a real year "2000
version" describing current organisations, RDF IG members, our
publications etc. that we might use for genuine collaboration and
query. More on that another time (I've
an RDF robot scuttering around just such a distributed database already,
but want to polish the system a little more before showing that half-baked
demo ;-)

Dan

ps. the CERN diag as RDF re-visualised in SVG and GIF is one of the demo
files I use in RDFViz,
http://www.ilrt.bris.ac.uk/discovery/rdf-dev/rudolf/rdfviz/
-- I think the GraphViz SVG support might need updating before this will
work, currently my SVG plugin doesn't display the file at all. Anyone
fancy having a look at that problem? See 'accessibility features of SVG
Note' for more context on SVG and metadata...
Oh, SVG= Scalar Vector Graphics, for those without built-in glossaries...

Received on Wednesday, 27 September 2000 10:48:50 UTC