Resource Token Exchange

Hi WG,

A few days ago I saw a note from Bob
Ferris<http://lists.w3.org/Archives/Public/public-rdf-comment/2011Mar/0000.html>
to
this list, referencing Pat Hayes'
message<http://lists.w3.org/Archives/Public/public-rdf-wg/2011Feb/0060.html>to
the WG suggesting ``token'' as a useful term.  Believe it or not I was
exploring more or less the same idea at roughly the same time, so not only
was I happy to see the message, I was motivated to scrape together some of
my notes, which I've posted as a PDF
doc.<http://blog.mobileink.com/2011/03/resource-token-exchange.html>
Sorry
I don't have an HTML version, I'm a little pressed for time, but if anybody
really wants one I could get to it.  Also, it's a little rough and
disorganized, but it is readable, and those interested in the token stuff
might want to take a look at it; I go into more detail that Pat's note does,
with more attention to the type side of the type/token distinction.

There are a few other bits some might find interesting (please consider it a
set of conversation starters rather than a finished thesis).  In particular
I've taken a crack at using Category Theory to provide a formal semantics
for just about everything - IRIs, resources, RDF etc.  It seems to work
pretty well, and leads to a simple, specific semantics that has the virtue
of eliminating the philosophical handwavery and smoke and mirrors that
usually accompany discussions of "resources".  For example the stuff about
whether an resource is an "info" resource or not just goes away, as a false
(or at least pointless) dichotomy.  Since I'm not a mathematician and
definitely not a CT specialist I would appreciate feedback from somebody who
is regarding the CT stuff.

While I'm at it, I've been following some of the WG discussions regarding
the "g-" stuff.  My reaction:  double and triple cringe.  I sure hope you
don't succumb to the temptation to annoint such stuff; if you look closely
at it it either doesn't work or is just a old wine in new bottles.  I'm
pretty sure my eyeballs would refuse to read anything that used these terms.
The fact that the WG has had a hard time figuring out just what "g-boxes"
etc. actually mean is telling.  A possible way of addressing the same
concerns is just to use plain old graph theory, particularly super- and
subgraphs, with notions of denotation suitable to open world semantics.
 There's some language about that in my notes.

Cheers,

Gregg Reynolds