a story of log:semantics and log:includes

regarding our SOURCE issue...
http://www.w3.org/2001/sw/DataAccess/rq23/#i_sourceImplExp

and in particular, my action...

ACTION DanC: explain evolution of log:semantics/log:includes from
uri-is-graph to uri-is-doc in cwm, to inform discussion of SOURCE

 -- http://www.w3.org/2001/sw/DataAccess/ftf2#brqlw


For a tutorial introduction to log:semantics/log:includes, see
  http://www.w3.org/2000/10/swap/doc/Reach

Also, keep in mind the relationships between URIs, resources,
and representations; see the figure in
  http://www.w3.org/TR/webarch/#intro

In the current/modern design, log:semantics relates
a resource to a graph obtained from a representation
of of the resource, and log:includes releates
2 graphs (log:includes is pretty much a synonym
for RDF simple entailment http://www.w3.org/TR/rdf-mt/#entail).

The current version of
  http://www.w3.org/2000/10/swap/test/includes/t7.n3
uses them like this...

 { [is log:semantics of <>] log:includes ...

In the original design, a resource was equated with
the graph you get from parsing its representation;
there was no need for log:semantics; it was owl:sameAs.

The original version of t7.n3 just said...

{ <> log:includes ...


I can't remember which was the convincing motivation, but
some of the motivations include

 -- a desire to talk about the resource and the
	graph independently

 -- a desire to model resources whose state changes
	(we haven't gotten there yet; cwm assumes
	that it can cache representations forever and
	that log:semantics is functional)


So I'm not sure how much this informs the source issue,
but I hope it stimulates some thinking.

In case I or somebody else can correlate these dates to
something else (#rdfig logs, for example) here's an
excerpt of the cvs log of t7.n3:


----------------------------
revision 1.4
date: 2002/02/21 04:12:29;  author: timbl;  state: Exp;  lines: +7 -6
after testing listfix branch
----------------------------
revision 1.3
date: 2001/09/21 20:44:07;  author: timbl;  state: Exp;  lines: +2 -2
Name change resolvesTo to semantics, hasContent to content. Some schema
cleanup to validate under check.n3
----------------------------
revision 1.2
date: 2001/05/21 03:12:00;  author: timbl;  state: Exp;  lines: +1 -1
after changes to namespace and use of this with forall
----------------------------
revision 1.1
date: 2001/05/10 01:59:36;  author: timbl;  state: Exp;
test files for log:includes


-- 
Dan Connolly, W3C http://www.w3.org/People/Connolly/

Received on Tuesday, 24 August 2004 14:27:36 UTC