[rdfweb-dev] Re: Atom-FOAF

[Henry Story]

Here are some of my own thoughts concerning this [1] work:

	- How is this an improvement over RSS1.0. I am very new to RDF, RSS 
and blogging,
		and have not yet had much time to look at what the RSS1.0 community 
has come up
		with. Perhaps I am just (badly) re-inventing the wheel. I would love 
feedback
		from that area on how they think one could improve on their work, and 
the 		pitfalls one should avoid. Perhaps there is an RSS1.0 meeting 
point I should 		send this to.
	- clearly there is a lot of overlap with Dublin Core. It would be 
interesting to
		make these obvious by stating the equality relations between the 
properties
		I have defined and the Dublin Core ones in the Atom.owl file.
	- Should the foaf:Agent class be accessed directly, as it is in the 
example, or
		should one interpose an interface class, say atom:Agent, as Danny 
Ayers does in
		his Atom-OWL[2] work from which mine is derived. An interface class 
would
		be one that specifies the minimal requirements. If one were to do 
this, then 		presumably somebody could still end up with something 
directly referencing
		foaf:Agent, but I suppose they would have to import some ontology 
that makes
		the equality explicit. What is the best way to deal with importing 
such rules?

  Anyway as a side comment I have to say that this RDF thing is really 
easy to get a grip on. With tools such as Protege [3], from Stanford 
University, writing object oriented RDF is a breeze. Languages such as 
N3 make reading RDF as easy as reading the
morning newspaper. And it really feels good to have something that is 
based not only on such widely adopted standards as XML, but also 
supported by the full mathematical infrastructure of set theory and 
logic [4]. I love power tools, and we have not found anything much 
better than mathematics yet. It is only by building on the shoulder of 
giants that we can hope to go any further. I think that the view from 
up on Tim Berner's Lee's shoulder is going to be pretty awesome.


Henry


On 15 Aug 2004, at 16:55, Henry Story wrote:

>
> I have written out a full description with a lot of illustrations on 
> how to FOAFify Atom [1]. The description is written out as a blog 
> which is using the format it is advocating to back it up. So it is 
> both a description and a proof of concept.
>
> I see this being useful to the foaf, atom, and bloged communities for 
> different reasons:
> 	- for the atom community it could bring bring some ontological 
> insight, and lead to clarifications - mapping always do that.
>     - for the foaf community, I suggest this as an opening into the 
> blogging space
>     - for bloged, the blog editor I am working on, this will be useful 
> at the very
> 	least to help nail down the needed data structures, though I hope to 
> use it both as
> 	an internal and external data-structure.
>
> I look very much forward to any criticism. This document is the result 
> of many such criticisms, for which I am very thankful.
>
> Henry Story

[1] http://bblfish.net/work/atom-owl/2004-08-12/blogexample.html
[2] http://semtext.org/atom/index.html
[3] http://protege.stanford.edu/
[4] http://www.w3.org/TR/rdf-mt/