Re: Deliverable 12.1.1

Danny Ayers wrote:
> 
> What have you got?? ;-)

Well hopes, aspirations that sort of thing ...  :-)

Actually, we do have some code from our "ePerson" community information
management experiment, though that was an intranet kind of thing with no
connection to blogging so we might end up reusing some ideas more than the code.

> I'm just in the middle of a rebuild, after deciding what I'd got was getting
> too crufty. I think the closest thing around to what I'm aiming for is
> Tinderbox, (though I haven't got a Mac on which to play), but more node &
> arc based & very web & RDF-aware. In other words a 'Personal Knowledge
> Manager', although I'm not sure how far I want to go into the email side of
> things. I'm approaching the first goal which is mindmapping/(semantic)
> blogging with a graphic UI.

Sounds great. 

> >If you have any
> >thoughts on requirements for such tools then we'd be really interested in
> >hearing about them.
> 
> I don't think we can have too many schema registries... but a loose thought
> closer to your deliverable would be a registry/discovery tool for not only
> RDF-aware apps, but a kind of Citeseer for related (primarily open source)
> software. 

Interesting thought. Blogged bibliography of software tools ... yes, worth
exploring.

> For example, when I started on the current project I had a pretty
> thorough search for node & arc graphing code & libraries. Nothing was really
> close enough, so I coded a lot from scratch. Months later, purely by chance
> I came across the library I'm using now (fortunately it was around the time
> I decided on the rewrite, so I can still save some effort).

Which library did you end up using?

> Something else which I reckon might be popular would be a 'Jena Lite'
> (mentioned on that list not long ago). An RDF API for folks such as RSS
> feeders that don't want everything Jena can offer, but want goes beyond
> straight XML or regexp-based hacking.

Good suggestion but probably outside the scope of this particular activity. One
challenge I'd see with a 'Jena lite' would be the parser. ARP is very much aimed
at quality and full standards conformance rather than 'liteness' - hence the use
of Xerces which dwarfs jena.

> PS. Winer/Wiener
> see : http://www.cs.umanitoba.ca/~djc/wiener/w8.html

Thank you for that!

Dave

Received on Wednesday, 6 November 2002 09:44:01 UTC