Re: Fresnel implementation for Jena

From: Andrew Green <ndrw_grn_at_yahoo.com.mx>
Date: Mon, 28 Nov 2005 11:42:48 -0600

El lun, 28-11-2005 a las 08:46 +0100, Emmanuel Pietriga escribió:
> Andrew Green wrote:
> > Emmanuel:
> >
> > How far along are you in the implementation of Fresnel Display
> > Vocabulary for Jena and IsaViz? Have you designed any API's yet? Any
> > skeleton code for classes? General sketches of how the implementation
> > will work?
>
> I've been focusing on FSL until now, providing implementations for Ejna,
> IsaViz and Sesame. Now that I'm done with it, I'll spend my "Fresnel
> time" on implementing the core lens and formatting vocabularies in IsaViz.
>
> I plan to reuse as much of the code Ryan wrote as possible, though I
> probably won't be able to reuse it "as is", as IIRC it is written for
> Sesame. But migrating it to Jena should not be too difficult (I did it
> the other way around for the FSL engine, and if you take a graph
> perspective on the Sesame repository, the Sesame API is fairly close to
> Jena's).
>

OK. In the Simile subversion repository, under fresnel I see a "trunk"
that runs with Sesame, and under "branches" I see "jena-base", and a
version that apparently runs under Jena. How do the two versions in CVS
compare? Is the Sesame version more complete, or are they roughly equal
in functionality and maturity?

> The rought sketch for the IsaViz/Jena implementation is:
> - a module A that builds an in-memory representation of lenses and
> formats from their RDF/XML, N3, ... representation (I have to look at
> what Ryan did and see if I can reuse is data structures "as is")
> - a module B that applies lenses and formats to a model and generates
> the XML intermediate result tree
> - a module C that builds the SVG final rendering from the XML result tree
>
> Module A will be IsaViz-independent. Module B will probably be
> IsaViz-independent too, but that is not yet 100% clear. Module C will
> clearly be IsaViz-specific.
>
>

Your sketch for modules A and B is roughly equivalent to what I'd
imagined. Our equivalent to module C will be the code that constructs
views in our MCV framework. (We're using Ruby on Rails, and sending data
from a Java component to Rails over XML-RPC. The Java component
interfaces with Jena; this is where the Fresnel processing will take
place.)

> > I'm asking because I'd like to implement a few of the most basic aspects
> > of Fresnel quite quickly, and if in doing so I could contribute to a
> > general implementation of the norm, I'd be very glad to.
>
> My problem is that I won't be able to contribute much code for the next
> two weeks. So depending on what "quite quickly" means and what basic
> aspects you are interested in, it might be possible. Can you elaborate
> on your needs?
>

Quickly means, unfortunately, that in the couple few days I have to code
some quite basic lens functionality. Something I expect won't be too
hard (especially given that I'm not starting from scratch, but from an
existing codebase).

This is of course not a "request" for an implementation... I was just
wondering how whatever I do might best contribute to Fresnel development
in general.

When I have something to show I'll let you know, and if any of our work
is useful to you... After all, it's all free software...

Greetings,
Andrew Green

__________________________________________________
Correo Yahoo!
Espacio para todos tus mensajes, antivirus y antispam ¡gratis!
Regístrate ya - http://correo.yahoo.com.mx/
Received on Mon Nov 28 2005 - 17:33:45 EST

This archive was generated by hypermail 2.3.0 : Thu Aug 09 2012 - 16:40:51 EDT