[Durito-dev] Re: HyperDE: need help?

[Daniel Schwabe]

Caro Andrew,
suponho que  você consiga ao menos ler Português, assim como nós 
conseguimos ler Espanhol...

ahg wrote:

> Dear Demetrius Nunes and Daniel Schwabe,
>
> First of all, thank you very much for writing and for looking at our
> project, and for sending your paper. Please also accept my apologies for
> not responding sooner.
>
> HyperDE caught our attention about two weeks ago, in part because it
> uses Rails to create an interface to Semantic Web data, as does (or
> will) our applicaiton.
>
> In addition--if I understand correctly--HyperDE (and SHDM) have as a
> starting point theoretical proposals regarding knowledge representation
> and user interfaces.
>  
Sim. OOHDM e SHDM são métodos para autoria de aplicações hipermídia, na 
web ou não. Adotam a abordagem de desenvolvimento baseado em modelos.

> Unfortunately, right now we are working under a very strict schedule
> that obliges us to have a collection of archival images and related
> metadata publicly available online by the middle of December. 
O site do Projeto Portinari, sobretudo a parte do acervo (collection), 
foi projetado com o OOHDM (www.portinari.org.br).

> Because of
> this, I thought that now was not the best time to learn SHDM and study
> that methodology and HyperDE more thoroughly; we already have a clear
> enough idea of how the application we are creating will function, and I
> felt that the surest course, for the time being, would be to simply
> follow through with the implementation of that design.
>
> However, this decision does not preclude future collaboration! Our
> project will continue next year, when we'll have more time to study
> proposals such as yours--and perhaps even consider ways of working
> together, or making our work complementary.
>  
Temos bastante experiencia neste tipo de aplicação, devido à colaboração 
com o Projeto Portinari, que contém a documentação tanto da obra, quanto 
da época (documentos, fotos, entrevistas, filmes, etc...) do pintor 
Candido Portinari.

> One thing that you should perhaps note is that our project is rooted in
> the Social Sciences and Archiving, as well as Computer
> Science--specifically, we seek to create tools that (1) improve the way
> Social Science researchers interact with primary source materials, and
> (2) organize archival data in a rigorous, flexible and useful manner. A
> few papers about the project are available from
> http://durito.nongnu.org.
>  
Sim, eu já tinha dado uma olhada. Muito interessante.

> (In fact, a longer text just appeared in the Janary-June issue of Anais
> do Musueo Paulista, published by the Museu Paulista de Universidade de
> Sao Paulo.)
>
> Of course, in this domain--as in any other--the attempt to create an
> effective human interface to a complex knowledge representation stands
> to benefit greatly from theoretical research into that process, and this
> is, in the medium and long term, why it is important that we study and
> understand work such as yours.
>
> On a more specific note:
>
> 1) Perhaps in January we could analyse our system's current and future
> interface and features using SHDM. (By the way, a partical mock-up of
> our planned interface can be found at
> http://durito.nongnu.org/maqueta3/prueba_interfaz3.html.
Eu vi, muito bom. Parabéns!

> Also, at the end of
> http://durito.alesoft.net/durito/attachment/wiki/DocumentacionDisenio/informe_etapa_VI.pdf?format=raw
> is a brief description of functions to be added to the system in future
> versions--note, however, that the rest of that document is a bit
> outdated.)
>  
O que o OOHDM/SHDM podem acrescentar são formas de compor visões 
navegacionais que apoiam as tarefas previstas para seus usuários.

> 2) Any comments or suggestions you have regarding any aspect of the
> project would be extremely welcome. (You might even want to subscribe to
> our developers' mailing list--see
> http://lists.nongnu.org/mailman/listinfo/durito-dev. By the way, I am
> forwarding your messages and this reply to that list.)
>
> 3) I would like to understand specifically how your system replaces
> Fresnel. How do you specify what data is displayed regarding a given
> resource? How do you specify the transformation of complex graphs into
> simpler structures? How does data ingestion/modification work, and what
> mechanisms do you have for data validation? And finally, what
> problems/limitations to you see in Fresnel?
>  
O que temos na metodologia é, na verdade, COMPLEMENTAR a Fresnel. A 
nossa abordagem trata a interface desde um ponto de vista mais abstrato 
inicialmente, focalizando apenas na função cumprida por cada widget 
(exibição de dado, captura de dado, disparo de evento). Posteriormente, 
a interface abstrata deve ser mapeada em uma interface concreta, e, 
neste caso, isto poderia ser especificado com Fresnel. Estamos agora 
investigando esta linha.

> Also, please let me know which language you prefer to use--English or
> Spanish?
>  
Me puedes contestar en español, pero en realidade no se escribir bien en 
español; espero que alcances a leer en Portugues. Al peor, mantenemos el 
Ingles...

>  
Abrazos,
Daniel