[Biojava-l] GO web service

Chervitz, Steve Steve_Chervitz@affymetrix.com
Tue, 10 Sep 2002 18:08:43 -0700


As a person who has been involved in both approaches, I can see the
strengths of each. With the top-down, consensus-based approach, it's great
to have an abstract model as the point of discussion and future reference.
The model developed by the gene expression standard was used to
auto-generate objects and parsers in both Java and Perl that can be used to
serialize/de-serialize gene expression data. That was a big win. It would be
great to have something similar within Bio*. 

One point to consider is that the gene expression work was entirely
data-centric. There are no methods in any of the objects in the model.
Things could get trickier when you're using this approach for
data+service-oriented objects and the underlying data exchange layer is not
XML. But maybe not.

I think a possible impediment to adopting a model-based approach in the Bio*
community has been the lack of good, freely available, Unix-based tools for
object modeling. The rise of ArgoUML/Poseidon could really help here
(http://argouml.tigris.org/, http://www.gentleware.com/).

There's also probably some cultural issues at play as I think many
developers in the Bio* community aren't so familiar with modeling concepts
and processes. We know best how to communicate with working code and
databases, so that's what we tend do. That's not saying that it can't
change, though. It may just take some time for the concepts & tools to
percolate till we develop a critical mass of interest in Bio* modeling.

For additional insights, I think Alan hits the nail on the head here:
http://biocorba.org/pipermail/biocorba-l/2002-June/000326.html

For more info about the gene expression standard:
 o http://www.mged.org/Workgroups/MAGE/mage.html
 o http://www.omg.org/news/newsletters.htm  (see p.7 of April 2002 issue)

Steve

> -----Original Message-----
> From: Dickson, Mike [mailto:mdickson@netgenics.com] 
> Sent: Tuesday, September 03, 2002 12:06 AM
> To: 'Ewan Birney'; Dickson, Mike
> Cc: Patrick McConnell; Simon Brocklehurst; Brian Gilman; 
> biojava-l@biojava.org
> Subject: RE: [Biojava-l] GO web service
> 
> 
> I didn't take it as a snipe.  I simply wanted to point out 
> that there have been successful examples of the top down 
> consensus driven approach (the gene expression standards work 
> is a very good example). The OMG brings a mature, consensus 
> driven adoption process and where appropriate it should be utilized.
> 
> Mike
> 
> > > > The "official" OMG way to do this is via UML and some buzz word 
> > > > about technology free modelling, but there is still a massive 
> > > > tendancy for top-down-design-by-committee, which I think just 
> > > > doesn't work, and
> > frankly
> > > > the Biojava/BioPerl data model (which are reasonably in 
> sync, give 
> > > > or
> > take
> > > > the odd split location thang) is a fine way to start, and the 
> > > > usual "propose idea on list, and whoever codes it wins the 
> > > > argument" is a
> > far
> > > > better resolving procedure than committees
> > >
> > > I'm not going to get into this one other than to say that this is
> > clearly an
> > > opinion and it doesn't represent everyone's position (not mine at
> > least).
> > > For the record, I'm on the board of directors of the OMG.
> > 
> > Apologies Mike - that was not meant as a low snipe. I do 
> have a more 
> > serious point which the "working code, open discussion and 
> progressive 
> > refactoring" which is how the Bio* projects work is bascially 
> > ExtremeProgramming in a geographically distributed way, and I think 
> > gives the best results. Of course, YMMV.
> 
> 
> _______________________________________________
> Biojava-l mailing list  -  Biojava-l@biojava.org 
> http://biojava.org/mailman/listinfo/biojava-l
>