Dear John,

I am definitely agreeing with you, but this would
require access to a well documented source code,
which does not seem to be openly available in
the case of the DSSAT family of models.

Pseudocode would provide deeper insights into a
model's theory, but isn't helpful either under
practical conditions, because a true understanding
of a model can only evolve in "debug mode".

The Java programming language and modern software
design techniques provide excellent mechanisms
for establishing standardized developing frameworks.
Are there any plans for applying these technologies
to the DSSAT suite and might even turning it into an
open source project ?

Best wishes

Matthias

John Roy Porter wrote:
> dear Aston
>
> the best way to calibrate a model is to understand what the model is
> trying to do and the processes it is trying to describe.
>
> best wishes
>
> John R Porter
>
> John R Porter DSc
> Professor of Agroecology
> Environment, Resources and Technology,
> Faculty of Life Sciences
> The University of Copenhagen
> Hoejbakkegaard Alle 9
> 2630 Taastrup DENMARK
>
> email: [log in to unmask]
> vox: +45 3528 3377/3375
> fax: +45 3528 3384
> mobil: +45 2652 1484
>
> .....Credeamus vita non ioca est...
>
>>>> "Chipanshi, Aston" <[log in to unmask]> 14/02/2007 17:32 >>>