Dear Gerrit,

thank you for your note. When I opened my copy of DSSAT v4
from April 2004 I found many chapters that were still empty
and labled "under construction". Is there a new version available ?

The available information in DSSAT 4, Volume 4 indeed
provides a wealth of quantitative information about
DSSAT, but it forces me to understand the full
breadth of a crop model, which is a tremendous task
and requires an enormous amount of time.
We actually worked with the Fortran code a couple
of years ago, but had to give up for this reason.

It would be much more convenient to have a testing
environment available, which provides quick
and documented access to a certain model component
(sub-model, routine, or even a single variable or constant)
that someone is interested in. It would also be desirable
to have mechanism available that maintains the integritry
of the entire model structure while allowing for component
changes or even replacements (So called "test-driven development").
This would allow for a true scientific testing of the underlying
parameterization. Procedual languages like Fortran do not have the
flexibility to make such testings possible, which
was the reason for mentioning Java that seems to
evolve as a lingua franca of quantitative model
development in many areas of science.

Regards,
Matthias



Gerrit Hoogenboom wrote:
> Dear colleagues,
>
> I appreciate the earlier discussion. The source code for the Cropping
> System Model of DSSAT is in the public domain, but not available as a
> download from the ICASA web site. If you are a registered user of
> DSSAT, we will be glad to provide you with the most recent source
> code. Many components of the CSM v4 source code are very well
> documented and can be found in DSSAT v4, Volume 4, as distributed on the
> CD.
>
> We are in the process of developing a new GenCalc for DSSAT v4 and it
> will be included in our next release. If you are interested in
> testing the Beta version, please contact Dr. Tony Hunt
>
> We are very open to improvements of the current models as well as
> development of new tools. If you would like to contribute, please let us
> know.
>
> PLEASE DO NOT REPLY TO this message if you would like a copy of the
> source code or need to contact Dr. Hunt.
>
> Regards,
>
> Gerrit
>