GeoModeller inversion development [SEC=UNCLASSIFIED]

12 views
Skip to first unread message

Richar...@ga.gov.au

unread,
Mar 29, 2007, 8:18:51 AM3/29/07
to geomo...@googlegroups.com
Hello All.

Paul Donchak raised the subject of the present state of the GeoModeller
inversion module.

"An issue not touched upon by Peter is the status of the inversion
module: it does not seem to be in a state that the average geologist can
really use, and now seems to be a standalone module requiring scripting
and higher level geophysical knowledge than originally planned. In the
light of this, the pricing policy for the inversion module is difficult
to understand."

The GeoModeller inversion approach is an exciting and innovative method
involving exploration of many different geological scenarios and assessment
of the likelihood of each of these scenarios with respect to gravity and/or
magnetic geophysical data.

I have no particular view on the pricing, but I have supported the decision
by the software developers to carry out the initial phase of the development
of this module in a batch environment.

A variety of stability issues have been noted by people when using the
geological mapping parts of GeoModeller, some aspects of which can be
attributed to programming of a GUI environment. The development of the
underlying computational foundation for the inversion procedure is a big
enough initial challenge without increasing the degree of difficulty by
concurrently implementing the prototype codes in a fancy user interface.

There have also been many calls for more thorough testing of software. In the
initial stages of the development of inversion options, a batch environment
has ensured that processing sequences can be reliably reproduced and easily
added to the automated testing sequence carried out for each new software
build.

Many of the inversion processes take a considerable length of time to run.
There is little point having a GUI for these processes.

Once the initial development phase has been completed and tested, I believe
that the plan calls for implementation of default and advanced inversion
workflows via a more intuitive user interface. In the interim, I would call
on users to support the developers to do what we have asked them so often to
do; to produce robust, well tested software which fulfils a demanding
technical requirement. I feel comfortable that they will be able to do this
more effectively and sooner if they continue to work in a batch environment.
I can accept for now that I will have to spend a short period of time
learning a few simple commands to work in this environment, and to invest
some extra effort to understand the concepts behind this inversion approach.


Regards ... Richard Lane

Reply all
Reply to author
Forward
0 new messages