[sdpd] Re: SDPD stagnation

[Armel Le Bail <armel...@fluo.univ-lemans.fr>]

>No one is saying that a good GUI
>(or for that matter, a good command
>line driven interface) overcomes the 
>limitations of a poor underlying program 
>or poor data.
>
>However, if the underlying program
>is good, a good interface can make all
>the difference between a program being
>taken up by people who are not steeped
>in crystallography, and its use being
>restricted to experts.  

There are 2 kinds of poor data :
  - poor raw data, only experts can cope with them ;
  - poor data due to inadequate data treatment
    (even if the raw data is splendid) by people
    who are not steeped (but some are) in crystallography.

The complete package in which you insert the
powder and get the structure (and the manuscript to be
published) is not already written. DASH, Endeavour or
PowderSolve are at a defined step of the SDPD process.
You do not simply enter a raw pattern inside. You still
need steeped crystallographers who may not care to GUIs
if equivalent freeware is available (moreover, they usually
cannot afford all those too many commercial programs).
The difference will be between people wasting their time
by trying those programs with false cells, or false space
groups, or inadequate starting models, and those who will
enter the convenient data inside (mostly with a text
editor ;-).

ORTEP is still freeware. The source is open. Unfortunately
no version with a GUI is delivered with open source...
Unfortunately, because such open sources would show
to interested people how to do GUIs with some given
compiler.

Don't believe that I would not build a GUI for some
programs if I could (ESPOIR for instance). In fact, I
continuously think to GUIs. The publication of a book
about the Digital Visual Fortran compiler was recently
cancelled. I subscribed and prepaid for a sample of that
book at fatbrain.com. Now you will have to wait more...
If I find time and skills to do a GUI, I will put it as an
open source. A question of culture, really. The
belief that knowledge should be shared between
scientists for free, for non-commercial purposes, at least.
Many guys succeeded a lot with that culture : Rietveld,
Sheldrick, etc. You may still find many people using
SHELX without any GUI. SHELXS is even the program
which solved the largest number of structures from
powder diffraction data, by Patterson or direct methods.
The use of GUI for driving SHELX has no clear
influence.

Best,

Armel Le Bail - Universite du Maine, Laboratoire des Fluorures,
CNRS ESA 6010, Av. O. Messiaen, 72085 Le Mans Cedex 9, France
http://www.cristal.org/