Re: [sdpd] Espoir 3.5 release coming soon

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

>Might it be possible to have ESPOIR in a mode similar to 
>EXPO Direct Methods software where ESPOIR can be given the 
>Cell Contents - and it can transparently handle
>locating atoms on special and general positions -
>and balancing the atoms out? 

Might be possible. In fact the new version is already doing that,
more or less well, when the option to enter in the STRUPLO 
part is selected, with automatic polyhedra recognition. But this is not 
optimized yet. Moreover, the Direct method software are 
able to explore thousand of structure models in a few minutes,
which is not the case for the molecule location and/or Monte Carlo
from "scratch" software for which 10-100 independent tests
may need hours or days. So that helping the program with good
special position hypotheses is more essential. Those methods are not
really working in the same application niche. The latter is
recommended only when the direct methods fail... and chances
to continue to fail remain large (scratch mode).

Armel Le Bail
http://www.cristal.org/course/