On Thu, 27 Feb 2003, Miguel wrote:

> The work I have been doing on jmol the last few months has been focused on
> performance and rendering quality. We are hoping to release it (Jmol v6)
> within the next few weeks.
> 
> I would be *very* interested in getting your feedback on Jmol 6 performance.
> 
> How big are the molecules you are working with?
> 
> What kind of hardware are you running on?
> 

I would consider a 20^3 configuration (~ 8000 atoms) to be small- but
jmol would not handle much larger than this. On a P4 1.8GHz jmol can
several minutes to load such a configuration- rotating it is very slow
(although wireframe mode is fine). In contrast- VMD works smoothly
with those sizes. It would be great to be able to display such
configurations on the web, using the jmol applet. I realise java will
never compete with native opengl code- what kind of improvement could
you achieve with your performance enhancements?

> 
> There may even be postscript output ... I am somewhat embarassed to say
> that I don't know.
> 

There are good postscipt and povray renderers out there
(eg. molscript)- perhaps it is not worth it to compete with them.

The best thing about jmol is the ability to run it as an applet- so
you should try to keep it small and fast. It should also be possible
to load modules as required- eg. to do protein specific
visualisations, charge fields etc. If it was quick enough it would be
nice to connect to the output of a simulation- perhaps you could view
the evolution of your simulation over the web?

a

-- 
Aonghus Lawlor           [EMAIL PROTECTED]
00 353 1 7062418         http://fiachra.ucd.ie/
Irish Centre for Colloid Science & Biomaterials
Department of Chemistry
University College Dublin

Attachment: pgp00000.pgp
Description: PGP signature

Reply via email to