VMD-L Mailing List

From: Axel Kohlmeyer (akohlmey_at_cmm.chem.upenn.edu)
Date: Sun May 24 2009 - 08:38:11 CDT

On Sun, 2009-05-24 at 02:31 -0700, Rob wrote:
> Axel Kohlmeyer wrote:
>
>
> > the best way to start, is to modify a plugin that does already
>
> > something similar for a similar software. but then there is also
> > the plugin programmer's guide at:
> > http://www.ks.uiuc.edu/Research/vmd/plugins/doxygen/
>
> A while ago I have improved and extended the VASP plugin.
> I am about to use those plugins as examples for the abinit plugin.

ok. i don't use VASP and never used those plugins, so i cannot
comment on them. they look a lot like they are modeled after
the cubeplugin. does that mean, you are mostly interested into
making gridded data readable? have you checked, whether this
is stored in reciprocal or in real space?

> > in VMD 1.8.7 the molfile plugin API was significantly expanded.
> > depending on what information you want to hand over to VMD,
> > you should start working off the CVS version.
>
> Hmm, no; I don't want the burden of compiling VMD itself.

you don't have to compile VMD to compile the plugins. still
with over two years since the last release you don't want to
start working off an antiquated version of the code.

> What's so neat about the plugins is, that they are independent
> of the VMD code. I can code, compile, and install plugins in
> the directory tree of the precompiled VMD.
>
> Hence, at present have only access to 1.8.6 and its plugins.
> I tried the 1.8.7beta release, but it doesn't provide the sources
> of the plugins....

all you need is to apply for read-only cvs access and that
is usually done very quickly.

> But, eh, do all plugins then need a severe rewrite with 1.8.7?
> Is the plugin API only extended, or completely changed?
> Is the plugin interface change already documented?

the main documentation is (in) the source code. the changes
are incremental, so old ABI plugins should continue to keep
working, but there are several exciting new features like
the ability to (optionally) store velocities and complete
wavefunctions, if projected on atomic orbitals. however, that
last part is still being worked on, so it is in a bit of flux.
there is a plugin for gamess and i'm currently working on
plugins for gaussian and CPMD.

there also is a project underway to have all output redirected
to a console log window which requires some (transparent) rewrites
of the plugins (they still compile for 1.8.6).

cheers,
   axel.

> Rob.
>
>
> 

-- 
=======================================================================
Axel Kohlmeyer   akohlmey_at_cmm.chem.upenn.edu   http://www.cmm.upenn.edu
   Center for Molecular Modeling   --   University of Pennsylvania
Department of Chemistry, 231 S.34th Street, Philadelphia, PA 19104-6323
tel: 1-215-898-1582,  fax: 1-215-573-6233,  office-tel: 1-215-898-5425
=======================================================================
If you make something idiot-proof, the universe creates a better idiot.