VMD-L Mailing List

From: Axel Kohlmeyer (akohlmey_at_cmm.chem.upenn.edu)
Date: Fri Jun 08 2007 - 08:35:04 CDT

On Fri, 8 Jun 2007, Harry Bullen wrote:

harry,

HB> Axel Kohlmeyer wrote:
HB> > On Thu, 7 Jun 2007, Harry Bullen wrote:
HB> >
HB> > harry,
HB> >
HB> > do you mean, to have VMD imported into other python
HB> > scripts or just use a python interpreter in VMD (instead of tcl).
HB> >
HB> > the latter has already been implemented (type 'gopython' in the
HB> > command line window, or use 'gopython script.py'), the former
HB> > is yet incomplete.
HB> >
HB> Yes I do mean the former, a VMD imported into other python scripts.
HB> While I know that I can use gopython to run a python interpreter in VMD,
HB> my mentor hear really wants to have a VMD module. Apparently, a lot of
HB> people here have trouble installing and using VMD because it has a lot
HB> of dependencies, and he believes that it will be easier to use and have

i canno agree there. this kind of problem does not go away from having
a python module, since this module, if precompiled, will still have the
same dependency problems as the regular VMD binary.

HB> less dependencies. We have already discussed alternatives to doing this
HB> and have come to the conclusion that this is the best way to go. Also,
HB> we don't need the module to implement all of the feature of the VMD,
HB> just atom selection, loading of files, vmdnumpy, and optionally the imd
HB> module.

ok, so your problem is the usual 'python problem', i.e. that the
python shipped with VMD is not compatible with the native python
on the respective machines, due to VMD being compiled with a different
and incompatible python version.

the remedy for this would be to compile VMD yourself using the
respective native libraries for python, tcl/tk and so on.
some comments on how to do this on a 32-bit linux machine and
even a matching .spec file to build a rpms for fedora linux
(core 4 and later, IIRC) are on the VMD public biocore project.
just create a (guest) account and join the VMD project.

HB> If someone could point me to so development code that would be helpfull
HB> as even a partial solution could help us here.

please check out the vmd homepage. it gives you the option of
downloading the released source and also instructions on how
to get (read-only) access to the cvs. if you want to continue
the vmd python module development, the latter is highly recommended.

also, joining the VMD biocore project is highly recommended in this
case, since this will give you more direct contact with the developers,
via the biocore chat feature and the option to exchange files easily
via the biofs.

let us know if you need more help with this.

cheers,
   axel.

HB> > HB> Hello, I am student working at NIST using the VMD. Several members of
HB> > HB> the lab really want a python module for the VMD, and have asked me to
HB> >
HB> > it would be useful to learn what kind of application
HB> > would be needing this. there might be alternatives.
HB> >
HB> > cheers,
HB> > axel.
HB> >
HB> > HB> try and build one. I noticed that their has been some work on such a
HB> > HB> module and I would really like to see it if at all possible.
HB> > HB>
HB> > HB> Harry Bullen
HB> > HB> NIST
HB> > HB>
HB> >
HB> >
HB>
HB> 

-- 
=======================================================================
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.