VMD-L Mailing List

From: Axel Kohlmeyer (akohlmey_at_cmm.chem.upenn.edu)
Date: Fri May 25 2007 - 17:00:20 CDT

On Fri, 25 May 2007, Yinglong Miao wrote:

YM> Dear all,

dear long,

YM> I have been haunted by a problem with installing VMD on a LINUX
YM> cluster, BigRed, at IU. BigRed consists of 512 IBM JS21 Blades, each
YM> with two dual-core PowerPC processors and runs SuSE Linux Enterprise
YM> Server operating system (see
YM> http://uits.iu.edu/scripts/ose.cgi?aueo.def.help).
YM>
YM> I was told that BigRed doesn't support OpenGL and even has issues with
YM> X. And VMD cannot be installed with GUI. But I want to use VMD for

the big question here is how to get a binary at all.
the last precompiled binary for Linux on PPC is a version
1.7.1 and dates to 2001. you probably don't want to use that,
so you'll have to compile VMD yourself. in principle, it
should be possible to compile VMD without X/OpenGL support,
but it is probably simpler to compile it including X (and
potentially using Mesa instead of the native OpenGL)
anyways, since there are be some plugins, that support
both textmode and graphical mode and may not be loadable
without X and Tk.

since you probably have to build supporting libraries and binaries
as well, it will need some experience in compiling diverse and
not always extensively documented software...

how about transferring your trajectories to a machine with
a supported platform and doing the analysis there?

YM> numerical analyses of NAMD simulation outputs by making use of VMD
YM> plugins. Also for my current NAMD simulations, I need to call TCL
YM> scripts which need to be interpreted by VMD with its plugins (some
YM> basic functions handling dcd, pdb files, atom selection , etc.). So I
YM> want to ask is there a way to have VMD installed as a TCL interpreter
YM> executed with "vmd -dispdev text -e myscript.tcl" without worrying
YM> about the graphics support?

if you use a VMD binary that has been compiled with X support
and start it with -dispdev text, it will need the X libraries,
that are dynamically linked to be present, but will not use them
(with linux there are ways to fake libraries, if needed).

YM> There were some discussions in previous posts about "vmd can't start up
YM> after being installed in linux" (lanhua Tue Apr 17 2007), but I didn't
YM> find my answers there. Do you have any more ideas?

in 9 out of 10 cases, those reports are referring to the
problem of libstdc++.so.5 not being installed on 32-bit x86 linux.
(which is a side effect of using redhat 9 to build x86 linux VMD
so the binaries will be compatible to almost all current x86 linux
machines).

cheers,
   axel.

YM>
YM> Best regards,
YM> Long
YM>
YM>
YM> --
YM> Yinglong Miao
YM> Ph.D. Candidate
YM> Center for Cell and Virus Theory
YM> Department of Chemistry, C203A
YM> Indiana University
YM> 800 E. Kirkwood Ave.
YM> Bloomington, Indiana 47405
YM> 1-812-856-0981 (office)
YM> 1-812-857-6205 (home)
YM> http://ylmiao.dict.cn/mypage/
YM> 

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