From: John Duff (jfduff_at_mtu.edu)
Date: Tue Apr 18 2000 - 17:35:03 CDT

Hello,
 
I'm a CS grad student at Michigan Tech University, and am having a problem
getting VMD to connect properly to a running NAMD simulation on Linux.
(on Sun/Sparc systems, I can do it no problem.) Here's the situation:

Setup: RedHat Linux 6.2 (latest versions of everything) on a
       Dell Inspiron 7000 laptop (366 MHz Pentium II) 128M RAM

Description: I downloaded and installed NAMD_2.1_Linux-i686 and
             VMD 1.4 (vmd-1.4.bin.LINUX.mesa.tar.gz). For VMD, I ran the
             configure script (after changing the install directories, nothing
              major) and then did a 'make install', as per the README file
              instructions. VMD runs fine on its own, I can load the sample
              alanin molecule and rotate/animate/see it no problem. Similarly,
              I can run NAMD just fine using the provided sample alanin
              files. So, running them each separately works beautifully.
              But, when I tried to attach VMD to a running NAMD simulation,
              (both running on the same machine, my laptop), VMD gives a
              warning message in the console popup window about
              "Connecting to opposite-endian machine", then the alanin molecule
             I had already loaded into VMD (as per the directions, and I can
             rotate/animate/see it no problem) basically explodes in the
             viewport (sounds over dramatic, but best way to describe it)
             and then VMD coredumps and NAMD also stops. I have no
              problem doing this (attaching VMD to running NAMD sim) on a
              Sun Ultra 5, so I know that my NAMD alanin file is set up
              correctly, and I'm following the directions correctly as well.
              The error message about opposite-endianness seems to be a clue
              to me, because that shouldn't be if both are running on the
              same machine, obviously. Is this all due to some known issue?
              I looked over the email archives, but didn't see
              anything useful (maybe I missed it?). I noticed some postings
              about having to use older versions of glibc in older versions
              of VMD, but assumed that since VMD runs fine on its own for me
              that that was not the problem.
 
 
Any help or pointers anyone could offer would be greatly appreciated. If you
need any more info from me, please let me know.

Thanks in advance,
John Duff
jfduff_at_mtu.edu