dear ganesh,
please always keep the list in cc. thanks.
Dear Axel,
Thank you for the response. I am using VMD-v.1.8.7, which I thought was the latest (it was downloaded 4-months back for a MAC). I was able to read POSCAR and generate a LAMMPS-trj. file from VMD before. But the problem lies when I try to read this lmmps.trj file back in VMD and look at periodic images of the cell/atoms. They don't agree with the original POSCAR, but that is because the lmmps.trj file does not have any information about the angles between the lattice vectors (i.e. no xy,xz,yz values)), so it creates images by translating along the orthogonal (x,y,z) instead of translating along (a,b,c) directions. I am afraid lammps might do the same. So I was trying to enter the cell directly into lammps.
support for skewed cells in the lammpstrj plugin was only added
to vmd after the 1.8.7 release.
you have to use an alpha test version to have that working.
if you load the coordinates from a poscar file and the system cell is not
aligned in a specific way, the poscar plugin will rotate them (and the cell)
to match VMD's expectations.
in order to write a data file from that (and not a lammps trjajectory) file,
you need the topotools plugin. you can keep using vmd-1.8.7 for that,
but you have to update the topotools plugin with the latest snapshot
which will be able to read and write tilted cell systems.
Perhaps a more recent VMD does this properly. I will try that.
The crash happens when I feed lammps trajectory-output from my skewed-cell input with {xy,xz,yz} data (which is probably wrong)
again, it would be easier to see what is going, if i had a sample,
but for now i will assume that you have run into a bug that has
already been fixed.
axel.