[gmx-users] Transition difficulties: version 3.3.3 to 4.0.3 regarding pull_geometry=distance

[chris.neale at utoronto.ca]

Hi Steve,

I also use the pull code and would be very interested to more fully  
understand the problem that you are experiencing. However, I don't  
generally modify any code myself, and therefore have no use for your  
private system. What I would be eager to receive is your  
POPC-buckyball system with a small instruction on how I might detect  
the difference between 4.0.3 and 4.0.99 behaviour.

chris.neale |at| utoronto.ca

Chris.

-- original message --

Hi,

The permeants in my bilayer system are improperly pulled in calculations
using all versions of Gromacs 4.0.x including the VERSION
4.0.99_development_20090120.  Since, due to restrictions, I can not
provide the topology file for my system to bugzilla, I constructed a
similar prototypical system using Prof. Tieleman's POPC bilayer
coordinates and field parameters.  A buckyball permenat doped in the
center of this system did indeed react similarly in a 4.0.3 calculation
with an improper pull away from the equilibrated 3.3.1 constraint force
coordinates.  It was heartening to see that this problem was fixed for
that system, by using VERSION 4.0.99_development_20090120.
Unfortunately, again the problem remained for my actual system, even
with use of the newest version of code.

Since the cylinder option is now operational, I have a viable
work-around for this problem.  I am quite appreciative of the attention
that this issue has received.  If there is continued interest, and a
willingness for an engineer to receive the topology file in a less
public forum than bugzilla, I believe we could proceed with the
debugging process.

Thank you,

Steve Fiedler