Unexpected restart behaviour when restart files are copied #668
Description
Summary
Sorry to keep on opening small issues, but I have noticed another slightly odd behaviour that caught me out:
When restarting MD calculations, the decision of which restart file to use is made on the basis of the modification times of the restart files:
This means that when I copied an MD calculation to a new directory and tried to use a restart, it picked a different restart than I expected as the copy operation had randomised the modification times.
I suppose that this is not so common a scenario, but it was a bit confusing. I guess this one is a bit of a mater of taste, but would it be more robust to use the timesteps in the filenames to pick the best restart file to use?
Platform
Linux 5.14.0-611.27.1.el9_7.x86_64 x86_64 GNU/Linux
Version
0.8.6
Python version
No response