From: Jérôme Hénin (jhenin_at_ifr88.cnrs-mrs.fr)
Date: Thu Jan 07 2010 - 04:28:09 CST
The code assumes that files from previous runs are named according to
the pattern <prefix>.<ext>. The colvars module itself writes state
files with the colvars.state extensions, but other components (e.g.
the ABF bias) also write files using different extensions. So the
convention in the current code allows it to find all necessary restart
files from a given run, based on a single prefix.
The downside is, all files have to conform to that pattern. In your
case, the main reason I can see why you would need the backup file is
if the last written state file is corrupted or otherwise unusable. For
now, there is no option but to rename the file to conform to the
expectations of the code.
If you have a suggestion on how to better handle this, we'll be glad
to discuss it.
2010/1/7 BIN ZHANG <zhngbn_at_gmail.com>:
> Dear all:
> I found a strange behavior of the keyword "colvarsInput" in NAMD 2.7b. Say,
> I specify the following line in the configure file:
> colvarsInput ../lg08/output/npt_equil.restart.colvars.state.old
> then, the code would actually strip of "colvars.state", and append it at the
> end of the string. That said, the final effect would be like I provided the
> following line:
> colvarsInput ../lg08/output/npt_equil.restart.old.colvars.state
> Of course, NAMD won't start because I don't have that file.
> Can someone explain to me why would the code do this? Or what's the
> advantage of doing this?
This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 15:53:40 CST