| Detailed description |
|
|
Excuse me for asking and adding to your workload, but I do really think this would be a useful improvement.
A rather troublesome problem now is that if anything goes wrong with an MD run at any point in the pipeline, one has no option but to start over again, possibly fixing -if possible- the original PDB file.
In addition, currently Zephyr cannot be detached from the underlying computation.
If an option to save as a script all the commands and configuration files that are going to be used in a run was available, then:
- it would be possible to launch the script on a remote computation machine (e.g. accessing the same directory over NFS or SSHFS)
- it would be possible to study and learn how to perform the computation, and even further tailor it to adapt to special needs
- it would be possible to fix a problem at intermediate steps and then continue from where it failed.
- it would be possible to have e.g. an execution tab where the user could enter a remote computer, username and password, and then the data and script could be copied over by ssh and run there on a shared folder, looking for the output in the local path of the shared folder. Or the execution be monitored by ssh and copied back when finished. |
|
