cpu_free was not set correctly due usage of two different keywords for the same: cpu_free and cpu_free_min.

forgot to remove all references to log file writing with print_both, give logger a file handler, not a path

when logging to do not print to stdout, press enter to continueif not logging to file, also add the cpu_free_min option to the argument parser

when the user does not have any jobs, the pbswrap.parse_qstat will not contain the current users uid. Fail gracously and set cpu_user count to zero instead

this is embaressing, but by developing on the laptop repo and then pushing to the remote, and only then pulling from the cluster, this kind of too verbose work flows are created...