segfault in object generation if nservers > nobjects
The maximum number of objects generated by the python code is set as:
py_max_objects = max_objs-num_servers;
This causes problems if the num_servers is > max_objs (see new input-generator utility or old triton-fault-sim).
We can put in a check to prevent this, but why is num_servers subtracted from the max_objs in the first place?