expand explanation of job state options

instead of location functions in sacct, squeue, and scancel

    (e.g. 8:00 or 08:00:00 or 8:00:00 or 08:00:00).

fix for better debugging when trying to start a new cluster in accounting that hasn't been added yet

add -n description to slurmd and slurmctld help message
fix support so that nice value of -2 is valid (-2 == NO_VAL).

    "scontrol update JobId=# NumNodes=#" or
    "scontrol update JobId=# NodeList=<names>". Subsequent job steps must
    explicitly specify an appropriate node count to work properly.

    daemon's nice value.

which gets group members

    See "man slurm.conf" for details about how these control when slurmctld
    updates its information of which users are in the groups allowed to use
    partitions.

ESLURM_REQUESTED_NODE_CONFIG_UNAVAILABLE when slurmctld has restarted and 
nodes have not yet registered (in unknown state).