steps rather than cancelling the job. This now matches the behavior of
    all other signals. "scancel <jobid>" still cancels the job and all steps.

determine actual minimum mem size and proc count before 
trying to use them. 
reorder tests in sbatch output to match actual output order
 (regular expression tests are done strictly in order (at least 
 for some versions of expect).

checkpoint as required, still need to perform actual checkpoint 
operation

update the perl api to mirror what the slurm.h file looks like now.  Haven't checked to make sure we don't have any more new functions in the api itself we need to look at.

fix bad data type for read call (ssize_t instead of size_t).
fix race condition between setsid and setpgid