great presentation! it adds more clarity to samples provided around hpc. question - if one invokes execution of some code on a particular node through this method, can that code call another operation (or the same) hosted in the same 'hpc-service' from which it was originally invoked? does this pseudo-recursive scenario pose any known problems (besides the need to accordingly structure ones code) when (in my shot at the plain vanilla non-wcf HPC equivalent) a node executing a task tries to start a new job? - thanks!