Currently working on the above that results in what I hope will be a utility for zero memcpy of invariant data grams.

Until I get my JIRA-jitsu back on, I'll list where I am at:

TODO/DOING:

• binary API (currently giving me negative energy)
• harden zero memcpy portion
• optimisation of "to zero memcpy or actually memcpy, that is the question!" based on packet length (simple decision on what is fastest)
• generalise into a (set) of utilities that others can use (I need 5 "copies" this invariant piece)
  COMPLETE
• zero memcpy works but not robust