E-order can be optimized out of CapTP for DeepFrozen objects

(Please keep discussion mainly on e-lang list for consistency.)

A novel observation that was made at last Friam while reviewing Cap'n Proto. In E terms:


If an object is known to be DeepFrozen, and therefore stateless, then it is not necessary to delay third-party messages to that object in the fashion required by E-order, as that requirement is to prevent mis-orderings of stateful operations.


I don't currently recall enough of the CapTP design to say how hard this alternate behavior would be to implement. It would also require holders of far references to have a remote-is-DeepFrozen flag, which is not especially difficult but does not fit neatly into the target-object-agnostic existing protocol.

-- 
Kevin Reid                                  <http://switchb.org/kpreid/>