🚀 go-pugleaf

	This is probably a naive question, but..

	When designing cryptosystems I'd imagine it's desireable to use
algorithms that don't parallelize well. Is it feasible to somehow add
a step to a process like exponentiation that won't decompose well? Lenstra's
experiment with using loads of machines to factor primes shows us that
massively parallel attacks are just a matter of networking. ;) Is there
anything like factorization that also has the property of not parallelizing
well or at all?

mjr.