[onetimePete]

The real trouble is the inability of chips to identify what haskell could completely avoid, doing the same algorithm for the same input repeatedly. Hashing over input skip and save output, that would be a optimization. Also Tools that give Programmers a DataOrientated-VisualFeedback of the OO-Cluttercosts they impart on there work.

A flag that guarantees side effect freedom to a set of operations and suboperations, for the processor that would be great.

[paulddraper]

Hashing is expensive.

[thrownaway2424]

Not as expensive as a cache miss.