-2. One of the
-[[criteria we established for classifying reduction strategies|
-topics/week3_evaluation_order]]
-strategies is whether they reduce subexpressions hidden under lambdas.
-That is, for a term like `(\x y. x z) (\x. x)`, do we reduce to
-`\y.(\x.x) z` and stop, or do we reduce further to `\y.z`? Explain
-what the corresponding question would be for CL. Using either the
-OCaml CL evaluator or the Haskell evaluator developed in the wiki
-notes, prove that the evaluator does reduce expressions inside of
-"functional" CL expressions. Then provide a modified evaluator that
-does not perform reductions in those positions.
-
-<!-- just add early no-op cases for Ka and Sab -->
-
-3. In the previous homework, one of the techniques for controlling
-evaluation order was wrapping expressions in a `let`: `let x = blah in
-foo`, you could be sure that `blah` would be evaluated by the time the
-interpreter considered `foo` (unless you did some fancy footwork with
-thunks). That suggests the following way to try to arrive at eager
-evaluation in our Haskell evaluator for CL:
-
- reduce4 t = case t of
- I -> I
- K -> K
- S -> S
- FA a b ->
- let b' = reduce4 b in
- let a' = reduce4 a in
- let t' = FA a' b' in
- if (is_redex t') then reduce4 (reduce_one_step t')
- else t'
-
-Will this work? That is, will `reduce4 (FA (FA K I) skomega)` go into
-an infinite loop? Run the code to find out, if you must, but write
-down your guess (and your rationale) first.
-
-<!-- Doesn't work: infinite loop. -->
-