[tpm] Complexity and language

[arocker at Vex.Net]

> But seriously, I'm talking about more than 'field validation' but
> algorithmic equivalence (or non-equivalence... where a+b != b+a)
> [Or would that be 'temporal significance'?]
>

That's more domain-specific. Once you've got sensible input, (for whatever
is your definition of "sensible"), do they map to the output you want?

In your example, the possible sets would presumably be "  " (nothing
happens), "a ", " a", "b ", " b", "ab", "ba", and perhaps multiple
occurrences of one event boiling down to a single event, (e.g. "bbba" =
"ba"). Then the problem becomes writing a test generator that is less
complex than the thing it's testing. (That way lies recursion.) Jigs
should not cost more than the product, unless they're going to get
re-used.

I think we're wandering off-topic a bit here; I was trying to determine
what sort of problems demand more complex structures and processes than
simple lists and basic logic and arithmetic operations, and how to order
the increasing demands.