Some thoughts on the 'reveal' operator

[Dan Bornstein]

Bill Frantz:
>I like this solution the best.  I would like to figure out a way to only
>have to code reveal at the highest level, while still retaining the
>"explicit revelation only" feature.  If we allow:
>
>    define factorial(n) {
>        reveal if (n <= 1) {
>            1
>        } else {
>            n * factorial(n-1)
>        }
>    }
>
>To mean reveal the value of the if expression, we are there.  What kind of
>trouble does considering the if to be an expression with a value cause?
>What other problems does the introduction of this pattern cause?  The
>Algol68 experience seems relevant here.

I think the problem with this is that when the example is expanded into
non-trivial code, the fact that the result of the if as a thing being
revealed may be lost:

   reveal if (...) {
      ... a page of code here ...
      1
   } else {
      ... a page of code here ...
      <inadvertent-dangerous-capability-revealing-expression>
   }

-dan