> I am thinking that a non-broken approach would involve (1) treating
> a domain as binary-compatible with its base type, and therefore with
> all other domains on the same base type, and (2) allowing a coercion
> function that produces the base type to be used to produce the
domain
> type. (The patch tries to do (2), but does it in the wrong places,
> leading to the mislabeled-expression problem.)
2 was the goal, and it worked enough for any default expression I
could come up with -- so I thought it did pretty good. Guess not. It
should be binary equivelent to the base type it's made out of.