On Tue, 20 Sep 2005, Michael Sperber wrote:

> Richard Kelsey <xxxxxx@s48.org> writes:

>> Actually, it does give an order.  "Parent init expressions, if any,
>> are evaluated before child init expressions."

Actually, the draft says this in the section on INIT! (with the bang), where it
seems fine.

It is quiet on the field initializations.  For the latter, if the current
draft were to stand, I would prefer that it remain unspecified.  In the most
obvious functional approach, the child fields are more naturally evaluated
first.

Cheers
Andre