On Wed, 2013-02-20 at 13:32 +0000, Rupert Swarbrick wrote:
. . .
> The thing is, I've been reading sections 3.1 and 3.2 of the hyperspec
> quite closely and I can't find anything where it says that an
> implementation can assume a function returns a single value until told
> otherwise (which seems to be what's happening).
IIUC a function call returns ALL the values returned by the last form in
its body; that's what I understand an implicit progn to be. It seems to
be the responsibility of the caller to handle multiple return values
with one of the operators that handle multiple values. Section 3.1.7
(Return Values) of the hyperspec seems pretty clear on this. So it
seems to me that an implementation cannot make such an assumption.
Going deeper into this will require a REAL expert, not me :-) If you go
off-line with this please keep me in the loop; I'm interested.
--
Bill Wood