Greetings!  Thought it wise to post a short note on the multiple value
type meaning used by the compiler at the moment.  As Bob had noted
earlier, '(values t t) probably gives no information at all about the
function according to the ansi spec.  For this reason, we are handling
values types separatly outside of subtypep and inside the compiler so
as not to run afoul of the spec, but with a semantics more useful to
the compiler.  Namely, (values t) == t, (values) == nil, (values x y
z) == (values x y z null), (type-and '(values integer keyword)
'(values fixnum symbol t)) == '(values fixnum keyword), (type-or
'(values integer keyword) '(values fixnum symbol character)) ==
'(values integer symbol (or null character)).  In other words, these
follow what a multiple value binding or setq could expect for a
variable assignment.

BTW, if anyone is experiencing an occasional segfault on calls to
random, please let me know, especially if they can be reliably
reproduced.  I need to post a note to gmp-devel about this.  This
should pertain to the latest gmp library only.

Take care,
-- 
Camm Maguire                                            [EMAIL PROTECTED]
==========================================================================
"The earth is but one country, and mankind its citizens."  --  Baha'u'llah


_______________________________________________
Gcl-devel mailing list
Gcl-devel@gnu.org
http://lists.gnu.org/mailman/listinfo/gcl-devel

Reply via email to