gcl-devel
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Gcl-devel] Re: on incf


From: Paul F. Dietz
Subject: Re: [Gcl-devel] Re: on incf
Date: Thu, 11 Aug 2005 22:47:34 -0500
User-agent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.7.10) Gecko/20050716

Camm Maguire wrote:
Greetings!
test-random-types3 spit this out after a 10000 10 run:

(((AND (OR (OR (NOT (AND NIL RATIO NIL))) (COMPLEX (INTEGER * 12800))))
  (NOT (CONS (CONS (CONS T T)
                   (CONS (COMPLEX (REAL 78460 78460)) (EQL -3905)))
             (EQL -285003020)))))

do I understand right that this means the tester thinks

(subtypep '(AND (OR (OR (NOT (AND NIL RATIO NIL))) (COMPLEX (INTEGER * 12800))))
' (NOT (CONS (CONS (CONS T T)
                   (CONS (COMPLEX (REAL 78460 78460)) (EQL -3905)))
             (EQL -285003020))))

is wrong?  We're returning nil t at present, which looks right here.

Take care,

Actually, there's a problem with that -- it should be producing a triple
of types, not a pair (I've checked in a fix).  Call these types t1, t2,
and t3.  The meaning is:

(1) If we determined that t1 is a subtypep of t2, then check (using
  various identities) that t1 is a subtype of (or t2 t3) and that
  (and t1 t3) is a subtype of t2.

(2) If we can determine that t1 is not a subtype of t2, then
   check that (or t1 t3) is not a subtype of t2, and that t1
   is not a subtype of (and t2 t3).

        Paul




reply via email to

[Prev in Thread] Current Thread [Next in Thread]