[Openmcl-devel] deftype bug?
nikodemus at random-state.net
Thu Mar 18 15:58:25 UTC 2010
On 18 March 2010 16:11, Paul Krueger <plkrueger at comcast.net> wrote:
> also include a "satisfies" type. I'm curious if anyone knows why the
> serial semantics that were so clearly defined in CLtL2 were not
> adopted for the standard; oversight or choice?
At a guess, because it would mean that set-theoretical notion of types
goes out of the window. Consider the type
(or (and x y) (and z y))
this is -- in ANSI Common Lisp -- equivalent to both types
(and (or x z) y)
(and y (or x z))
but under CLtL2 is certainly not equivalent to the last one. Consider
further what should should
(subtypep '(and (or x z) y) '(and y (or x z)))
return? In ANSI CL the answer is a clear and resounding T, T. CLtL2 I
have no clue about.
More information about the Openmcl-devel