Subject: Re: CMUCL18d on Alpha?
From: Erik Naggum <>
Date: Tue, 23 Apr 2002 17:36:33 GMT
Newsgroups: comp.lang.lisp
Message-ID: <>

* Raymond Toy
| I do think it is a small problem in lisp where 
|    (= (1+ double-float-epsilon) 1)
| may or may not be T depending on whether there was a store between the
| comparison or not.

  Hm.  Neither Allegro CL, CMUCL, or CLISP manage to distinguish 1d0 from
  (1+ double-float-epsilon), but with an identical epsilon, LispWorks does.

  (scale-float double-float-epsilon (float-precision double-float-epsilon))
  should in theory be identical to (1+ double-float-epsilon), but couriously,
  (1- (scale-float ...)) yields a value twice as big as the epsilon.  This
  is really weird."
  In a fight against something, the fight has value, victory has none.
  In a fight for something, the fight is a loss, victory merely relief.

  Post with compassion: