On Monday 28 April 2003 15:18, Paul F. Dietz wrote:
> Running the gcl ansi-tests on ecl cvs, I get a catastrophic failure (se=
> fault) in the defclass tests. The problem appears to be with
> SLOT-EXISTS-P, which is returning some weird object:
> > (defclass foo () ())
> > (slot-exists-p (make-instance 'foo) 'b)
Fixed. Ancient bug -- comes from old EcoLisp.
I see that you have been working on the test suite a lot. Bad news. I had=
a copy of ECL to run with only 70 bugs, and now they become 450 :-)
Anyway, I also submitted fixes for the (SETF fname) issue: now the compil=
and the interpreter should be able to recognize all occurrences of (SETF=20
fname) function names. Furthermore, I changed the way in which (SETF fnam=
definitions are stored: formerly, it used to be an uninterned symbol; now=
is a system property (GET/PUT-SYSPROP).