
The following bugs have been fixed:

Feb 28,  1992: There was a  bug  on line 141 of   eunify.lsp.  (- i 1)
should have been  (+ i 1).  Because  of the bug, Ken  Mano at NTT labs
was  able  to derive s(0) = p(0)  for an axiomatization of arithmetic.
I'm amazed this didn't cause serious problems  and that nobody noticed
it  for two years.  I've  rerun a number of problems  with the correct
code, and haven't noticed any other problems  that seemed to have been
affected by the bug.  (It only would have affected problems that cause
permutative unification to be   activated.)  I've changed the  version
number and date in the start-up banner (v1.0.1, February 1992.)

May 28, 1989:
There were cases when the gen-permutations function would 
not generate all possible permutations (when the arity of the
permutative symbol was greater than or equal to three).  This resulted in
an incomplete unification algorithm, etc.  This has been fixed.


On a Symbolics during a run of the group1 example, one of the queue
routines complains about NIL being where a fixnum should be.


******************************************************************
To date, the following bugs have been reported:



On a port to the Apollo workstation, the system loads and compiles
(with a number of minor warnings about unused variables and 
the KCL-specific OBJECT declaration), but runs much more slowly
than it should.


When compiling the file queues.lsp using plain KCL (not AKCL) dated
Sept. 1986 and 1987 and the Sun 3/2*0 C compiler (under SunOS
3.2 and 3.5), the C compiler
reports a syntax error.  Use of the GNU C compiler (gcc 1.34) 
solved the problem.  (I have not noticed this problem using
AKCL and the Sun compiler on rascal, a Sun 3/280.)





