>>>>> "Wolfgang" == Wolfgang Jenkner <wjenkner@inode.at> writes:
Wolfgang> Anyway, in the case where CMUCL gives the warning could you please
Wolfgang> check if the generated lisp code is essentially the same as the code
Wolfgang> generated by Clisp? (Sorry for asking you to do this but I have only
Wolfgang> Clisp installed.)
They appear to be identical. In any case, here is the generated code:
(PROGN
(DEFPROP |$harm_foo| T TRANSLATED)
(ADD2LNC '|$harm_foo| $PROPS)
(DEFMTRFUN (|$harm_foo| $ANY MDEFINE NIL NIL) (|$l| |$u|)
(DECLARE (SPECIAL |$u| |$l|) (FIXNUM |$u| |$l|))
((LAMBDA ()
NIL
NIL
((LAMBDA (|tr-gensym~1| |tr-gensym~2|)
(COND
((NOT (< |tr-gensym~2| |tr-gensym~1|))
(DO ((|$i| |tr-gensym~1| (F1+ |$i|))
(|tr-gensym~0| 0 (ADD* |tr-gensym~0| (RREMAINDER 1 |$i|))))
((< |tr-gensym~2| |$i|) |tr-gensym~0|)))
((= |tr-gensym~2| (F1- |tr-gensym~1|)) 0)
(T (INTERVAL-ERROR '$SUM |tr-gensym~1| |tr-gensym~2|))))
|$l| |$u|)))))
When I load this up and run it, I don't get the warning message about
$i anymore. An mistake on my part, I guess. Sorry about that.
Ray