On Sat, Nov 23, 2002 at 11:43:45AM +0000, Christophe Rhodes wrote:
> Valtteri Vuorik pointed me to another nasty compiler bug in current CVS:
> * (defun u () (member (char "foo" 0) '(#\. #\/) :test #'char=))
> -> TYPE-ERROR "NIL is not of type REAL".
> The attached patch is a band-aid which seems to cure the problem, but
> I'm slightly uncertain as to what the type algebra for function types is
> meant to be doing, so it's probably worth checking for sanity.
I haven't played with my new CVS updated version yet, but seeing the fix
from Alexey I gather this is no longer a showstopper.
and also (in another post)
> PS: I have some more LOOP fixes in my tree; also, things that I think
> probably ought to go in before release are Matthew Danish's PEEK-CHAR
> patch (version 2) and the highly embarrassing "sbcl --core" Segmentation
> fault bug... or have these been left too close to the wire?
I notice you merged the "sbcl --core" bug since then, thanks.
As far as I could see the PEEK-CHAR version 2 patch is still
outstanding. I'll probably deal with that later today.
AFAIK, after the PEEK-CHAR patch there's nothing else planned with
high fixingness/complexity ratio for 0.7.10. But since this time we've
done more last-minute patching than usual this time, I think I'll push
the release back a few days, probably to around Wednesday, to give
other things a chance to come up.
William Harold Newman <william.newman@...>
"If you can't remember what mnemonic means, you've got a problem."
-- Wall, Christiansen, and Schwartz, _Programming Perl_, 2d edn., p. 548.
PGP key fingerprint 85 CE 1C BA 79 8D 51 8C B9 25 FB EE E0 C3 E5 7C