Re: [Sbcl-devel] CLOS possible bug

[Attila L. <att...@gm...>]

>
> Could you please try with no init files?  If that works, could you
> please send me the contents of your user and system init files?


uh?!

just now i tried to reproduce it and i couldn't... not with my core file,
nor with my standard init files, nor with  "sbcl --sysinit /dev/null
--userinit /dev/null"

then i tried it from inside slime, our usual dev environment, and it was
working there, too for a one or two sbcl restarts. and after it failed for
the first time in emacs/slime, i could reproduce it with the noinit from
shell. i don't know sbcl internals, but this is quite strange for me!

could it be that based on the loaded dll's or something, sbcl heap is loade=
d
at a different address and there's some identity hash involved in this bug?
due to a missing hash function identity hash is used? but i'd better stop
making wild guesses... :)

and the backtrace is: http://paste.lisp.org/display/19188

hope it helps,

- attila

(alias 101 on irc &no 'its not lisp code :)