Nikodemus Siivola writes: > I'm under the impression that 2.4 kernels in general have threading > problems, and fixing them is a lower priority eg. threadsafetyfing > sbcl internals, and that for serious sb-thread use 2.6 kernels are > recommended, but dunno. In this context, the tests in threads.impure.lisp involve "serious sb-thread use" and so are not really expected to work too well in 2.4 -dan -- "please make sure that the person is your friend before you confirm"