> Richard, can you say what failed on the OS X 10.4 (Tiger) buildbot?

There seems to be a problem which depends on the order in which you run 
the test, and it happens on Linux also.  For example if I do

   ./python -m test -v \
       test_multiprocessing_fork \
       test_multiprocessing_forkserver

Then I get lots of failures when forkserver runs.  I have tracked down 
the changeset which caused the problem, but I have not had time to look 
in to it.

 > The only vaguely suspicious message when running with -v was:
 > [...]
 > [semaphore_tracker] '/mp18203-0': [Errno 22] Invalid argument
 > [semaphore_tracker] '/mp18203-1': successfully unlinked
 > [...]

That is expected and it shows the semaphore tracker is working as 
expected.  Maybe I should print a note to stderr to expect this.