as reported previously by another submitter in June 2002:
It appears that the previously reported bug was not
adequately fixed despite several new releases since
then.
middleman crashes about every 10-20 requests.
there are no documented debug options to diagnose the
reasons the daeman is quitting unexpectedly.
I had high hopes for this program. But given this
behavior of the daemon program crashing after a such a
short period of service, middle-man is pretty useless to
me. Until this is problem is either fixed or some means
of helping to determine why the program crashes is
provided, middle-man is unusable.
Logged In: YES
user_id=432781
Well, I'm sorry to hear it doesn't meet your expectations.
Keep in mind I am the sole developer behind this project,
and it's not always possible for me to test it as thoroughly
as I'd like before each release. I depend on people like you
to provide thorough and accurate bug reports to help me
diagnose problems. This problem you're having is almost
certainly unreleated to the one the previous bug submitter
was having, quite a few crash bugs have been identified and
fixed since then. Could you recompile middleman using the
--enable-debug option in the configure script, type ulimit
-c unlimited in your shell, run it until it crashes, and
then email me the compiled binary and core file so I could
diagnose the problem better ? Also, the last few lines of
the log file may be helpful. If you feel daring, you could
grab the latest CVS which fixes several race conditions in
the caching code and a few other things.