[Python-Dev] Release candidate followed by feature freeze?

Jeremy Hylton jeremy@beopen.com
Tue, 3 Oct 2000 11:37:21 -0400 (EDT)

>>>>> "NS" == Neil Schemenauer <nas@arctrix.com> writes:

  [GvR writes:]
  >> Anyway, I'm not yet ready to disable GC by default -- that's
  >> sensible but also defeats the purpose...

  NS> Okay.  I think I've found a possible problem.  When I set
  NS> threshold0 to 1 in the C code, weird things start happening.
  NS> I'm looking into it.

I agree that it isn't too useful to ship the garbage collector, but
have it turned off.  On the other hand, we're going to get a lot of
unhappy users if GC bugs cause inexplicable crashes in large
programs.  It would be hard to sell Python for important applications,
e.g. long-running server processes, if there was a small possibility
that the program could crash through no fault of its own.