[Python-Dev] New thread death in test_bsddb3

Barry Warsaw barry@python.org
22 Apr 2003 22:33:21 -0400

On Tue, 2003-04-22 at 18:27, Mark Hammond wrote:

> Yes, this is my fault.  The assertion is detecting the fact that bsddb is
> creating and using its own interpreter/thread states than using the
> thread-state already seen for that thread.
> As Tim says, the assertion is new, but the check it makes is valid.  I
> believe that removing the assertion would allow it to work, but the right
> thing to do is fix bsddb to use the new PyGILState_ API, and therefore share
> the threadstate with the rest of Python.
> I will do this very shortly (ie, within a couple of hours)

Thanks for taking care of this Mark!  Yes, as PEP 291 states, bsddb.c
has to be compatible with Python 2.1.  At some point we may want to
re-evaluate that, but for now, if it's easy to do, we should keep