> It doesn't look like any PyLucene function is on the stack 
> (except for the thread start). This could mean that memory got trashed and 
> that, later (now), python itself is crashing or it could mean that this crash 
> is unrelated to PyLucene.

Sure, it could mean that.  But this application has been running
stably (no crashes) for two years now, and on PyLucene 1.9 for six
months (no crashes).  Upgrading to PyLucene 2.0.0 at the beginning of
June has produced four crashes in 10 days, each time apparently in a
thread which is running a search via PyLucene.  I'd suspect this new
hand-written code you've got in there.  I'll see if I can find out more.

Bill
_______________________________________________
pylucene-dev mailing list
[email protected]
http://lists.osafoundation.org/mailman/listinfo/pylucene-dev

Reply via email to