[Python-Dev] Berkeley DB

Fred L. Drake, Jr. fdrake@beopen.com
Tue, 25 Jul 2000 12:29:02 -0400 (EDT)

Andrew Kuchling writes:
 > It's not; in fact, db_wrap.c is SWIG-generated code while the old
 > bsddb module was a hand-written extension.  Is it OK to add a
 > SWIG-generated module to the core?  (I have an irrational dislike
 > SWIG's trick of encoding pointers as strings, because it provides a
 > way to deliberately engineer core dumps; it's just sort of
 > unsettling.)

  You're right -- SWIG code stinks for exactly this reason.  I think
Jeremy did some work to avoid this, but I imagine that's no longer
available (it's at CNRI).
  It seems to me that the right thing is to fix SWIG (probably as part
of the next generation of the tool).  I'm not sure what Dave's current
plans are, or if there's any expectation for the schedule.  I don't
see anything relevant at www.swig.org, but I might be missing
something obvious to SWIG users.


Fred L. Drake, Jr.  <fdrake at beopen.com>
BeOpen PythonLabs Team Member