Let me start my saying that I know Guido is giving the following issue keen attention, and my understanding is that the issue will go away in 2.3. Will be a tremendous help to PyGeo. But its worth noting, that in the real world, as to PyGeo, the IDLE situation has *considerably* complicated life . Particularly since PyGeo *needs* the forked version, rather than the version distributed with Python2.2. That is if a user chooses to use IDLE. And that would be the most natural and sensible choice for a true newbie. A look into the IDLE fork mailing list about a week ago seemed to indicate that the forked version is now encountering some problems running VPython scripts, and will therefore the same would be true as to PyGeo. I am sure this is being attended to. I guess I will be more proactive and download the current IDLE fork version (or first see if it is part of 2.3a) and do some testing of PyGeo scripts with it. Results to be reported on the IDLE fork list, not here. Art