[Python-Dev] GC and ExtensionClass - a summary of the problem and a workaround

[James Henstridge]

On Thu, 17 May 2001, Guido van Rossum wrote:

> My other point is that Skip's problem is clearly a gtk bug: it
> shouldn't have exposed the type before fully initializing it.

On further investigation, it turned out that it was caused by a bug in my
code generator that caused one extension class to be initialised before
its base class (in fact, that particular extension class shouldn't have
had any base classes).  It was just the cyclic GC code triggering the bug.

It will be fixed in the next snapshot of pygtk for GTK+ 2.0

James.

-- 
Email: james@daa.com.au
WWW:   http://www.daa.com.au/~james/