Why does 2.3a2 change "slots" in complex?

[Terry Reedy]

"David Mertz" <mertz at gnosis.cx> wrote in message
news:mailman.1046488322.24610.python-list at python.org...
> Our gnosis.xml.pickle package makes an effort to pickle various
objects,
> generally trying even harder than the standard pickle module.  In
> particular, we try to do sensible things with new-style classes that
> have slots, and that don't.

The standard pickle has (or will have) a new protocol added to do the
same.

> One of the unit tests involves adding extra attributes to
descendents of
> standard data types.  It turns out, that test breaks with Python
2.3a2,
> because of a change in the implementation of complex.

Looking at Sourceforge CVS, there was one small patch to
complexobject.c on Jan 29, first since Oct 5, that added .getnewargs()
.  Unclear how could have affected attribute setting.

Terry J. Reedy