This changeset isn't just a clean-up, it fixes a rather large bug.  When
I talked about possible subclass issues, I was referring to classes that
weren't explicitly tested in the SUnit tests, but have their own
implementation of some of the add/remove methods.  These should not hold
up inclusion in 3.6 for several reasons.

First, they occur in classes that wouldn't cause distress to everyday
Squeaking if problems turned up, such as ScreeningMorph and
PianoRollMorph.

Second, I was tired last night, and was being cautious about the claims
I was making on behalf of Andreas' code.  Faced with the prospect of
postponing this until 3.7, and refreshed after a night's sleep, I have
looked again at the code in question more closely, and manually
exercised it.  None of the code is doing anything fancy.  For example,
PianoScrollMorph>>removedMorph: overrides the Morph method, but is
almost an exact copy of the previously-misused and now-removed
PianoScrollMorph>>privateRemoveMorph:.  The closer examination and
manual exercise turned up no problems.

Another reason is that the next Croquet release will probably be based
on 3.6, and almost certainly not on 3.7a.  Since hardware-accelerated 3D
windows are what brought this bug to light in the first place, it would
be nice to include this now rather than having to include it separately
in Croquet.  

Basically, I think that any risks posed by the changeset (which, upon
closer examination, are very small), are outweighed by the immediate fix
of an outstanding bug.  The code cleanup is icing.













< I'm a bug-fixing machine! >

This post brought to you by the BugFixArchiveViewer, a handy tool that
makes it easy to comment on proposed fixes and enhancements for Squeak. 
For more information, check out the Web page for the BugFixArchiveViewer
project: http://minnow.cc.gatech.edu/squeak/3214 

< I'm a bug-fixing machine! >