I've been working on the integration of the Pocket Smalltalk IDE into the
Squeak IDE.  The basic idea I'm working on, and its close to working now, is
that I have a Pocket Smalltalk Project.  The project holds onto an
environment which is comprised solely of the Pocket Smalltalk classes.  I
have entry and exit code to replace the globals like SystemOrganization and
a few others with ones that will only have the Pocket Smalltalk organization
and such.  So at the moment, you can create a Pocket Smalltalk project and
do Pocket Smalltalk code within it.  It is kind of cool to be able to dip in
and out of Pocket Smalltalk and Squeak.

The reason for this for doing this is that I would like to be able to use
the Squeak tools to work with Pocket Smalltalk code, rather than having to
build yet another set of tools.  Some of the problems I am running into is
that a lot of the tools work with off of many globals.  Globals such as
Smalltalk, for the environment, SystemOrganization for the organization of
the classes.

What I was wondering was whether it made sense to look at changing a lot of
the code that depends on globals of this nature, to go through the currently
active project.  I'm starting to do this, but through the creation of
subclasses, and installing these subclasses into their corresponding super
class while in a Pocket ST project.  It seems to me that using project
boundaries for these makes a lot of sense, especially with the module (I
think their called image segments ???) stuff that is being implemented.

What do you guys think ?