"Stephen Pair" <spair at acm.org> wrote:
> Anthony wrote:
> > So I think both static scope (pools) and dynamic scope 
> > (contexts) are appropriate.  The significance is dynamic 
> > scope is not an extra concept like islands.  I believe this 
> > is more in line with the objects-as-capabilities concept.
> 
> Have you looked at RuntimeEnvironments (on SqueakMap)?  They accomplish
> the same effect, but they are cleaner than using exceptions...they also
> work when a process is forked.  To do what you describe would look like:
> 
> ----
> Env clamp: [
> 	Env at: #world put: aWorld.
> 	self doSomethingThatRequiresAWorld.
> ].
> ----
> SomeClass>>doSomethingThatRequiresAWorld
> 
> 	(Env at: #world) doSomething.
> ----

Sorry, Stephen, I forgot to mention your work.  Our two implementations
are basically equivalent.  I think mine is a little faster since it
reuses the context sender chain (hierarchy) instead of creating a new
parallel environment hierarchy.  This allows me to re-implement
Exceptions to use my general context handlers/environments, without loss
of efficiency.  Your point about environments being inherited by forked
processes is a nice option, but I wouldn't leave out exception handlers
from being inherited also.  I believe exception handlers and dynamic
variables should be treated the same, not differently.  (It would not be
hard to fork a process with the initial context holding context handlers
from its parent process).

Cheers,
Anthony