Re: Mother ship App. (distributing FASL files)

[cfry@MIT.EDU (Christopher Fry)]

> There where some talks recently about using FASL files
> as a code distribution means. The advantages of this approach
> have been highlited, but I also see two problems with it:
> 
> First, there's the speed problem. For big programs, people
> won't be happy waiting a few minutes for the application to
> load.
> 
> Second and more serious, as these applications would share the
> same mac application (the MCL mother ship) they would all have
> the same mac signature, so you could not have them receive
> specific apple events.

If we can add some extra bits to the supported apple-events, then 
they could indicate what "lisp-app" they were suppose to go to.
It would be up to the loaded lisp-environment to branch
on this extra "lisp-app" info. If each lisp-app was represented as 
a CLOS object that contained its lisp-app-signature, then when an
apple-event came in , each lisp-app object would be asked if it was
for them. As soon as one of them returns T, no more are asked.
Of course, a lisp-app might deviously run some code based
on the event then still return NIL allowing other lisp-apps a shot 
at it.

In any case, I'd consider this whole issue of apple-events not 
significant enough to kill off the whole idea esp because the
rest of it seems relatively easy to implement per gained 
functionality.