Smalltalk evolution

Hello Craig,

        Do you plan to, perhaps, include some enhancements to the Smalltalk  
language itself?  For example, silently ignoring message failure  
instead of raising DoesNotUnderstand exception?  Myself, I used to be  
completely against this until I started working with Objective-C and  
it made reading code so much easier because it was just so much  
clearer.  Not understanding a message really isn't an exceptional  
circumstance.  Besides, the compiler draws your attention to when such  
things occur by providing a message 'message not found.  Unknown  
methods are expected to return type 'id' and take a variable number of  
arguments'.  Useful stuff.  Elegance is, in my opinion, underrated in  
its utility.

        Namespaces perhaps?  Prototypes with Slots?

        Unicode message names?  I certainly think the last would be useful  
especially in the case of advanced mathematics and unary operators  
using Unicode math symbols.  Of course, this is just an idea out of my  
head so there is probably some compelling reason why I've not heard of  
it before.

        And, to clarify my understanding:  All message sends are asynchronous  
in Spoon?

        These and more questions are probably answered elsewhere but I've not  
had the occasion to run across them yet.  If anyone else wants to jump  
in, please do.  I've dabbled with Smalltalk over the years and am now  
finding myself spending more and more time with it as a personal  
preference.  I am, by no means, versed well enough yet to not be  
dangerous :)

        All these questions point out one final question:  Where can one find  
out about the current evolutionary efforts related to Squeak and  
Smalltalk?  I see lots of random articles but no real driving process...

BR,
Dave
_______________________________________________
Spoon mailing list
[hidden email]
http://lists.squeakfoundation.org/cgi-bin/mailman/listinfo/spoon

Hi David--

 > Do you plan to, perhaps, include some enhancements to the Smalltalk
 > language itself?

      Well, no, but I have made changes to some fundamental parts of the
Smalltalk environment. Many might consider them to be so fundamental as
to be changes to the language. The main ones are getting rid of the
system dictionary and fileouts.

 > For example, silently ignoring message failure instead of raising
 > DoesNotUnderstand exception?

      No, I don't plan to do that.

 > Namespaces perhaps?

      I don't think namespaces per se are necessary. In Spoon the name
of each class can be anything, and classes are referenced between
systems without using their names at all. Developers will probably want
to create groups of conceptually-related classes, but they won't be
namespaces. In effect, every class has its own namespace.

 > Prototypes with Slots? Unicode message names?

      I have no plans for those.

 > And, to clarify my understanding:  All message sends are asynchronous
 > in Spoon?

      No, at the moment all messages are synchronous. This is just
what's needed to bootstrap a system. Developers can provide modules
which provide fancier forms of messaging later (like what Croquet uses).

 > Where can one find out about the current evolutionary efforts related
 > to Squeak and Smalltalk?  I see lots of random articles but no real
 > driving process...

      You have to talk to people, you can't read it in an article
somewhere (such a thing would be obsolete as soon as it was written,
anyway). Asking here and on the Squeak group are good places to start.


      thanks,

-C


_______________________________________________
Spoon mailing list
[hidden email]
http://lists.squeakfoundation.org/cgi-bin/mailman/listinfo/spoon