How does a newbie get past the feeling thay he is trying to understand an elephant whilst looking through a keyhole?

Stephan Rudlof sr at evolgo.de
Sat Apr 29 16:35:05 UTC 2006


On 29.04.2006 17:12, Trygve Reenskaug wrote:
>...

> The most revolutionary part of Smalltalk is, IMO, the Stored Program Object 
> Computer. Seen in this light, its current programming language is a 
> language for microcode because it defines new operations on the objects, 
> just as regular microcode defines new operations on the binary words.

Interesting POV!

> 
> We need one or more languages that lets uts describe higher level 
> constructs where a program = Data + Communication + Algorithm. I have 
> written an exploratory example in Java that is based on strongly 
> encapsulated components. It seems to help readability, but at the cost of a 
> very bureaucratic object strucure.
> 

> What does the FORTRAN of OO look like?

Good question.

To go into this direction evolutionarily Squeak could have
1. namespaces;
2. multiple inheritance of code (e.g. Traits) - related to the former
point! -;
3. rigid conventions or tools or language extensions (not meant as xor)
for answering the questions - 'outside' more or less fine granular being
another class or package or ? -:
3.a) what is the 'official' interface, which *can/should* be used from
outside,
3.b) which part of above interface *has* been used from the outside;
4. BlockClosures.

Currently I have no opinion about package oriented code extensions
(class boxes), since this could lead to difficult to grasp code
variations being valid all together... (related to 1., 2., 3.).


Regards,
Stephan

> 
> Cheers
> --Trygve
>...
-- 
Stephan Rudlof (sr at evolgo.de)
   "Genius doesn't work on an assembly line basis.
    You can't simply say, 'Today I will be brilliant.'"
    -- Kirk, "The Ultimate Computer", stardate 4731.3



More information about the Squeak-dev mailing list