Self and pure object thinking

Mark Mayfield mmayfield at netexplorer.com
Fri Mar 20 23:20:21 UTC 1998


Thanks!

Having published two books with Peter Coad on object model patterns and
Java design, I want to get to the heart of object-oriented thinking. Having
seen Smalltalk-80 on an Alto at Xerox during the summers of 1981 and 1982,
and reading from cover to cover many times my first computer book
"Smalltalk-80: The Language and its Implementation," (Yes...I have my
original Blue book sitting here on my bookshelf at work) I want to move
pure object thinking forward into the formal analysis and design
activities. I'm tired of having to build class oriented object models that
to be understood, are too rigid. When I add dynamic capabilities (object
inheritance, i.e. composition and delegation) to a class-oriented model,
the diagrams become harder to understand and therefore communicate less. My
goal is a simpler notation that expresses the powerful ideas that were
presented in languages like Self, NewtonScript, and ObjectLogo. Models
built using this pure object notation can still be implemented in languages
like Smalltalk and Java through the use of notation-to-implementation
patterns. The only way to grow complex systems is with simple principles.
Complex notations are counter productive when trying to model complex
systems.

If this is a crazy idea...let me know.

Mark

>>You might find the page on prototype based languages in Rainer Blome's
>>Bookmarks
>>interesting:
>>
>>http://www.physik3.gwdg.de/~rainer/
>
>That page has some really good info.
>I've summarized it in this chart:
>
> http://www.slip.net/~dekorte/Proto/Chart.html
>
>Steve



___________________________________________________________
Mark Mayfield                     mmayfield at netexplorer.com

   "The best way to predict the future is to invent it."

       Alan Kay





More information about the Squeak-dev mailing list