I am really interested in Marvin, but I would like to know a bit more
about the need of a new language, what is the vision, how do you want
the language to be in 5 years from now...
In what will it be useful for? What are the pattern where prototypes
fit bettern then objects (issued from classes) ?
Cheers,
Alexandre
I didn't want to make new language like Slate, but the current design
of Marvin is IMHO the best way how to enhance Squeak with prototypes.
The reasons why Self/Marvin may be better than Smalltalk:
- simpler and cleaner semantics
- multiple inheritance/mixins
- dynamic inheritance
- namespaces
Self is the logical ancestor of Smalltalk and the fact, that it has
improved capabilities of Smalltalk by the simplification of this
language (in contrast with Python, Ruby etc.), is amazing.
Prototype-based systems are predetermined for visual programming so
they may be very useful for future three-dimensional collaborative
environments (like Croquet).
Many people need to use Self for their research projects but Self is
almost dead today. Marvin may be the interesting alternative.
Because Marvin is integrated with Squeak, it doesn't need complex
environment, set of traits and prototypes. But in future I'd like to
see it more independent (within Squeak).
Currently I'm working on an outliner
(http://www.comtalk.net/Squeak/uploads/95/outliner.PNG).
By the way, have you ever had a Squeak presentation in Brno? :-)
-- Pavel
I like your screenshot, it looks very cool!
Yes, I presented Squeak in Brno, a couple of years ago. I was in
contact with Vladimir Janousek.
Please, let me know about your project.
Cheers,
Alexandre