The cleverest thing in Java is the combination of the Java VM spec,
and the way classes are defined as complete entitities with well
defined interfaces.
Lisp can fight back by adopting these two very good ideas and building
a lisp-like system based on them, as a superset of the java vm/java class
spec. Lisp would then be able to "just use" any existing java code.
The lisp VM would naturally include provision for multiple inheritance,
multiple dispatch, fully dynamic runtime behavior, etc. This would *not*
be a VM for common lisp - the slate would have to be clean to have any chance
of success. Even so, I suspect that the "VM spec" would be widely regarded as
bloated and impossible to implement effeciently, but that's a different battle.
The other front to fight back on is to write a binding of a common
lisp-like language to the existing JVM, with the necessary compromises
skewed 'way toward full functionality over effeciency. The discipline of
reorganizing common lisp's morass of functions into java-like classes
will help focus the lisp VM effort.
Chance of this actualy happening? 0.00001
---
Tack #two: admit that java has won, and play microsoft's game
(with better motives) by defining extensions to java that add
missing behavior. Start with macros and lisp-like syntax. Once
you have those, you're free to make vast additions to the *apparent*
language, while still implementing them in the same-old java vm.
Chance of this actually happening? 0.1
---
BTW, I think there is no way lisp can fight back based
directly on any of the existing lisp implementations; they're
doomed to the backwaters of computing.
--
My home page: http://www.andromeda.com/people/ddyer/home.html
In article <······················@192.0.2.1>, ·····@netcom.com (Dave Dyer)
wrote:
> The other front to fight back on is to write a binding of a common
> lisp-like language
How about EuLisp?
--
http://www.lavielle.com/~joswig/
On Mon, 07 Apr 1997 22:18:44 +0200, ······@lavielle.com (Rainer
Joswig) wrote:
>In article <······················@192.0.2.1>, ·····@netcom.com (Dave Dyer)
>wrote:
>
>> The other front to fight back on is to write a binding of a common
>> lisp-like language
>
>How about EuLisp?
>
>--
>http://www.lavielle.com/~joswig/
If you really want to Lisp to prosper, work to convince Gates et al
that Lisp can help him counter the Java-everywhere trend.
Ian Garmaise
----------------
Ian Garmaise
····@cris.com