I would also like to point out some very important issues with respect to C#, Java, and Python.
C# has precisely one implementation.
Java has precisely one useful implementation (there are others, but all come with sacrifices).
Python has precisely one implementation.
Lisp? Not so much.
I'll tell you what -- you make A Lisp implementation, note the singular, and rally a community around that one implementation, work hard at pushing it, make it portable across Linux, BSD, Windows, et. al., contribute sensible, orthogonal libraries to it, and provide a SINGLE source for downloads, documentation, an easy-to-use and easy-to-contribute module system (which I think rules out 90% of ANSI CL), easy and standardized Lisp-to-C foreign function interface, and reject elitism in the user community, accept that portability is a myth and embrace POSIX- and Windows-specific APIs where it makes sense to (true portability comes from factoring common code -- witness how Python deals with OS-specific stuff via is 'os' module), and I can guarantee you you'll have a fair modicum of success.
This is happening with Factor today; with increasing frequency, I see people whom I'd never expect to use a stack-based language enjoying Factor. In five years time, I fully expect Factor to have a larger userbase than Arc.
I would also like to point
I would also like to point out some very important issues with respect to C#, Java, and Python.
C# has precisely one implementation.
Java has precisely one useful implementation (there are others, but all come with sacrifices).
Python has precisely one implementation.
Lisp? Not so much.
I'll tell you what -- you make A Lisp implementation, note the singular, and rally a community around that one implementation, work hard at pushing it, make it portable across Linux, BSD, Windows, et. al., contribute sensible, orthogonal libraries to it, and provide a SINGLE source for downloads, documentation, an easy-to-use and easy-to-contribute module system (which I think rules out 90% of ANSI CL), easy and standardized Lisp-to-C foreign function interface, and reject elitism in the user community, accept that portability is a myth and embrace POSIX- and Windows-specific APIs where it makes sense to (true portability comes from factoring common code -- witness how Python deals with OS-specific stuff via is 'os' module), and I can guarantee you you'll have a fair modicum of success.
This is happening with Factor today; with increasing frequency, I see people whom I'd never expect to use a stack-based language enjoying Factor. In five years time, I fully expect Factor to have a larger userbase than Arc.