[Openmcl-devel] CCL on Solaris sparc architecture

Ron Garret ron at flownet.com
Wed Jan 27 12:12:59 PST 2010


On Jan 27, 2010, at 11:22 AM, Helmut Eller wrote:

> * Ron Garret [2010-01-27 19:39+0100] writes:
> 
>> Ah.  Well, in that case I am at a loss.  I can't imagine why anyone
>> would care about a compiler front end (or use a Sparc for that matter)
>> unless they cared about performance.
> 
> Good compiler messages and debugging information need a reasonably good
> front-end.
> 

Why not use SBCL then?  Its front end is arguably better than CCL and it has a Sparc back end.

>> And I can't imagine why anyone
>> would use a JVM back end unless they didn't.
> 
> There exist plenty of libraries for the JVM that would be available to
> Lisp without the need to go through the FFI/JNI.  It would also be
> easier to make a Lisp library available to Java/JRuby/Jython/Jwhatever.
> 

Ah.  Well, if that's part of your quality metric then that is a quandry.  Have you considered Clojure?  Or Kawa?  SISC?  Maybe even Jython?  Or do you really need Common Lisp?  If so, have you considered paying Clozure Associates to write a JVM back end for you?

Feel free to treat those as rhetorical questions.

rg




More information about the Openmcl-devel mailing list