On Sun, Feb 23, 2003 at 10:51:07AM -0500, rif wrote:
> Thank you very much for the clear and detailed explanation.  It
> certainly makes sense, although it does seem to have some unfortuante
> implications for the abstract design of fast programs in Lisp as
> compared to C/C++.  

Yes. CL is reasonably good at being efficient considering that it
refuses to compromise on expressiveness (where C and C++ are
reasonably good at being expressive considering that they refuse to
compromise on efficiency). Sometimes the efficiency cost of Lisp is
the extra indirection discussed above, and sometimes it's other stuff
like dynamic typing overhead, automatic GC, or support for dynamic
redefinition of stuff. The cost is seldom huge, but it can easily be
significant.

IMO Lisp is often appropriate when the big risk is that the program
will take three times too long to develop, or that it will run 333
times too slowly, or that it will fail 3% of the time; but you should
think twice about using Lisp where the big risk is that the program
will run three times too slowly.

-- 
William Harold Newman <[EMAIL PROTECTED]>
Remember: don't show fear.  Testers can sense fear.
  -- Paul F. Dietz, cmucl-imp 2003-01-27
PGP key fingerprint 85 CE 1C BA 79 8D 51 8C  B9 25 FB EE E0 C3 E5 7C

Reply via email to