Tracy Harms wrote:
> This article has been catching some attention. There is no mention of
> J in it, but J has come up in at least some resulting discussions.

I can't find the discussions, but this point seems to come down to the
assumption that functional programming is lazy evaluation (which IS a good
idea), and function level programming is ignored.  Mathematica can do the
latter quite well, and so can J.

   newton=:1 : '-(u % u d. 1)'
   f=:_2+*:
   f newton^:_ ] 2
1.41421
   f newton^:(i.4) 2
2 1.5 1.41667 1.41422
   f newton^:(i.4) 2r1
2 3r2 17r12 577r408
   1&o. newton^:_ ] 3
3.14159

Best wishes,

John


----------------------------------------------------------------------
For information about J forums see http://www.jsoftware.com/forums.htm

Reply via email to