James C Field wrote:
> Your parlour trick exposes an inadequacy in J. String theory (really a
> model) is untestable.
The fact that I can use J to easily express a mathematical concept, and
communicate it via email, and know with complete confidence
that my audience will see what I wrote as I wrote it (all without using any
extraneous or special tools), is a strong advantage of
J, and a weakness of SMN and APL.
Also, The J phrase +/ 2 ^ - i. _ is no less "testable" than:
OO
-----
\ -i
\ 2
/
/
-----
i=0
... it's just more succinct, consistent, and easier to type (and easier to
extend, even dramatically extend) . They are simply
different notational expressions of the same mathematical concept; neither
calculable, but amenable to "solution" by symbolic
manipulation and insight. So if the J is inadequate, so is the SMN.
If you don't believe that, then you must explain the thread to which you
responded, wherein the mathematical concept is discussed
and dissected, and furthermore where Roger explains how he could easily extend
the J interpreter to make the J formulation
calculable! That is, the J expression already already has all the value of the
SMN expression (because the two isomorphic), and one
day it might have the overwhelming advantage of being *executable*.
> I would love to see EVIDENCE of the obvious practical or commercial or
> demonstrated benefits of J.
Well, I made my living (and several others') for a number of years, exclusively
with J. And I'm not alone. Have you seen
http://www.jsoftware.com/jwiki/Stories ?
-Dan
----------------------------------------------------------------------
For information about J forums see http://www.jsoftware.com/forums.htm
