How about:
> !.1e_33
>!.1.0000000000000001e_33
>"0!.1e_33
|domain error
| >"0!.1e_33
This is "the same" as my >/0$a: solution. The question is do
solutions that exploit the difference between the primitive > and
the derived entity >"0 count? That is, can I stand upon J's
self-knowledge, or does that fall loosely under the rubrik "special
code"? If not, could the challenge be phrased as
Produce an expression in which the noun result of:
[x] f"0 y
differs from the noun result of:
[x] f"0"0 y
for some noun(s) [x] y and all verbs f ,
in a way guarunteed by the Dictionary of J,
excluding the appendicies and entries for !: ?
-Dan
----------------------------------------------------------------------
For information about J forums see http://www.jsoftware.com/forums.htm
