On Mar 22, 2:13 pm, William James <w_a_x_...@[EMAIL PROTECTED]
> wrote:
> We hear that locals discourage factoring. The proof?
> Newbies use locals and don't factor.
No, the proof is that when you use locals, it makes it
harder to factor. Something that makes an activity
easier, encourages it, something that makes an activity
harder, discourages it.
Where things go awry is when things are placed through
the "take everything to ridiculous extremes" filter and
people read "locals discourage factoring" as:
* "locals make it impossible to factor", which is a bald-
faced lie, since C relies heavily on locals, and it
is possible to write well-factored C code ... not
always encouraged, and oft-times with a runtime
penalty, but possible.
* "locals are antithetical to forth" which is silly ...
the argument goes that locals introduce syntactic
restrictions on how we program, and the point of Forth
is the opportunities that arise from a substantial
freedom from syntax ... but if the second premise is
correct, it contradicts the supposed conclusion,
since the point is how little is "antithetical" to Forth.
Locals introduce a capacity and an inflexibility, and like
any other construct, the question is when the benefit of
the capacity justifies working with the inflexibility.
Certainly when implementing Forth wrappers for C library
calls, they almost always pay their own way, and the same
with quickly and correctly prototyping published
mathematical expressions. For me, that's enough to
justify having them.
|
