On Mar 28, 2005, at 10:57 AM, Aleksandar Donev wrote:
> Bill Long wrote:
>> I'd guess that because the first letter of NOT() was already one of
>> the "integer letters", there seemed to be no need to add the I.
> Actually I did not know about NOT (I should, I understand), and would
> have expected any "bit model" intrinsics to follow the "I" naming.
Why would you expect that? Or perhaps I should say that they *DO*
follow the only "I" naming convention that I know about, which is that
"I" was prefixed to names that didn't already begin with one of the
default implicit integer letters. I know of no particular naming
convention unique to the bit functions. Perhaps the reason you
associate "I" names with the bit functions is that many other
"interesting" Fortran intrinsics are more for reals instead of
integers, so you don't see the "I" convention as much as you see the
corresponding "A" one (for things such as AIMAG).
Although I wasn't there and so can't testify first hand, my fairly
strong guess is that there was no intent to associate bit intrinsics
with names starting with "I". I think that any such interpretation is
an after-the-fact impression (and an incorrect one, as evidenced by the
NOT function).
--
Richard Maine | Good judgment comes from experience;
[log in to unmask] | experience comes from bad judgment.
| -- Mark Twain
|