Re: GNAT Optimization of Constant Expressions

["Dmitry A. Kazakov" <mailbox@dmitry-kazakov.de>]

On Thu, 17 May 2007 18:38:34 -0500, Randy Brukardt wrote:

[...]
> And the use of constants as arguments to those expensive
> routines is pretty silly. (Besides, good Ada practice would be to name your
> constants, and when you did that, you got essentially the same code as with
> Fortran.)

I think that it not a good idea to rely on good practice. There is a
language problem of specifying that the calls to a function with static
constant shall yield static result. This is not just sin (Pi/3.0), but also

   for ... loop
      if Match (Text, <pattern-expression>) then

or

   Create_Combo_Box ("ada" + "c" +  "c++" + ...);

or

   5.0 * m + 10.0 * feet

Folding self-descriptive expressions into constants declared miles away
from their only use is not a good style.

-- 
Regards,
Dmitry A. Kazakov
http://www.dmitry-kazakov.de