| Re: Praise for Gfortran (finally) |
|
 |
|
 |
|
 |
|
 |
Group: comp.lang.fortran · Group Profile
Author: nospamnospam Date: Sep 20, 2008 12:21
glen herrmannsfeldt ugcs.caltech.edu> wrote:
> I could imagine more differences in the results of compile
> time and run time intrinsic functions. Another reason not
> to do equality testing of floating point values.
Yes, and I think the current thinking is more along exactly those lines
- that this is a case of the reasons to avoid dependence on exact
equality of floating point things in general, as opposed to a reason to
limit the language just because such inequality might happen in a cross
compilation.
--
Richard Maine | Good judgement comes from experience;
email: last name at domain . net | experience comes from bad judgement.
domain: summertriangle | -- Mark Twain
| 
