[NTLUG:Discuss] The wrong computation example from the newsgroup

[Peter Koren]

Here is a good piece on numerics which should make one wary of the
hidden problems and snares waiting in floating point programming:

http://www.ddj.com/articles/1997/9711/9711a/9711a.htm?topic=scientific

-- Pete Koren


Steve Baker wrote:
> 
> sysmail at glade.net wrote:
> >
> > Hmmm...  Pardon spewage from a newcomer, but gcc is an application, not
> > an operating system.
> 
> It is *possible* for this to be Linux's "fault" (except that it's not a fault)
> on the PC. Even though it's already been shown to work "as expected" under
> Linux on Alpha's and on the PC with optimisation turned on.
> 
> I would imagine that the hardware floating point unit is initialised by Linux
> and not by GCC and it's libraries.  I believe that the low level rounding
> policy is determined by the chip setup - and hence there is a teeny-tiny
> chance that this could be Linux's "fault" (bearing in mind that this isn't
> a fault, etc, etc).
> 
> I know that some PC games (under Windoze) deliberately change the rounding
> mode of the hardware for some arcane reason.
> 
> --
> Steve Baker   HomeEmail: <sjbaker1 at airmail.net>
>               WorkEmail: <sjbaker at link.com>
>               HomePage : http://web2.airmail.net/sjbaker1
>               Projects : http://plib.sourceforge.net
>                          http://tuxaqfh.sourceforge.net
>                          http://tuxkart.sourceforge.net
>                          http://prettypoly.sourceforge.net
>                          http://freeglut.sourceforge.net
> _______________________________________________
> http://ntlug.org/mailman/listinfo/discuss