Re: [Bug-gnubg] 3d board - emails

[Jim Segrave]

On Sun 24 Aug 2003 (11:20 +0000), Jonathan Kinsey wrote:
> 
> Thanks for fixing the timing problem, I've committed the fix - Jim: does 
> this work for you now?

Under FreeBSD, no, I still get no activity during the idle loop,
although there's 100%CPU.
Under Linux, a laptop with no 3D acceleration, trying returning time
+= 1000, time += 100, time += 10000 all result in moves which take a
minute or more to display - there was no indication that any of the
three values made any difference.

> And the little endian problem - I was hoping it wouldn't go wrong :)
> And the sqrt/NAN problem - I've committed a fix, does it work ok?

It looks like this is a gcc/MacOS bug with floating point ops,
although a work around is possible, it would be useful if Olivier
could reduce this to the simplest possible example and submit ig to
the gcc bug mailing list.

-- 
Jim Segrave           address@hidden