I also see this kind of bug with MSVC. It appears that the call of
calc_gc_course_dist( dest, start, &course, &dist ); at line 195 of cloud.cxx
returns an invalid course.

This course is invalid because around line 182 of polar3d.hxx, there is the
calculation of acos from a value greater than 1. I don't know why the math
is
bogus (perhaps a problem of precision) but I tried to clamp the value to
[-1,1]
and it doesn't loop anymore.

Could someone double-check the math in calc_gc_course_dist to see why
(sin(dest.y()) - sin_start_y * cos(d)) / (sin(d) * cos_start_y) is sometimes
greater than 1 ?

-Fred

----- Original Message -----
From: "Christian Mayer" <[EMAIL PROTECTED]>
To: "FGFS" <[EMAIL PROTECTED]>
Sent: Tuesday, February 12, 2002 3:30 PM
Subject: [Flightgear-devel] Latest CVS doesn't run


> Hi,
>
> the latest version (updated everything, incl. base) doesn't run for me
> (MSVC as usual) as it's hooked up in an endless loop in cloud.cxx again.
>
> I thought we had that fixed already :( Can anybody remember the patch?
>
> CU,
> Christian
>
> --
> The idea is to die young as late as possible.        -- Ashley Montague
>
> Whoever that is/was; (c) by Douglas Adams would have been better...
>
> _______________________________________________
> Flightgear-devel mailing list
> [EMAIL PROTECTED]
> http://mail.flightgear.org/mailman/listinfo/flightgear-devel
>


_______________________________________________
Flightgear-devel mailing list
[EMAIL PROTECTED]
http://mail.flightgear.org/mailman/listinfo/flightgear-devel

Reply via email to