Mathias Fröhlich writes:
>
> On Samstag 05 November 2005 17:21, Vassilii Khachaturov wrote:
> > With either jsbsim or yasim aircraft, when is in the vicinity of the
> > (North) pole, the AGL (as seen in the HUD) goes into 2*10^7 ranges. You
> > can either start up with --lat=90 (and any longtitude you please), or, if
> > you dislike the singularity of --lat=90 at the startup, use --lat=88 and
> > head north. Soon past 89 degrees you'll see it happening. (I initially
> > discovered it by trying to start up with santa at the pole :) ).
> >
> > When this happens, one can fly below earth (altitude-wise, as indicated on
> > the altimeter) down and down w/o a problem on an aircraft (like hunter)
> > that doesn't allow it normally.
> >
> > I don't think this is a particuarly annoying aspect of the flightgear
> > universe, but maybe somebody will get a hint to another bug from this
> > report.
> It is just that there is no scenery for that area, I think.
>
> The poles have a singularity in the lat/lon/altitude coordinate system. That
> singularity might be the reason these places do not even have a water
> surface, at least I guess that reason ...

The code use to have this test which prevented you from getting *too* close
to the pole and seemed to work well enough

note: one second of latitude is approx 20 meters

void sgGeocToGeod( const double& lat_geoc, const double& radius,
                                   double *lat_geod, double *alt, double 
*sea_level_r )
{
 ............
        if( ( (PI_OVER_2 - lat_geoc) < SG_ONE_SECOND )        // near North pole
                  || ( (PI_OVER_2 + lat_geoc) < SG_ONE_SECOND ) )   // near 
South pole
        {
                *lat_geod = lat_geoc;
                *sea_level_r = Geocent_a*E;
                *alt = radius - *sea_level_r;
        } else {

Not sure if an equiv test is included in the current code

Of course all of this nonsense could be avoided if latitude and longitude were
not used for internal representaion.  eg  use Cartesian XYZ form instead

Norman


_______________________________________________
Flightgear-devel mailing list
Flightgear-devel@flightgear.org
http://mail.flightgear.org/mailman/listinfo/flightgear-devel
2f585eeea02e2c79d7b1d8c4963bae2d

Reply via email to