People tracking CVS will probably have noticed I'm pushing and pulling SGWayPoint around in various ways, which is foundation work for some route-manager changes. The major thing I've committed today, is removed every use of temporary waypoints to run SGWayPoint::CourseAndDistance, which is a proxy for wgs84_geo_inverse() ... and a rather tedious amount of boilerplate code to make the call.
I've converted all the calls in the main codebase to use Mathias' SGGeodesy helpers, which are generally much cleaner, and can work with SGGeod directly - in many places this means less passing around of lat/ lon/alt doubles. And it means the only code using SGWayPoint is the route-manager code, so I can now make further changes without affecting any other subsystems. Along the way I've converted a few more objects to store positions as SGGeod instead of raw doubles - notably FGViewer and the ground network nodes. In the FGViewer case this made the code considerably simpler, since it was already using Mathias' SGQuat and SGCart types, which convert cleanly from SGGeod. None of this should cause any user-visible change at all. I'm pretty confident about the FGviewer changes, but slightly less so about the AI traffic / schedule / groundnet changes; I ran the sim at Schiphol for some time and everything looks sane, but please do let me know if you think I might have regressed something. Comments or feedback welcome, as always. James ------------------------------------------------------------------------------ Crystal Reports - New Free Runtime and 30 Day Trial Check out the new simplified licensing option that enables unlimited royalty-free distribution of the report engine for externally facing server and web deployment. http://p.sf.net/sfu/businessobjects _______________________________________________ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
