LaRCsim maintains all it's variables internally. I had been copying all of

them back and forth to the FG struture everytime I updated the flight model.
However, I have realized that this is not necessary.  I just need to copy
the control positions and environmental parameters into the LaRCsim structure
before updating the FDM, then copy every thing back out into the publick FGFS
structure afterwords.  This seems to solve (or at least help) a westward
drift problem some poeple had been observing.

diff --git a/FDM/LaRCsim.cxx b/FDM/LaRCsim.cxx
index 4c8f644cf65fe573efbc89b48121060c713fc315..5e206dd24e244f20200e7c5dc5ceb7b11fc413ac 100644 (file)
--- a/FDM/LaRCsim.cxx
+++ b/FDM/LaRCsim.cxx
@@ -51,8 +51,20 @@ int fgLaRCsimUpdate(FGState& f, int multiloop) {
        f.set_Altitude( 0.0 );
     }
 
+    // copy control positions into the LaRCsim structure
+    Lat_control = controls.get_aileron();
+    Long_control = controls.get_elevator();
+    Long_trim = controls.get_elevator_trim();
+    Rudder_pedal = controls.get_rudder();
+    Throttle_pct = controls.get_throttle( 0 );
+    Brake_pct = controls.get_brake( 0 );
+
+    // Inform LaRCsim of the local terrain altitude
+    Runway_altitude =   f.get_Runway_altitude();
+
+    // old -- FGstate_2_LaRCsim() not needed except for Init()
     // translate FG to LaRCsim structure
-    FGState_2_LaRCsim(f);
+    // FGState_2_LaRCsim(f);
     // printf("FG_Altitude = %.2f\n", FG_Altitude * 0.3048);
     // printf("Altitude = %.2f\n", Altitude * 0.3048);
     // printf("Radius to Vehicle = %.2f\n", Radius_to_vehicle * 0.3048);
@@ -79,13 +91,6 @@ int fgLaRCsimUpdate(FGState& f, int multiloop) {
 // Convert from the FGState struct to the LaRCsim generic_ struct
 int FGState_2_LaRCsim (FGState& f) {
 
-    Lat_control = controls.get_aileron();
-    Long_control = controls.get_elevator();
-    Long_trim = controls.get_elevator_trim();
-    Rudder_pedal = controls.get_rudder();
-    Throttle_pct = controls.get_throttle( 0 );
-    Brake_pct = controls.get_brake( 0 );
-
     Mass =      f.get_Mass();
     I_xx =      f.get_I_xx();
     I_yy =      f.get_I_yy();
@@ -382,6 +387,15 @@ int fgLaRCsim_2_FGState (FGState& f) {
 
 
 // $Log$
+// Revision 1.7  1998/12/14 13:31:06  curt
+// LaRCsim maintains all it's variables internally.  I had been copying all of
+// them back and forth to the FG struture everytime I updated the flight model.
+// However, I have realized that this is not necessary.  I just need to copy
+// the control positions and environmental parameters into the LaRCsim structure
+// before updating the FDM, then copy every thing back out into the publick FGFS
+// structure afterwords.  This seems to solve (or at least help) a westward
+// drift problem some poeple had been observing.
+//
 // Revision 1.6  1998/12/05 15:54:08  curt
 // Renamed class fgFLIGHT to class FGState as per request by JSB.
 //