1
0
Fork 0
flightgear/src/Main/fg_init.cxx

831 lines
24 KiB
C++
Raw Normal View History

// fg_init.cxx -- Flight Gear top level initialization routines
//
// Written by Curtis Olson, started August 1997.
//
// Copyright (C) 1997 Curtis L. Olson - curt@infoplane.com
//
// This program is free software; you can redistribute it and/or
// modify it under the terms of the GNU General Public License as
// published by the Free Software Foundation; either version 2 of the
// License, or (at your option) any later version.
//
// This program is distributed in the hope that it will be useful, but
// WITHOUT ANY WARRANTY; without even the implied warranty of
// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
// General Public License for more details.
//
// You should have received a copy of the GNU General Public License
// along with this program; if not, write to the Free Software
// Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
//
//
// $Id$
1997-08-23 01:46:20 +00:00
#ifdef HAVE_CONFIG_H
# include <config.h>
#endif
// For BC 5.01 this must be included before OpenGL includes.
2001-03-23 22:42:49 +00:00
#ifdef SG_MATH_EXCEPTION_CLASH
# include <math.h>
#endif
#ifdef HAVE_WINDOWS_H
# include <windows.h>
#endif
#include <GL/glut.h>
1997-08-23 01:46:20 +00:00
#include <stdio.h>
#include <stdlib.h>
#if defined( unix ) || defined( __CYGWIN__ )
# include <unistd.h> // for gethostname()
#endif
// work around a stdc++ lib bug in some versions of linux, but doesn't
// seem to hurt to have this here for all versions of Linux.
#ifdef linux
# define _G_NO_EXTERN_TEMPLATES
#endif
2000-02-15 03:30:01 +00:00
#include <simgear/compiler.h>
#include STL_STRING
1997-08-23 01:46:20 +00:00
2000-02-15 03:30:01 +00:00
#include <simgear/constants.h>
2000-02-16 23:01:03 +00:00
#include <simgear/debug/logstream.hxx>
#include <simgear/math/point3d.hxx>
#include <simgear/math/polar3d.hxx>
2000-09-27 20:16:22 +00:00
#include <simgear/math/sg_geodesy.hxx>
#include <simgear/misc/sg_path.hxx>
#include <simgear/timing/sg_time.hxx>
2000-02-15 03:30:01 +00:00
1998-10-17 01:33:52 +00:00
#include <Aircraft/aircraft.hxx>
#include <FDM/UIUCModel/uiuc_aircraftdir.h>
#include <Airports/runways.hxx>
1998-08-25 16:59:08 +00:00
#include <Airports/simple.hxx>
#include <Autopilot/auto_gui.hxx>
#include <Autopilot/newauto.hxx>
#include <Cockpit/cockpit.hxx>
#include <Cockpit/radiostack.hxx>
#include <Cockpit/panel.hxx>
#include <Cockpit/panel_io.hxx>
#include <FDM/ADA.hxx>
1999-10-15 22:21:09 +00:00
#include <FDM/Balloon.h>
1999-11-19 02:10:24 +00:00
#include <FDM/External.hxx>
2000-07-24 17:27:12 +00:00
#include <FDM/JSBSim.hxx>
1999-10-14 01:53:43 +00:00
#include <FDM/LaRCsim.hxx>
#include <FDM/MagicCarpet.hxx>
#include <Include/general.hxx>
#include <Joystick/joystick.hxx>
#include <Objects/matlib.hxx>
#include <Navaids/fixlist.hxx>
#include <Navaids/ilslist.hxx>
2001-03-16 23:59:02 +00:00
#include <Navaids/mkrbeacons.hxx>
#include <Navaids/navlist.hxx>
#include <Scenery/scenery.hxx>
#include <Scenery/tilemgr.hxx>
#include <Time/event.hxx>
#include <Time/light.hxx>
#include <Time/sunpos.hxx>
#include <Time/moonpos.hxx>
#include <Time/tmp.hxx>
#ifndef FG_OLD_WEATHER
# include <WeatherCM/FGLocalWeatherDatabase.h>
#else
# include <Weather/weather.hxx>
#endif
1997-08-23 01:46:20 +00:00
#include "fg_init.hxx"
1999-11-19 02:10:24 +00:00
#include "fg_io.hxx"
#include "options.hxx"
2000-07-08 06:29:19 +00:00
#include "globals.hxx"
#include "bfi.hxx"
#if defined(FX) && defined(XMESA)
#include <GL/xmesa.h>
#endif
2001-03-23 22:59:18 +00:00
SG_USING_STD(string);
extern const char *default_root;
1997-08-23 01:46:20 +00:00
// Read in configuration (file and command line) and just set fg_root
bool fgInitFGRoot ( int argc, char **argv ) {
string root;
char* envp;
// First parse command line options looking for fg-root, this will
// override anything specified in a config file
root = fgScanForRoot(argc, argv);
#if defined( unix ) || defined( __CYGWIN__ )
// Next check home directory for .fgfsrc.hostname file
if ( root == "" ) {
envp = ::getenv( "HOME" );
if ( envp != NULL ) {
SGPath config( envp );
config.append( ".fgfsrc" );
char name[256];
gethostname( name, 256 );
config.concat( "." );
config.concat( name );
root = fgScanForRoot(config.str());
}
}
#endif
// Next check home directory for .fgfsrc file
if ( root == "" ) {
envp = ::getenv( "HOME" );
if ( envp != NULL ) {
SGPath config( envp );
config.append( ".fgfsrc" );
root = fgScanForRoot(config.str());
}
}
// Next check if fg-root is set as an env variable
if ( root == "" ) {
envp = ::getenv( "FG_ROOT" );
if ( envp != NULL ) {
root = envp;
}
}
// Otherwise, default to a random compiled-in location if we can't
// find fg-root any other way.
if ( root == "" ) {
#if defined( __CYGWIN__ )
root = "/FlightGear";
#elif defined( WIN32 )
root = "\\FlightGear";
#elif defined( macintosh )
root = "";
#else
root = PKGLIBDIR;
#endif
}
2001-03-24 06:03:11 +00:00
SG_LOG(SG_INPUT, SG_INFO, "fg_root = " << root );
globals->set_fg_root(root);
return true;
}
// Read in configuration (file and command line)
bool fgInitConfig ( int argc, char **argv ) {
// First, set some sane default values
fgSetDefaults();
// Read global preferences from $FG_ROOT/preferences.xml
SGPath props_path(globals->get_fg_root());
props_path.append("preferences.xml");
2001-03-24 06:03:11 +00:00
SG_LOG(SG_INPUT, SG_INFO, "Reading global preferences");
if (!readProperties(props_path.str(), globals->get_props())) {
2001-03-24 06:03:11 +00:00
SG_LOG(SG_INPUT, SG_ALERT, "Failed to read global preferences from "
<< props_path.str());
} else {
2001-03-24 06:03:11 +00:00
SG_LOG(SG_INPUT, SG_INFO, "Finished Reading global preferences");
}
// Attempt to locate and parse the various config files in order
// from least precidence to greatest precidence
// Check for $fg_root/system.fgfsrc
SGPath config( globals->get_fg_root() );
config.append( "system.fgfsrc" );
fgParseOptions(config.str());
char name[256];
#if defined( unix ) || defined( __CYGWIN__ )
// Check for $fg_root/system.fgfsrc.hostname
gethostname( name, 256 );
config.concat( "." );
config.concat( name );
fgParseOptions(config.str());
#endif
// Check for ~/.fgfsrc
char* envp = ::getenv( "HOME" );
if ( envp != NULL ) {
config.set( envp );
config.append( ".fgfsrc" );
fgParseOptions(config.str());
}
#if defined( unix ) || defined( __CYGWIN__ )
// Check for ~/.fgfsrc.hostname
gethostname( name, 256 );
config.concat( "." );
config.concat( name );
fgParseOptions(config.str());
#endif
// Parse remaining command line options
// These will override anything specified in a config file
fgParseOptions(argc, argv);
return true;
}
// find basic airport location info from airport database
bool fgFindAirportID( const string& id, FGAirport *a ) {
if ( id.length() ) {
SGPath path( globals->get_fg_root() );
path.append( "Airports" );
path.append( "simple.mk4" );
FGAirports airports( path.c_str() );
2001-03-24 06:03:11 +00:00
SG_LOG( SG_GENERAL, SG_INFO, "Searching for airport code = " << id );
if ( ! airports.search( id, a ) ) {
2001-03-24 06:03:11 +00:00
SG_LOG( SG_GENERAL, SG_ALERT,
"Failed to find " << id << " in " << path.str() );
return false;
}
} else {
return false;
}
2001-03-24 06:03:11 +00:00
SG_LOG( SG_GENERAL, SG_INFO,
"Position for " << id << " is ("
<< a->longitude << ", "
<< a->latitude << ")" );
return true;
}
// Set current_options lon/lat given an airport id
bool fgSetPosFromAirportID( const string& id ) {
FGAirport a;
// double lon, lat;
2001-03-24 06:03:11 +00:00
SG_LOG( SG_GENERAL, SG_INFO,
"Attempting to set starting position from airport code " << id );
if ( fgFindAirportID( id, &a ) ) {
fgSetDouble("/position/longitude", a.longitude );
fgSetDouble("/position/latitude", a.latitude );
2001-03-24 06:03:11 +00:00
SG_LOG( SG_GENERAL, SG_INFO,
"Position for " << id << " is ("
<< a.longitude << ", "
<< a.latitude << ")" );
return true;
} else {
return false;
}
}
// Set current_options lon/lat given an airport id and heading (degrees)
bool fgSetPosFromAirportIDandHdg( const string& id, double tgt_hdg ) {
FGRunway r;
FGRunway found_r;
double found_dir = 0.0;
if ( id.length() ) {
// set initial position from runway and heading
SGPath path( globals->get_fg_root() );
path.append( "Airports" );
path.append( "runways.mk4" );
FGRunways runways( path.c_str() );
2001-03-24 06:03:11 +00:00
SG_LOG( SG_GENERAL, SG_INFO,
"Attempting to set starting position from runway code "
<< id << " heading " << tgt_hdg );
// SGPath inpath( globals->get_fg_root() );
// inpath.append( "Airports" );
// inpath.append( "apt_simple" );
// airports.load( inpath.c_str() );
// SGPath outpath( globals->get_fg_root() );
// outpath.append( "Airports" );
// outpath.append( "simple.gdbm" );
// airports.dump_gdbm( outpath.c_str() );
if ( ! runways.search( id, &r ) ) {
2001-03-24 06:03:11 +00:00
SG_LOG( SG_GENERAL, SG_ALERT,
"Failed to find " << id << " in database." );
return false;
}
double diff;
double min_diff = 360.0;
while ( r.id == id ) {
// forward direction
diff = tgt_hdg - r.heading;
while ( diff < -180.0 ) { diff += 360.0; }
while ( diff > 180.0 ) { diff -= 360.0; }
diff = fabs(diff);
2001-03-24 06:03:11 +00:00
SG_LOG( SG_GENERAL, SG_INFO,
"Runway " << r.rwy_no << " heading = " << r.heading <<
" diff = " << diff );
if ( diff < min_diff ) {
min_diff = diff;
found_r = r;
found_dir = 0;
}
// reverse direction
diff = tgt_hdg - r.heading - 180.0;
while ( diff < -180.0 ) { diff += 360.0; }
while ( diff > 180.0 ) { diff -= 360.0; }
diff = fabs(diff);
2001-03-24 06:03:11 +00:00
SG_LOG( SG_GENERAL, SG_INFO,
"Runway -" << r.rwy_no << " heading = " <<
r.heading + 180.0 <<
" diff = " << diff );
if ( diff < min_diff ) {
min_diff = diff;
found_r = r;
found_dir = 180.0;
}
runways.next( &r );
}
2001-03-24 06:03:11 +00:00
SG_LOG( SG_GENERAL, SG_INFO, "closest runway = " << found_r.rwy_no
<< " + " << found_dir );
} else {
return false;
}
double heading = found_r.heading + found_dir;
while ( heading >= 360.0 ) { heading -= 360.0; }
double lat2, lon2, az2;
double azimuth = found_r.heading + found_dir + 180.0;
while ( azimuth >= 360.0 ) { azimuth -= 360.0; }
2001-03-24 06:03:11 +00:00
SG_LOG( SG_GENERAL, SG_INFO,
"runway = " << found_r.lon << ", " << found_r.lat
2001-03-24 04:56:46 +00:00
<< " length = " << found_r.length * SG_FEET_TO_METER * 0.5
<< " heading = " << azimuth );
geo_direct_wgs_84 ( 0, found_r.lat, found_r.lon,
2001-03-24 04:56:46 +00:00
azimuth, found_r.length * SG_FEET_TO_METER * 0.5 - 5.0,
&lat2, &lon2, &az2 );
2001-03-24 00:18:01 +00:00
if ( fabs( fgGetDouble("/sim/startup/offset-distance") ) > SG_EPSILON ) {
double olat, olon;
double odist = fgGetDouble("/sim/startup/offset-distance");
2001-03-24 04:56:46 +00:00
odist *= SG_NM_TO_METER;
double oaz = azimuth;
2001-03-24 00:18:01 +00:00
if ( fabs(fgGetDouble("/sim/startup/offset-azimuth")) > SG_EPSILON ) {
oaz = fgGetDouble("/sim/startup/offset-azimuth") + 180;
}
while ( oaz >= 360.0 ) { oaz -= 360.0; }
geo_direct_wgs_84 ( 0, lat2, lon2, oaz, odist, &olat, &olon, &az2 );
lat2=olat;
lon2=olon;
}
fgSetDouble("/position/longitude", lon2 );
fgSetDouble("/position/latitude", lat2 );
fgSetDouble("/orientation/heading", heading );
2001-03-24 06:03:11 +00:00
SG_LOG( SG_GENERAL, SG_INFO,
"Position for " << id << " is ("
<< lon2 << ", "
<< lat2 << ") new heading is "
<< heading );
return true;
}
// General house keeping initializations
bool fgInitGeneral( void ) {
string root;
1999-11-19 02:10:24 +00:00
#if defined(FX) && defined(XMESA)
char *mesa_win_state;
1999-11-19 02:10:24 +00:00
#endif
2001-03-24 06:03:11 +00:00
SG_LOG( SG_GENERAL, SG_INFO, "General Initialization" );
SG_LOG( SG_GENERAL, SG_INFO, "======= ==============" );
root = globals->get_fg_root();
1998-09-08 21:40:08 +00:00
if ( ! root.length() ) {
// No root path set? Then bail ...
2001-03-24 06:03:11 +00:00
SG_LOG( SG_GENERAL, SG_ALERT,
"Cannot continue without environment variable FG_ROOT"
<< "being defined." );
exit(-1);
}
2001-03-24 06:03:11 +00:00
SG_LOG( SG_GENERAL, SG_INFO, "FG_ROOT = " << '"' << root << '"' << endl );
#if defined(FX) && defined(XMESA)
// initialize full screen flag
global_fullscreen = false;
if ( strstr ( general.get_glRenderer(), "Glide" ) ) {
// Test for the MESA_GLX_FX env variable
if ( (mesa_win_state = getenv( "MESA_GLX_FX" )) != NULL) {
// test if we are fullscreen mesa/glide
1999-04-05 02:13:58 +00:00
if ( (mesa_win_state[0] == 'f') ||
(mesa_win_state[0] == 'F') ) {
global_fullscreen = true;
}
}
}
#endif
return true;
}
// This is the top level init routine which calls all the other
// initialization routines. If you are adding a subsystem to flight
// gear, its initialization call should located in this routine.
// Returns non-zero if a problem encountered.
bool fgInitSubsystems( void ) {
fgLIGHT *l = &cur_light_params;
2001-03-24 06:03:11 +00:00
SG_LOG( SG_GENERAL, SG_INFO, "Initialize Subsystems");
SG_LOG( SG_GENERAL, SG_INFO, "========== ==========");
From David Megginson: The files in the attached tarball make the following changes to FlightGear: - rename the existing FGInterface::init() method to FGInterface::_setup to get it out of the way - move *all* FDM initialization code out of src/Main/fg_init.cxx and into FGInterface::init(), and clean up fg_init.cxx a little (especially by removing the zillions of attempts to place the plane on the ground at various locations in the code) - modify FGInterface::bind() so that no values are picked up automatically at bind time (they are set previously by init() instead) - modify the init() methods of the classes derived from FGInterface (i.e. larcsim, jsbsim, balloon, magic, and ada) to invoke FGInterface::init() explicitly before doing their own setup I don't claim that the code in FGInterface::init() is optimal (or even correct), but it seems to work for on-ground starts with both LaRCSim and JSBSim on runways pointing various directions from near sea level to about 700' ASL (the range I happened to test). I expect that Jon and Tony will want to look at the code and refactor and correct it now that they can see what's going on in one place. Here's a quick outline of what is invoked: cur_fdm_state = new <whatever>(dt); cur_fdm_state->init(); cur_fdm_state->bind(); The constructor allocates memory and sets default values only (with the help of the FGInterface::_setup() method). The init() method pull any required properties out of the property tree and sets up the initial state of the FDM. The bind() method takes ownership of FDM-related properties so that the FDM can publish them to the rest of the sim. Note that bind() and init() are virtual, so any implementation in a subclass will hide the implementation in FGInterface; that's why subclass implementations of init() and bind() have to invoke FGInterface::init() and FGInterface::bind() explicitly, probably at the start, so that they get the basic initialization.
2001-04-05 20:20:44 +00:00
////////////////////////////////////////////////////////////////////
// Initialize the material property subsystem.
////////////////////////////////////////////////////////////////////
SGPath mpath( globals->get_fg_root() );
mpath.append( "materials" );
if ( material_lib.load( mpath.str() ) ) {
} else {
2001-03-24 06:03:11 +00:00
SG_LOG( SG_GENERAL, SG_ALERT, "Error loading material lib!" );
exit(-1);
}
From David Megginson: The files in the attached tarball make the following changes to FlightGear: - rename the existing FGInterface::init() method to FGInterface::_setup to get it out of the way - move *all* FDM initialization code out of src/Main/fg_init.cxx and into FGInterface::init(), and clean up fg_init.cxx a little (especially by removing the zillions of attempts to place the plane on the ground at various locations in the code) - modify FGInterface::bind() so that no values are picked up automatically at bind time (they are set previously by init() instead) - modify the init() methods of the classes derived from FGInterface (i.e. larcsim, jsbsim, balloon, magic, and ada) to invoke FGInterface::init() explicitly before doing their own setup I don't claim that the code in FGInterface::init() is optimal (or even correct), but it seems to work for on-ground starts with both LaRCSim and JSBSim on runways pointing various directions from near sea level to about 700' ASL (the range I happened to test). I expect that Jon and Tony will want to look at the code and refactor and correct it now that they can see what's going on in one place. Here's a quick outline of what is invoked: cur_fdm_state = new <whatever>(dt); cur_fdm_state->init(); cur_fdm_state->bind(); The constructor allocates memory and sets default values only (with the help of the FGInterface::_setup() method). The init() method pull any required properties out of the property tree and sets up the initial state of the FDM. The bind() method takes ownership of FDM-related properties so that the FDM can publish them to the rest of the sim. Note that bind() and init() are virtual, so any implementation in a subclass will hide the implementation in FGInterface; that's why subclass implementations of init() and bind() have to invoke FGInterface::init() and FGInterface::bind() explicitly, probably at the start, so that they get the basic initialization.
2001-04-05 20:20:44 +00:00
////////////////////////////////////////////////////////////////////
// Initialize the scenery management subsystem.
////////////////////////////////////////////////////////////////////
if ( fgSceneryInit() ) {
// Material lib initialized ok.
} else {
2001-03-24 06:03:11 +00:00
SG_LOG( SG_GENERAL, SG_ALERT, "Error in Scenery initialization!" );
exit(-1);
}
if ( global_tile_mgr.init() ) {
// Load the local scenery data
global_tile_mgr.update( fgGetDouble("/position/longitude"),
fgGetDouble("/position/latitude") );
} else {
2001-03-24 06:03:11 +00:00
SG_LOG( SG_GENERAL, SG_ALERT, "Error in Tile Manager initialization!" );
exit(-1);
}
2001-03-24 06:03:11 +00:00
SG_LOG( SG_GENERAL, SG_DEBUG,
1999-04-05 02:13:58 +00:00
"Current terrain elevation after tile mgr init " <<
scenery.cur_elev );
From David Megginson: The files in the attached tarball make the following changes to FlightGear: - rename the existing FGInterface::init() method to FGInterface::_setup to get it out of the way - move *all* FDM initialization code out of src/Main/fg_init.cxx and into FGInterface::init(), and clean up fg_init.cxx a little (especially by removing the zillions of attempts to place the plane on the ground at various locations in the code) - modify FGInterface::bind() so that no values are picked up automatically at bind time (they are set previously by init() instead) - modify the init() methods of the classes derived from FGInterface (i.e. larcsim, jsbsim, balloon, magic, and ada) to invoke FGInterface::init() explicitly before doing their own setup I don't claim that the code in FGInterface::init() is optimal (or even correct), but it seems to work for on-ground starts with both LaRCSim and JSBSim on runways pointing various directions from near sea level to about 700' ASL (the range I happened to test). I expect that Jon and Tony will want to look at the code and refactor and correct it now that they can see what's going on in one place. Here's a quick outline of what is invoked: cur_fdm_state = new <whatever>(dt); cur_fdm_state->init(); cur_fdm_state->bind(); The constructor allocates memory and sets default values only (with the help of the FGInterface::_setup() method). The init() method pull any required properties out of the property tree and sets up the initial state of the FDM. The bind() method takes ownership of FDM-related properties so that the FDM can publish them to the rest of the sim. Note that bind() and init() are virtual, so any implementation in a subclass will hide the implementation in FGInterface; that's why subclass implementations of init() and bind() have to invoke FGInterface::init() and FGInterface::bind() explicitly, probably at the start, so that they get the basic initialization.
2001-04-05 20:20:44 +00:00
////////////////////////////////////////////////////////////////////
// Initialize the flight model subsystem.
////////////////////////////////////////////////////////////////////
double dt = 1.0 / fgGetInt("/sim/model-hz");
// cout << "dt = " << dt << endl;
aircraft_dir = fgGetString("/sim/aircraft-dir");
const string &model = fgGetString("/sim/flight-model");
if (model == "larcsim") {
cur_fdm_state = new FGLaRCsim( dt );
} else if (model == "jsb") {
cur_fdm_state = new FGJSBsim( dt );
} else if (model == "ada") {
cur_fdm_state = new FGADA( dt );
} else if (model == "balloon") {
cur_fdm_state = new FGBalloonSim( dt );
} else if (model == "magic") {
cur_fdm_state = new FGMagicCarpet( dt );
} else if (model == "external") {
cur_fdm_state = new FGExternal( dt );
} else {
2001-03-24 06:03:11 +00:00
SG_LOG(SG_GENERAL, SG_ALERT,
"Unrecognized flight model '" << model
<< ", can't init aircraft");
exit(-1);
}
From David Megginson: The files in the attached tarball make the following changes to FlightGear: - rename the existing FGInterface::init() method to FGInterface::_setup to get it out of the way - move *all* FDM initialization code out of src/Main/fg_init.cxx and into FGInterface::init(), and clean up fg_init.cxx a little (especially by removing the zillions of attempts to place the plane on the ground at various locations in the code) - modify FGInterface::bind() so that no values are picked up automatically at bind time (they are set previously by init() instead) - modify the init() methods of the classes derived from FGInterface (i.e. larcsim, jsbsim, balloon, magic, and ada) to invoke FGInterface::init() explicitly before doing their own setup I don't claim that the code in FGInterface::init() is optimal (or even correct), but it seems to work for on-ground starts with both LaRCSim and JSBSim on runways pointing various directions from near sea level to about 700' ASL (the range I happened to test). I expect that Jon and Tony will want to look at the code and refactor and correct it now that they can see what's going on in one place. Here's a quick outline of what is invoked: cur_fdm_state = new <whatever>(dt); cur_fdm_state->init(); cur_fdm_state->bind(); The constructor allocates memory and sets default values only (with the help of the FGInterface::_setup() method). The init() method pull any required properties out of the property tree and sets up the initial state of the FDM. The bind() method takes ownership of FDM-related properties so that the FDM can publish them to the rest of the sim. Note that bind() and init() are virtual, so any implementation in a subclass will hide the implementation in FGInterface; that's why subclass implementations of init() and bind() have to invoke FGInterface::init() and FGInterface::bind() explicitly, probably at the start, so that they get the basic initialization.
2001-04-05 20:20:44 +00:00
cur_fdm_state->init();
cur_fdm_state->bind();
// allocates structures so must happen before any of the flight
// model or control parameters are set
fgAircraftInit(); // In the future this might not be the case.
From David Megginson: The files in the attached tarball make the following changes to FlightGear: - rename the existing FGInterface::init() method to FGInterface::_setup to get it out of the way - move *all* FDM initialization code out of src/Main/fg_init.cxx and into FGInterface::init(), and clean up fg_init.cxx a little (especially by removing the zillions of attempts to place the plane on the ground at various locations in the code) - modify FGInterface::bind() so that no values are picked up automatically at bind time (they are set previously by init() instead) - modify the init() methods of the classes derived from FGInterface (i.e. larcsim, jsbsim, balloon, magic, and ada) to invoke FGInterface::init() explicitly before doing their own setup I don't claim that the code in FGInterface::init() is optimal (or even correct), but it seems to work for on-ground starts with both LaRCSim and JSBSim on runways pointing various directions from near sea level to about 700' ASL (the range I happened to test). I expect that Jon and Tony will want to look at the code and refactor and correct it now that they can see what's going on in one place. Here's a quick outline of what is invoked: cur_fdm_state = new <whatever>(dt); cur_fdm_state->init(); cur_fdm_state->bind(); The constructor allocates memory and sets default values only (with the help of the FGInterface::_setup() method). The init() method pull any required properties out of the property tree and sets up the initial state of the FDM. The bind() method takes ownership of FDM-related properties so that the FDM can publish them to the rest of the sim. Note that bind() and init() are virtual, so any implementation in a subclass will hide the implementation in FGInterface; that's why subclass implementations of init() and bind() have to invoke FGInterface::init() and FGInterface::bind() explicitly, probably at the start, so that they get the basic initialization.
2001-04-05 20:20:44 +00:00
////////////////////////////////////////////////////////////////////
// Initialize the event manager subsystem.
////////////////////////////////////////////////////////////////////
global_events.Init();
// Output event stats every 60 seconds
global_events.Register( "fgEVENT_MGR::PrintStats()",
fgMethodCallback<fgEVENT_MGR>( &global_events,
&fgEVENT_MGR::PrintStats),
fgEVENT::FG_EVENT_READY, 60000 );
From David Megginson: The files in the attached tarball make the following changes to FlightGear: - rename the existing FGInterface::init() method to FGInterface::_setup to get it out of the way - move *all* FDM initialization code out of src/Main/fg_init.cxx and into FGInterface::init(), and clean up fg_init.cxx a little (especially by removing the zillions of attempts to place the plane on the ground at various locations in the code) - modify FGInterface::bind() so that no values are picked up automatically at bind time (they are set previously by init() instead) - modify the init() methods of the classes derived from FGInterface (i.e. larcsim, jsbsim, balloon, magic, and ada) to invoke FGInterface::init() explicitly before doing their own setup I don't claim that the code in FGInterface::init() is optimal (or even correct), but it seems to work for on-ground starts with both LaRCSim and JSBSim on runways pointing various directions from near sea level to about 700' ASL (the range I happened to test). I expect that Jon and Tony will want to look at the code and refactor and correct it now that they can see what's going on in one place. Here's a quick outline of what is invoked: cur_fdm_state = new <whatever>(dt); cur_fdm_state->init(); cur_fdm_state->bind(); The constructor allocates memory and sets default values only (with the help of the FGInterface::_setup() method). The init() method pull any required properties out of the property tree and sets up the initial state of the FDM. The bind() method takes ownership of FDM-related properties so that the FDM can publish them to the rest of the sim. Note that bind() and init() are virtual, so any implementation in a subclass will hide the implementation in FGInterface; that's why subclass implementations of init() and bind() have to invoke FGInterface::init() and FGInterface::bind() explicitly, probably at the start, so that they get the basic initialization.
2001-04-05 20:20:44 +00:00
////////////////////////////////////////////////////////////////////
// Initialize the view manager subsystem.
////////////////////////////////////////////////////////////////////
// Initialize win_ratio parameters
for ( int i = 0; i < globals->get_viewmgr()->size(); ++i ) {
globals->get_viewmgr()->get_view(i)->
set_win_ratio( fgGetInt("/sim/startup/xsize") /
fgGetInt("/sim/startup/ysize") );
}
// Initialize pilot view
FGViewerRPH *pilot_view =
(FGViewerRPH *)globals->get_viewmgr()->get_view( 0 );
pilot_view->set_geod_view_pos( cur_fdm_state->get_Longitude(),
cur_fdm_state->get_Lat_geocentric(),
cur_fdm_state->get_Altitude() *
2001-03-24 04:56:46 +00:00
SG_FEET_TO_METER );
pilot_view->set_sea_level_radius( cur_fdm_state->get_Sea_level_radius() *
2001-03-24 04:56:46 +00:00
SG_FEET_TO_METER );
pilot_view->set_rph( cur_fdm_state->get_Phi(),
cur_fdm_state->get_Theta(),
cur_fdm_state->get_Psi() );
// set current view to 0 (first) which is our main pilot view
globals->set_current_view( pilot_view );
2001-03-24 06:03:11 +00:00
SG_LOG( SG_GENERAL, SG_DEBUG, " abs_view_pos = "
<< globals->get_current_view()->get_abs_view_pos());
From David Megginson: The files in the attached tarball make the following changes to FlightGear: - rename the existing FGInterface::init() method to FGInterface::_setup to get it out of the way - move *all* FDM initialization code out of src/Main/fg_init.cxx and into FGInterface::init(), and clean up fg_init.cxx a little (especially by removing the zillions of attempts to place the plane on the ground at various locations in the code) - modify FGInterface::bind() so that no values are picked up automatically at bind time (they are set previously by init() instead) - modify the init() methods of the classes derived from FGInterface (i.e. larcsim, jsbsim, balloon, magic, and ada) to invoke FGInterface::init() explicitly before doing their own setup I don't claim that the code in FGInterface::init() is optimal (or even correct), but it seems to work for on-ground starts with both LaRCSim and JSBSim on runways pointing various directions from near sea level to about 700' ASL (the range I happened to test). I expect that Jon and Tony will want to look at the code and refactor and correct it now that they can see what's going on in one place. Here's a quick outline of what is invoked: cur_fdm_state = new <whatever>(dt); cur_fdm_state->init(); cur_fdm_state->bind(); The constructor allocates memory and sets default values only (with the help of the FGInterface::_setup() method). The init() method pull any required properties out of the property tree and sets up the initial state of the FDM. The bind() method takes ownership of FDM-related properties so that the FDM can publish them to the rest of the sim. Note that bind() and init() are virtual, so any implementation in a subclass will hide the implementation in FGInterface; that's why subclass implementations of init() and bind() have to invoke FGInterface::init() and FGInterface::bind() explicitly, probably at the start, so that they get the basic initialization.
2001-04-05 20:20:44 +00:00
////////////////////////////////////////////////////////////////////
// Initialize the lighting subsystem.
////////////////////////////////////////////////////////////////////
// fgUpdateSunPos() needs a few position and view parameters set
// so it can calculate local relative sun angle and a few other
// things for correctly orienting the sky.
fgUpdateSunPos();
fgUpdateMoonPos();
1998-09-15 04:27:27 +00:00
global_events.Register( "fgUpdateSunPos()", fgUpdateSunPos,
fgEVENT::FG_EVENT_READY, 60000);
global_events.Register( "fgUpdateMoonPos()", fgUpdateMoonPos,
fgEVENT::FG_EVENT_READY, 60000);
// Initialize Lighting interpolation tables
l->Init();
// update the lighting parameters (based on sun angle)
global_events.Register( "fgLight::Update()",
fgMethodCallback<fgLIGHT>( &cur_light_params,
&fgLIGHT::Update),
fgEVENT::FG_EVENT_READY, 30000 );
From David Megginson: The files in the attached tarball make the following changes to FlightGear: - rename the existing FGInterface::init() method to FGInterface::_setup to get it out of the way - move *all* FDM initialization code out of src/Main/fg_init.cxx and into FGInterface::init(), and clean up fg_init.cxx a little (especially by removing the zillions of attempts to place the plane on the ground at various locations in the code) - modify FGInterface::bind() so that no values are picked up automatically at bind time (they are set previously by init() instead) - modify the init() methods of the classes derived from FGInterface (i.e. larcsim, jsbsim, balloon, magic, and ada) to invoke FGInterface::init() explicitly before doing their own setup I don't claim that the code in FGInterface::init() is optimal (or even correct), but it seems to work for on-ground starts with both LaRCSim and JSBSim on runways pointing various directions from near sea level to about 700' ASL (the range I happened to test). I expect that Jon and Tony will want to look at the code and refactor and correct it now that they can see what's going on in one place. Here's a quick outline of what is invoked: cur_fdm_state = new <whatever>(dt); cur_fdm_state->init(); cur_fdm_state->bind(); The constructor allocates memory and sets default values only (with the help of the FGInterface::_setup() method). The init() method pull any required properties out of the property tree and sets up the initial state of the FDM. The bind() method takes ownership of FDM-related properties so that the FDM can publish them to the rest of the sim. Note that bind() and init() are virtual, so any implementation in a subclass will hide the implementation in FGInterface; that's why subclass implementations of init() and bind() have to invoke FGInterface::init() and FGInterface::bind() explicitly, probably at the start, so that they get the basic initialization.
2001-04-05 20:20:44 +00:00
////////////////////////////////////////////////////////////////////
// Initialize the local time subsystem.
////////////////////////////////////////////////////////////////////
// update the current timezone each 30 minutes
global_events.Register( "fgUpdateLocalTime()", fgUpdateLocalTime,
fgEVENT::FG_EVENT_READY, 1800000);
From David Megginson: The files in the attached tarball make the following changes to FlightGear: - rename the existing FGInterface::init() method to FGInterface::_setup to get it out of the way - move *all* FDM initialization code out of src/Main/fg_init.cxx and into FGInterface::init(), and clean up fg_init.cxx a little (especially by removing the zillions of attempts to place the plane on the ground at various locations in the code) - modify FGInterface::bind() so that no values are picked up automatically at bind time (they are set previously by init() instead) - modify the init() methods of the classes derived from FGInterface (i.e. larcsim, jsbsim, balloon, magic, and ada) to invoke FGInterface::init() explicitly before doing their own setup I don't claim that the code in FGInterface::init() is optimal (or even correct), but it seems to work for on-ground starts with both LaRCSim and JSBSim on runways pointing various directions from near sea level to about 700' ASL (the range I happened to test). I expect that Jon and Tony will want to look at the code and refactor and correct it now that they can see what's going on in one place. Here's a quick outline of what is invoked: cur_fdm_state = new <whatever>(dt); cur_fdm_state->init(); cur_fdm_state->bind(); The constructor allocates memory and sets default values only (with the help of the FGInterface::_setup() method). The init() method pull any required properties out of the property tree and sets up the initial state of the FDM. The bind() method takes ownership of FDM-related properties so that the FDM can publish them to the rest of the sim. Note that bind() and init() are virtual, so any implementation in a subclass will hide the implementation in FGInterface; that's why subclass implementations of init() and bind() have to invoke FGInterface::init() and FGInterface::bind() explicitly, probably at the start, so that they get the basic initialization.
2001-04-05 20:20:44 +00:00
////////////////////////////////////////////////////////////////////
// Initialize the weather subsystem.
////////////////////////////////////////////////////////////////////
// Initialize the weather modeling subsystem
#ifndef FG_OLD_WEATHER
// Initialize the WeatherDatabase
2001-03-24 06:03:11 +00:00
SG_LOG(SG_GENERAL, SG_INFO, "Creating LocalWeatherDatabase");
sgVec3 position;
sgSetVec3( position, current_aircraft.fdm_state->get_Latitude(),
current_aircraft.fdm_state->get_Longitude(),
2001-03-24 04:56:46 +00:00
current_aircraft.fdm_state->get_Altitude() * SG_FEET_TO_METER );
FGLocalWeatherDatabase::theFGLocalWeatherDatabase =
2000-12-06 04:13:16 +00:00
new FGLocalWeatherDatabase( position,
globals->get_fg_root() );
// cout << theFGLocalWeatherDatabase << endl;
// cout << "visibility = "
// << theFGLocalWeatherDatabase->getWeatherVisibility() << endl;
WeatherDatabase = FGLocalWeatherDatabase::theFGLocalWeatherDatabase;
2000-12-06 04:13:16 +00:00
double init_vis = fgGetDouble("/environment/visibility");
2000-12-06 04:13:16 +00:00
if ( init_vis > 0 ) {
WeatherDatabase->setWeatherVisibility( init_vis );
}
// register the periodic update of the weather
global_events.Register( "weather update", fgUpdateWeatherDatabase,
fgEVENT::FG_EVENT_READY, 30000);
#else
current_weather.Init();
#endif
1997-08-23 01:46:20 +00:00
From David Megginson: The files in the attached tarball make the following changes to FlightGear: - rename the existing FGInterface::init() method to FGInterface::_setup to get it out of the way - move *all* FDM initialization code out of src/Main/fg_init.cxx and into FGInterface::init(), and clean up fg_init.cxx a little (especially by removing the zillions of attempts to place the plane on the ground at various locations in the code) - modify FGInterface::bind() so that no values are picked up automatically at bind time (they are set previously by init() instead) - modify the init() methods of the classes derived from FGInterface (i.e. larcsim, jsbsim, balloon, magic, and ada) to invoke FGInterface::init() explicitly before doing their own setup I don't claim that the code in FGInterface::init() is optimal (or even correct), but it seems to work for on-ground starts with both LaRCSim and JSBSim on runways pointing various directions from near sea level to about 700' ASL (the range I happened to test). I expect that Jon and Tony will want to look at the code and refactor and correct it now that they can see what's going on in one place. Here's a quick outline of what is invoked: cur_fdm_state = new <whatever>(dt); cur_fdm_state->init(); cur_fdm_state->bind(); The constructor allocates memory and sets default values only (with the help of the FGInterface::_setup() method). The init() method pull any required properties out of the property tree and sets up the initial state of the FDM. The bind() method takes ownership of FDM-related properties so that the FDM can publish them to the rest of the sim. Note that bind() and init() are virtual, so any implementation in a subclass will hide the implementation in FGInterface; that's why subclass implementations of init() and bind() have to invoke FGInterface::init() and FGInterface::bind() explicitly, probably at the start, so that they get the basic initialization.
2001-04-05 20:20:44 +00:00
////////////////////////////////////////////////////////////////////
// Initialize vor/ndb/ils/fix list management and query systems
From David Megginson: The files in the attached tarball make the following changes to FlightGear: - rename the existing FGInterface::init() method to FGInterface::_setup to get it out of the way - move *all* FDM initialization code out of src/Main/fg_init.cxx and into FGInterface::init(), and clean up fg_init.cxx a little (especially by removing the zillions of attempts to place the plane on the ground at various locations in the code) - modify FGInterface::bind() so that no values are picked up automatically at bind time (they are set previously by init() instead) - modify the init() methods of the classes derived from FGInterface (i.e. larcsim, jsbsim, balloon, magic, and ada) to invoke FGInterface::init() explicitly before doing their own setup I don't claim that the code in FGInterface::init() is optimal (or even correct), but it seems to work for on-ground starts with both LaRCSim and JSBSim on runways pointing various directions from near sea level to about 700' ASL (the range I happened to test). I expect that Jon and Tony will want to look at the code and refactor and correct it now that they can see what's going on in one place. Here's a quick outline of what is invoked: cur_fdm_state = new <whatever>(dt); cur_fdm_state->init(); cur_fdm_state->bind(); The constructor allocates memory and sets default values only (with the help of the FGInterface::_setup() method). The init() method pull any required properties out of the property tree and sets up the initial state of the FDM. The bind() method takes ownership of FDM-related properties so that the FDM can publish them to the rest of the sim. Note that bind() and init() are virtual, so any implementation in a subclass will hide the implementation in FGInterface; that's why subclass implementations of init() and bind() have to invoke FGInterface::init() and FGInterface::bind() explicitly, probably at the start, so that they get the basic initialization.
2001-04-05 20:20:44 +00:00
////////////////////////////////////////////////////////////////////
2001-03-24 06:03:11 +00:00
SG_LOG(SG_GENERAL, SG_INFO, "Loading Navaids");
2000-04-25 17:33:13 +00:00
2001-03-24 06:03:11 +00:00
SG_LOG(SG_GENERAL, SG_INFO, " VOR/NDB");
current_navlist = new FGNavList;
SGPath p_nav( globals->get_fg_root() );
p_nav.append( "Navaids/default.nav" );
current_navlist->init( p_nav );
2001-03-24 06:03:11 +00:00
SG_LOG(SG_GENERAL, SG_INFO, " ILS and Marker Beacons");
2001-03-16 23:59:02 +00:00
current_beacons = new FGMarkerBeacons;
current_beacons->init();
current_ilslist = new FGILSList;
SGPath p_ils( globals->get_fg_root() );
p_ils.append( "Navaids/default.ils" );
current_ilslist->init( p_ils );
2001-03-24 06:03:11 +00:00
SG_LOG(SG_GENERAL, SG_INFO, " Fixes");
current_fixlist = new FGFixList;
SGPath p_fix( globals->get_fg_root() );
p_fix.append( "Navaids/default.fix" );
current_fixlist->init( p_fix );
From David Megginson: The files in the attached tarball make the following changes to FlightGear: - rename the existing FGInterface::init() method to FGInterface::_setup to get it out of the way - move *all* FDM initialization code out of src/Main/fg_init.cxx and into FGInterface::init(), and clean up fg_init.cxx a little (especially by removing the zillions of attempts to place the plane on the ground at various locations in the code) - modify FGInterface::bind() so that no values are picked up automatically at bind time (they are set previously by init() instead) - modify the init() methods of the classes derived from FGInterface (i.e. larcsim, jsbsim, balloon, magic, and ada) to invoke FGInterface::init() explicitly before doing their own setup I don't claim that the code in FGInterface::init() is optimal (or even correct), but it seems to work for on-ground starts with both LaRCSim and JSBSim on runways pointing various directions from near sea level to about 700' ASL (the range I happened to test). I expect that Jon and Tony will want to look at the code and refactor and correct it now that they can see what's going on in one place. Here's a quick outline of what is invoked: cur_fdm_state = new <whatever>(dt); cur_fdm_state->init(); cur_fdm_state->bind(); The constructor allocates memory and sets default values only (with the help of the FGInterface::_setup() method). The init() method pull any required properties out of the property tree and sets up the initial state of the FDM. The bind() method takes ownership of FDM-related properties so that the FDM can publish them to the rest of the sim. Note that bind() and init() are virtual, so any implementation in a subclass will hide the implementation in FGInterface; that's why subclass implementations of init() and bind() have to invoke FGInterface::init() and FGInterface::bind() explicitly, probably at the start, so that they get the basic initialization.
2001-04-05 20:20:44 +00:00
////////////////////////////////////////////////////////////////////
// Initialize the radio stack subsystem.
////////////////////////////////////////////////////////////////////
// A textbook example of how FGSubsystem
// should work...
current_radiostack = new FGRadioStack;
current_radiostack->init();
current_radiostack->bind();
From David Megginson: The files in the attached tarball make the following changes to FlightGear: - rename the existing FGInterface::init() method to FGInterface::_setup to get it out of the way - move *all* FDM initialization code out of src/Main/fg_init.cxx and into FGInterface::init(), and clean up fg_init.cxx a little (especially by removing the zillions of attempts to place the plane on the ground at various locations in the code) - modify FGInterface::bind() so that no values are picked up automatically at bind time (they are set previously by init() instead) - modify the init() methods of the classes derived from FGInterface (i.e. larcsim, jsbsim, balloon, magic, and ada) to invoke FGInterface::init() explicitly before doing their own setup I don't claim that the code in FGInterface::init() is optimal (or even correct), but it seems to work for on-ground starts with both LaRCSim and JSBSim on runways pointing various directions from near sea level to about 700' ASL (the range I happened to test). I expect that Jon and Tony will want to look at the code and refactor and correct it now that they can see what's going on in one place. Here's a quick outline of what is invoked: cur_fdm_state = new <whatever>(dt); cur_fdm_state->init(); cur_fdm_state->bind(); The constructor allocates memory and sets default values only (with the help of the FGInterface::_setup() method). The init() method pull any required properties out of the property tree and sets up the initial state of the FDM. The bind() method takes ownership of FDM-related properties so that the FDM can publish them to the rest of the sim. Note that bind() and init() are virtual, so any implementation in a subclass will hide the implementation in FGInterface; that's why subclass implementations of init() and bind() have to invoke FGInterface::init() and FGInterface::bind() explicitly, probably at the start, so that they get the basic initialization.
2001-04-05 20:20:44 +00:00
////////////////////////////////////////////////////////////////////
// Initialize the cockpit subsystem
////////////////////////////////////////////////////////////////////
if( fgCockpitInit( &current_aircraft )) {
// Cockpit initialized ok.
} else {
2001-03-24 06:03:11 +00:00
SG_LOG( SG_GENERAL, SG_ALERT, "Error in Cockpit initialization!" );
exit(-1);
1997-09-04 02:17:18 +00:00
}
From David Megginson: The files in the attached tarball make the following changes to FlightGear: - rename the existing FGInterface::init() method to FGInterface::_setup to get it out of the way - move *all* FDM initialization code out of src/Main/fg_init.cxx and into FGInterface::init(), and clean up fg_init.cxx a little (especially by removing the zillions of attempts to place the plane on the ground at various locations in the code) - modify FGInterface::bind() so that no values are picked up automatically at bind time (they are set previously by init() instead) - modify the init() methods of the classes derived from FGInterface (i.e. larcsim, jsbsim, balloon, magic, and ada) to invoke FGInterface::init() explicitly before doing their own setup I don't claim that the code in FGInterface::init() is optimal (or even correct), but it seems to work for on-ground starts with both LaRCSim and JSBSim on runways pointing various directions from near sea level to about 700' ASL (the range I happened to test). I expect that Jon and Tony will want to look at the code and refactor and correct it now that they can see what's going on in one place. Here's a quick outline of what is invoked: cur_fdm_state = new <whatever>(dt); cur_fdm_state->init(); cur_fdm_state->bind(); The constructor allocates memory and sets default values only (with the help of the FGInterface::_setup() method). The init() method pull any required properties out of the property tree and sets up the initial state of the FDM. The bind() method takes ownership of FDM-related properties so that the FDM can publish them to the rest of the sim. Note that bind() and init() are virtual, so any implementation in a subclass will hide the implementation in FGInterface; that's why subclass implementations of init() and bind() have to invoke FGInterface::init() and FGInterface::bind() explicitly, probably at the start, so that they get the basic initialization.
2001-04-05 20:20:44 +00:00
////////////////////////////////////////////////////////////////////
// Initialize the joystick subsystem.
////////////////////////////////////////////////////////////////////
if ( ! fgJoystickInit() ) {
2001-03-24 06:03:11 +00:00
SG_LOG( SG_GENERAL, SG_ALERT, "Error in Joystick initialization!" );
}
From David Megginson: The files in the attached tarball make the following changes to FlightGear: - rename the existing FGInterface::init() method to FGInterface::_setup to get it out of the way - move *all* FDM initialization code out of src/Main/fg_init.cxx and into FGInterface::init(), and clean up fg_init.cxx a little (especially by removing the zillions of attempts to place the plane on the ground at various locations in the code) - modify FGInterface::bind() so that no values are picked up automatically at bind time (they are set previously by init() instead) - modify the init() methods of the classes derived from FGInterface (i.e. larcsim, jsbsim, balloon, magic, and ada) to invoke FGInterface::init() explicitly before doing their own setup I don't claim that the code in FGInterface::init() is optimal (or even correct), but it seems to work for on-ground starts with both LaRCSim and JSBSim on runways pointing various directions from near sea level to about 700' ASL (the range I happened to test). I expect that Jon and Tony will want to look at the code and refactor and correct it now that they can see what's going on in one place. Here's a quick outline of what is invoked: cur_fdm_state = new <whatever>(dt); cur_fdm_state->init(); cur_fdm_state->bind(); The constructor allocates memory and sets default values only (with the help of the FGInterface::_setup() method). The init() method pull any required properties out of the property tree and sets up the initial state of the FDM. The bind() method takes ownership of FDM-related properties so that the FDM can publish them to the rest of the sim. Note that bind() and init() are virtual, so any implementation in a subclass will hide the implementation in FGInterface; that's why subclass implementations of init() and bind() have to invoke FGInterface::init() and FGInterface::bind() explicitly, probably at the start, so that they get the basic initialization.
2001-04-05 20:20:44 +00:00
////////////////////////////////////////////////////////////////////
// Initialize the autopilot subsystem.
////////////////////////////////////////////////////////////////////
current_autopilot = new FGAutopilot;
current_autopilot->init();
// initialize the gui parts of the autopilot
NewTgtAirportInit();
fgAPAdjustInit() ;
NewHeadingInit();
NewAltitudeInit();
From David Megginson: The files in the attached tarball make the following changes to FlightGear: - rename the existing FGInterface::init() method to FGInterface::_setup to get it out of the way - move *all* FDM initialization code out of src/Main/fg_init.cxx and into FGInterface::init(), and clean up fg_init.cxx a little (especially by removing the zillions of attempts to place the plane on the ground at various locations in the code) - modify FGInterface::bind() so that no values are picked up automatically at bind time (they are set previously by init() instead) - modify the init() methods of the classes derived from FGInterface (i.e. larcsim, jsbsim, balloon, magic, and ada) to invoke FGInterface::init() explicitly before doing their own setup I don't claim that the code in FGInterface::init() is optimal (or even correct), but it seems to work for on-ground starts with both LaRCSim and JSBSim on runways pointing various directions from near sea level to about 700' ASL (the range I happened to test). I expect that Jon and Tony will want to look at the code and refactor and correct it now that they can see what's going on in one place. Here's a quick outline of what is invoked: cur_fdm_state = new <whatever>(dt); cur_fdm_state->init(); cur_fdm_state->bind(); The constructor allocates memory and sets default values only (with the help of the FGInterface::_setup() method). The init() method pull any required properties out of the property tree and sets up the initial state of the FDM. The bind() method takes ownership of FDM-related properties so that the FDM can publish them to the rest of the sim. Note that bind() and init() are virtual, so any implementation in a subclass will hide the implementation in FGInterface; that's why subclass implementations of init() and bind() have to invoke FGInterface::init() and FGInterface::bind() explicitly, probably at the start, so that they get the basic initialization.
2001-04-05 20:20:44 +00:00
////////////////////////////////////////////////////////////////////
// Initialize I/O subsystem.
////////////////////////////////////////////////////////////////////
2000-09-10 00:04:50 +00:00
#if ! defined( macintosh )
1999-11-19 02:10:24 +00:00
fgIOInit();
#endif
// Initialize the 2D panel.
string panel_path = fgGetString("/sim/panel/path",
"Panels/Default/default.xml");
current_panel = fgReadPanel(panel_path);
if (current_panel == 0) {
2001-03-24 06:03:11 +00:00
SG_LOG( SG_INPUT, SG_ALERT,
"Error reading new panel from " << panel_path );
} else {
2001-03-24 06:03:11 +00:00
SG_LOG( SG_INPUT, SG_INFO, "Loaded new panel from " << panel_path );
current_panel->init();
current_panel->bind();
}
From David Megginson: The files in the attached tarball make the following changes to FlightGear: - rename the existing FGInterface::init() method to FGInterface::_setup to get it out of the way - move *all* FDM initialization code out of src/Main/fg_init.cxx and into FGInterface::init(), and clean up fg_init.cxx a little (especially by removing the zillions of attempts to place the plane on the ground at various locations in the code) - modify FGInterface::bind() so that no values are picked up automatically at bind time (they are set previously by init() instead) - modify the init() methods of the classes derived from FGInterface (i.e. larcsim, jsbsim, balloon, magic, and ada) to invoke FGInterface::init() explicitly before doing their own setup I don't claim that the code in FGInterface::init() is optimal (or even correct), but it seems to work for on-ground starts with both LaRCSim and JSBSim on runways pointing various directions from near sea level to about 700' ASL (the range I happened to test). I expect that Jon and Tony will want to look at the code and refactor and correct it now that they can see what's going on in one place. Here's a quick outline of what is invoked: cur_fdm_state = new <whatever>(dt); cur_fdm_state->init(); cur_fdm_state->bind(); The constructor allocates memory and sets default values only (with the help of the FGInterface::_setup() method). The init() method pull any required properties out of the property tree and sets up the initial state of the FDM. The bind() method takes ownership of FDM-related properties so that the FDM can publish them to the rest of the sim. Note that bind() and init() are virtual, so any implementation in a subclass will hide the implementation in FGInterface; that's why subclass implementations of init() and bind() have to invoke FGInterface::init() and FGInterface::bind() explicitly, probably at the start, so that they get the basic initialization.
2001-04-05 20:20:44 +00:00
////////////////////////////////////////////////////////////////////
// Initialize the BFI.
////////////////////////////////////////////////////////////////////
FGBFI::init();
From David Megginson: The files in the attached tarball make the following changes to FlightGear: - rename the existing FGInterface::init() method to FGInterface::_setup to get it out of the way - move *all* FDM initialization code out of src/Main/fg_init.cxx and into FGInterface::init(), and clean up fg_init.cxx a little (especially by removing the zillions of attempts to place the plane on the ground at various locations in the code) - modify FGInterface::bind() so that no values are picked up automatically at bind time (they are set previously by init() instead) - modify the init() methods of the classes derived from FGInterface (i.e. larcsim, jsbsim, balloon, magic, and ada) to invoke FGInterface::init() explicitly before doing their own setup I don't claim that the code in FGInterface::init() is optimal (or even correct), but it seems to work for on-ground starts with both LaRCSim and JSBSim on runways pointing various directions from near sea level to about 700' ASL (the range I happened to test). I expect that Jon and Tony will want to look at the code and refactor and correct it now that they can see what's going on in one place. Here's a quick outline of what is invoked: cur_fdm_state = new <whatever>(dt); cur_fdm_state->init(); cur_fdm_state->bind(); The constructor allocates memory and sets default values only (with the help of the FGInterface::_setup() method). The init() method pull any required properties out of the property tree and sets up the initial state of the FDM. The bind() method takes ownership of FDM-related properties so that the FDM can publish them to the rest of the sim. Note that bind() and init() are virtual, so any implementation in a subclass will hide the implementation in FGInterface; that's why subclass implementations of init() and bind() have to invoke FGInterface::init() and FGInterface::bind() explicitly, probably at the start, so that they get the basic initialization.
2001-04-05 20:20:44 +00:00
////////////////////////////////////////////////////////////////////
// Initialize the controls subsystem.
////////////////////////////////////////////////////////////////////
controls.init();
controls.bind();
From David Megginson: The files in the attached tarball make the following changes to FlightGear: - rename the existing FGInterface::init() method to FGInterface::_setup to get it out of the way - move *all* FDM initialization code out of src/Main/fg_init.cxx and into FGInterface::init(), and clean up fg_init.cxx a little (especially by removing the zillions of attempts to place the plane on the ground at various locations in the code) - modify FGInterface::bind() so that no values are picked up automatically at bind time (they are set previously by init() instead) - modify the init() methods of the classes derived from FGInterface (i.e. larcsim, jsbsim, balloon, magic, and ada) to invoke FGInterface::init() explicitly before doing their own setup I don't claim that the code in FGInterface::init() is optimal (or even correct), but it seems to work for on-ground starts with both LaRCSim and JSBSim on runways pointing various directions from near sea level to about 700' ASL (the range I happened to test). I expect that Jon and Tony will want to look at the code and refactor and correct it now that they can see what's going on in one place. Here's a quick outline of what is invoked: cur_fdm_state = new <whatever>(dt); cur_fdm_state->init(); cur_fdm_state->bind(); The constructor allocates memory and sets default values only (with the help of the FGInterface::_setup() method). The init() method pull any required properties out of the property tree and sets up the initial state of the FDM. The bind() method takes ownership of FDM-related properties so that the FDM can publish them to the rest of the sim. Note that bind() and init() are virtual, so any implementation in a subclass will hide the implementation in FGInterface; that's why subclass implementations of init() and bind() have to invoke FGInterface::init() and FGInterface::bind() explicitly, probably at the start, so that they get the basic initialization.
2001-04-05 20:20:44 +00:00
////////////////////////////////////////////////////////////////////////
// End of subsystem initialization.
////////////////////////////////////////////////////////////////////
2001-03-24 06:03:11 +00:00
SG_LOG( SG_GENERAL, SG_INFO, endl);
// Save the initial state for future
// reference.
globals->saveInitialState();
return true;
1997-08-23 01:46:20 +00:00
}
void fgReInitSubsystems( void )
{
2001-03-24 06:03:11 +00:00
SG_LOG( SG_GENERAL, SG_INFO,
2001-01-17 23:30:35 +00:00
"/position/altitude = " << fgGetDouble("/position/altitude") );
2000-07-08 06:29:19 +00:00
bool freeze = globals->get_freeze();
if( !freeze )
globals->set_freeze( true );
// Initialize the Scenery Management subsystem
if ( ! fgSceneryInit() ) {
2001-03-24 06:03:11 +00:00
SG_LOG( SG_GENERAL, SG_ALERT, "Error in Scenery initialization!" );
exit(-1);
}
if( global_tile_mgr.init() ) {
// Load the local scenery data
global_tile_mgr.update( fgGetDouble("/position/longitude"),
fgGetDouble("/position/latitude") );
} else {
2001-03-24 06:03:11 +00:00
SG_LOG( SG_GENERAL, SG_ALERT, "Error in Tile Manager initialization!" );
exit(-1);
}
// Initialize view parameters
FGViewerRPH *pilot_view =
(FGViewerRPH *)globals->get_viewmgr()->get_view( 0 );
pilot_view->set_view_offset( 0.0 );
pilot_view->set_goal_view_offset( 0.0 );
pilot_view->set_geod_view_pos( cur_fdm_state->get_Longitude(),
cur_fdm_state->get_Lat_geocentric(),
cur_fdm_state->get_Altitude() *
2001-03-24 04:56:46 +00:00
SG_FEET_TO_METER );
pilot_view->set_sea_level_radius( cur_fdm_state->get_Sea_level_radius() *
2001-03-24 04:56:46 +00:00
SG_FEET_TO_METER );
pilot_view->set_rph( cur_fdm_state->get_Phi(),
cur_fdm_state->get_Theta(),
cur_fdm_state->get_Psi() );
// set current view to 0 (first) which is our main pilot view
globals->set_current_view( pilot_view );
2001-03-24 06:03:11 +00:00
SG_LOG( SG_GENERAL, SG_DEBUG, " abs_view_pos = "
<< globals->get_current_view()->get_abs_view_pos());
cur_fdm_state->init();
controls.reset_all();
current_autopilot->reset();
fgUpdateSunPos();
fgUpdateMoonPos();
cur_light_params.Update();
fgUpdateLocalTime();
2000-07-08 06:29:19 +00:00
if( !freeze )
globals->set_freeze( false );
}