This is the initial commit for a reworked environment controller.
The main intention is to decouple the individual modules like
metar fetch, metar properties, environment interpolation etc.
to make it easier for other weather module developers to modify
environment settings.
As a side effect, the dialogs for weather-scenario, weather-conditions,
clouds and precipitations have been merged into a single dialog
Issue a warning when finding duplicate tail numbers. Note that this is far from a fatal error, but does interfere with the startup heuristics data collection mechanism.
* A few new features to allow database management and traffic pattern developement:
--prop:/sim/traffic-manager/dumpdata=true: dumps the currently loaded datafile to the console for inspection.
--prop:/sim/traffic-manager/datafile=/path/to/somefile.xml: Skips autoscanning and reads traffic only from the specified xml file.
--prop:/sim/traffic-manager/datafile=/path/to/somefile.conf: Slips autoscanning and reads traffic from a plain text file, as described on the forum/wiki
* Removed the use of pointers for heuristics calculations, as urged by Frederic Bouvier
* Fixed domain error in proportion calculation during loading, resulting in random skipping of aircraft, as reported by Jorg van der Venne.
Walk the AImodels property tree only once at init, rather than at update. Should have a beneficial effect on frame rate, but in practice makes no discernible improvement. It is cleaner code though.
Durk Talsma [Mon, 30 Aug 2010 19:13:16 +0000 (21:13 +0200)]
Enable the collection of AI aircraft position statistics. This will allow for a dramatic speedup of the establishment of AI traffic after program initialization.
Durk Talsma [Sun, 29 Aug 2010 17:25:34 +0000 (19:25 +0200)]
Major update to the AI code:
* New features
- More realistic descent paths
- Separation during descent and approach
- ATC approach controller (still silent)
- inbound traffic flow will start immediately
* Bug fixes
- Properly handle vertical speed when on ground
- Departing aircraft now wait for taxiclerance before moving
- Traffic manager waits for proper weather initialization
- Fixed instabilities in the preferential runway usage code
- Fine tuning of waypoint following code.
Vivian Meazza [Sat, 28 Aug 2010 23:00:09 +0000 (00:00 +0100)]
Major update of Ballistic objects and submodels. Introduce submodels slaved to AI Objects. Remove deprecated code. Try to increase framerate. Use quaternion calculations for Initial Conditions.
Torsten Dreyer [Mon, 23 Aug 2010 15:47:57 +0000 (17:47 +0200)]
Some terrainsampler improvements
- use max. computation time instead of max computations per frame
- restart computation of an area only after significant movememt
- let the area move with a heading and a speed
Torsten Dreyer [Wed, 18 Aug 2010 16:50:07 +0000 (18:50 +0200)]
Autopilot is more than just an autopilot: Introducing: "property rule"
The autopilot has grown beyond being just an autopilot. It's
components, filters and controllers could be used to set property
values based on complex rules and driven from other property values.
That's why I chose to give it the name "property-rule".
This patch allows to define an arbitrary number of property rules under
a/sim/property-rule[n]/path
using the same syntax as autopilot configurations.
Torsten Dreyer [Sat, 14 Aug 2010 20:42:31 +0000 (22:42 +0200)]
First step into implementing the local-weather system
This is the first part of the local-weather implementation
of Thorsten Renk, currently written entirly in Nasal. Here
comes the terrain-sampling subsystem as a first step. It is
not (yet) included in the build process which will be
performed when the system has a little matured.
James Turner [Sat, 14 Aug 2010 18:16:28 +0000 (19:16 +0100)]
Fix bug 150 (airports not found in GPS search)
Overhaul and simplify the GPS search logic based
on experience and testing with the API. Also fix
search-by-name, which was broken.
James Turner [Sat, 14 Aug 2010 09:23:13 +0000 (10:23 +0100)]
Thorsten Brehm: GPWS: Fixed permanent blocking of lower prio warnings
Only block callouts which were actually issued to the pilot. When
multiple callouts trigger at the same time, allow lower prio warnings
to be issued once the higher prio ones are resolved.
=> Gear/flap warnings are now working reliably (no longer blocked
permanently when altitude callouts are enabled).
Torsten Dreyer [Fri, 6 Aug 2010 07:06:32 +0000 (09:06 +0200)]
Interpret environment variables and hostname on Windows
- Point the homedir variable to %APPDATA%/flightgear.org. This
enables .fgfsrc in this path for windows users.
- Evaluate the returncode of gethostname() to not strdup() bogus
characters on failure. Also mainly on windows.
James Turner [Tue, 13 Jul 2010 20:50:44 +0000 (21:50 +0100)]
Create TimeManager subsystem, and collect the time related code out of main.cxx and fg_init.cxx into it.
Remove the unfortunately named 'tmp.[cxx|hxx]', pushing the remaining code in FGLight.
(second try, with init bug fixed)
James Turner [Tue, 13 Jul 2010 20:50:44 +0000 (21:50 +0100)]
Create TimeManager subsystem, and collect the time related code out of main.cxx and fg_init.cxx into it.
Remove the unfortunately named 'tmp.[cxx|hxx]', pushing the remaining code in FGLight.
Durk Talsma [Sun, 1 Aug 2010 14:16:33 +0000 (16:16 +0200)]
Added some stubs in the flightgear main code to eventually add a full, FGSubsystems based ATIS system that will serve as a replacement for the ATCDCL code.
James Turner [Fri, 30 Jul 2010 08:20:36 +0000 (09:20 +0100)]
Thorsten: EGPWS fixes!
* Fix issue #139, uninitialized variable causing sim deadlock in MK_VIII::Mode5Handler::get_soft_bias
* Fix more uninitialized variables sometimes causing warnings not to work.
* Fix some warnings only working on a single approach (missing reset).
* EGPWS self-test can now be triggered more than once (missing reset).
* Implement configuration options for attitude and altitude input selection.