]> git.mxchange.org Git - flightgear.git/log
flightgear.git
18 years agoVivian Meazza:
ehofman [Thu, 9 Feb 2006 12:29:05 +0000 (12:29 +0000)]
Vivian Meazza:

This patch is a combined effort by Gregor Richards, Oliver Schroeder, and
Vivian Meazza (and code cleanups and improvements by Erik Hofman). It corrects
the bug in which a Multiplayer model responds to local inputs, and the view
number bug which caused certain aircraft to appear as cockpit only models. It
passes remote properties over the net, and all major control surfaces and gear
are now animated correctly, providing that the local ~model.xml file contains
no leading "/" in the <property></property> data entries. MP objects
are now extrapolated using 1st and 2nd derivatives to make their movement
appear more smooth. The sim is now halted while a new client joins the net.

Known problems with MP are non-display of the remote client under certain
circumstances of starting/resetting, and a freeze on starting. These bugs are
long standing, and are not addressed by this patch.

Special thanks must go to AJ Macleod for his patient testing of this patch over many evenings.

We have also moved part of multiplayer into AIModels as part of the ongoing
development of MP.

18 years agoadd workaround for gdb bug
mfranz [Thu, 9 Feb 2006 11:08:17 +0000 (11:08 +0000)]
add workaround for gdb bug
s/calltree/callgrind/

18 years agomake sure the index is in the array, which isn't necessarily the case,
mfranz [Wed, 8 Feb 2006 17:18:24 +0000 (17:18 +0000)]
make sure the index is in the array, which isn't necessarily the case,
because a dialog button can get pressed when its callback isn't available
any more. Ideally, we would close the dialog or update its buttons when
this happens. But this requires bigger changes.

18 years agoMathias Fröhlich:
ehofman [Wed, 8 Feb 2006 10:25:56 +0000 (10:25 +0000)]
Mathias Fröhlich:

Make use of the automatic reference counting class for ssg* data in flightgear.

18 years agoSync. w. JSBSim CVS.
ehofman [Wed, 8 Feb 2006 09:15:57 +0000 (09:15 +0000)]
Sync. w. JSBSim CVS.

18 years agoemit to different message properties:
mfranz [Tue, 7 Feb 2006 22:35:22 +0000 (22:35 +0000)]
emit to different message properties:
/sim/messages/{atc,pilot,ai-plane,ground,approach}
These can the be used for differenct screen.log colors and for different
voices. (Aircraft & tutoruals may want to write to /sim/messages/copilot)

18 years agoadd al-info
mfranz [Sun, 5 Feb 2006 16:41:47 +0000 (16:41 +0000)]
add al-info

18 years agoVivian MEAZZA:
mfranz [Sun, 5 Feb 2006 11:33:08 +0000 (11:33 +0000)]
Vivian MEAZZA:

"Bug fix - ensure all TACAN channels are included"

18 years agocosmetics
mfranz [Sat, 4 Feb 2006 23:27:31 +0000 (23:27 +0000)]
cosmetics

18 years agopublic domain
mfranz [Sat, 4 Feb 2006 23:07:41 +0000 (23:07 +0000)]
public domain

18 years agoexport ATC message to /sim/messages/atc
mfranz [Sat, 4 Feb 2006 16:55:05 +0000 (16:55 +0000)]
export ATC message to /sim/messages/atc

18 years agostore garbage collector id (again) in the listener class, and free the
mfranz [Sat, 4 Feb 2006 16:37:25 +0000 (16:37 +0000)]
store garbage collector id (again) in the listener class, and free the
handler on destruction. There's no explicit Nasal command for removing
a listener, but deleting the node should fully remove them.

18 years agomake room for more font specifications in style files
mfranz [Sat, 4 Feb 2006 13:06:47 +0000 (13:06 +0000)]
make room for more font specifications in style files

18 years agotypo
mfranz [Sat, 4 Feb 2006 01:03:41 +0000 (01:03 +0000)]
typo

18 years agoRemove unused local variables
fredb [Fri, 3 Feb 2006 21:04:37 +0000 (21:04 +0000)]
Remove unused local variables

18 years agoonly write coordinates on button press, and not again on release
mfranz [Fri, 3 Feb 2006 11:39:04 +0000 (11:39 +0000)]
only write coordinates on button press, and not again on release

18 years agofix for deadlock problem (as suggested by Andy)
mfranz [Thu, 2 Feb 2006 18:03:07 +0000 (18:03 +0000)]
fix for deadlock problem (as suggested by Andy)

18 years agoVassilii Khachaturov:
ehofman [Thu, 2 Feb 2006 09:57:33 +0000 (09:57 +0000)]
Vassilii Khachaturov:

Fix the current buggy rain orientation behaviour for the views attached to the
aircraft (while still inheriting bugs with the views attached to anything else).

18 years agoPrevent a warning.
ehofman [Thu, 2 Feb 2006 09:40:27 +0000 (09:40 +0000)]
Prevent a warning.

18 years agoSync. w. JSBSim CVS
ehofman [Thu, 2 Feb 2006 09:35:25 +0000 (09:35 +0000)]
Sync. w. JSBSim CVS

18 years agocosmetics
mfranz [Thu, 2 Feb 2006 00:47:30 +0000 (00:47 +0000)]
cosmetics

18 years agobetter use zero sensitive area to detect a missing (or inoperable) button,
mfranz [Thu, 2 Feb 2006 00:00:27 +0000 (00:00 +0000)]
better use zero sensitive area to detect a missing (or inoperable) button,
because a missing <button> default to 0, which makes a lot of sense

18 years agopanel_io.cxx: if an action doesn't have a button, fire its bindings
mfranz [Wed, 1 Feb 2006 23:52:36 +0000 (23:52 +0000)]
panel_io.cxx: if an action doesn't have a button, fire its bindings
    immediately and then throw it away; this is useful for initialization
    purposes (see dme.xml)

cockpit.cxx: warnings--; typos--

18 years agofix 2D compass
mfranz [Wed, 1 Feb 2006 20:54:17 +0000 (20:54 +0000)]
fix 2D compass

18 years agoAdd OpenAL 1.0 backwards compatibility and OS-X support.
ehofman [Wed, 1 Feb 2006 13:27:36 +0000 (13:27 +0000)]
Add OpenAL 1.0 backwards compatibility and OS-X support.

18 years agoAdd a gl-info equivalent for OpenAL.
ehofman [Tue, 31 Jan 2006 18:20:49 +0000 (18:20 +0000)]
Add a gl-info equivalent for OpenAL.

18 years agoTie the sun halo texture to point sprites.
ehofman [Tue, 31 Jan 2006 15:14:45 +0000 (15:14 +0000)]
Tie the sun halo texture to point sprites.

18 years agoVivian MEAZZA:
mfranz [Mon, 30 Jan 2006 22:08:53 +0000 (22:08 +0000)]
Vivian MEAZZA:

"Implement the repeat function for submodels. Default value is 'false' "

18 years agoAdd the fact that an absolute value will be used for log and log10.
ehofman [Mon, 30 Jan 2006 20:34:31 +0000 (20:34 +0000)]
Add the fact that an absolute value will be used for log and log10.

18 years agoOlaf Flebbe:
ehofman [Mon, 30 Jan 2006 13:44:56 +0000 (13:44 +0000)]
Olaf Flebbe:

I had massive problems with tower.cxx, so I decided to doublecheck all
containers and iterators in order and found lot of potential problems.

Now I can fly with "debug on" of KSFO for minutes without problems.

The fix for AIMgr is from Mathias, but I reworked it. So blame me, not him.

18 years agoDurk Talsma:
ehofman [Mon, 30 Jan 2006 13:29:49 +0000 (13:29 +0000)]
Durk Talsma:

- Feet to meter conversion mistake (in AI getGround elev)
- Improved ground following code (not yet perfect, but for now no one will
  notice it within the marginal altitiude differences at the taxitrack or
  runway)
- Exclusion of the "AI" directory witihin data/Aircraft in
  main/init/fgSearchAircraft, to prevent AI aircraft to be picked up by the
  aircraft search function

18 years agosupport optional setlistener() arg to fire callback function initially
mfranz [Mon, 30 Jan 2006 12:29:58 +0000 (12:29 +0000)]
support optional setlistener() arg to fire callback function initially

18 years agoMathias Fröhlich:
ehofman [Mon, 30 Jan 2006 10:54:10 +0000 (10:54 +0000)]
Mathias Fröhlich:

Jon Stockill wrote:
> I've just discovered that when using the null fdm I'm not getting
> updates to /position/ground-elev-m any more. So I can't actually
> retrieve the terrain elevation. Is there somewhere else in the property
> tree I could read this from?

The attached patch fixes this problem.

18 years agoMathias Fröhlich:
ehofman [Mon, 30 Jan 2006 10:50:28 +0000 (10:50 +0000)]
Mathias Fröhlich:

This patch limits the maximum simtime we do simulation computations for.
That is highly sensible if you run flightgear in valgrind or some realy slow
debug build. In such a case it is possible that flightgear gets totally
unresponsible, because simulation time might increase slower than real time.
That patch introduces a maximum simulation time per rendered frame to limit
that effect.

If the property /sim/max-simtime-per-frame is set to something strictly
positive, the simulation time is limited to that value.
The default is unchanged - no limit.

Anyway, from the point of view of gui responsiveness and responsiveness to
realtime controls like joystick inputs it might be a good idea to limit that
by default to say 1 second. If you have less than 1fps, flightgear is
unplayable anyway and I believe we do not longer need to care for realtime
correctness for that case ...

18 years agoon closing a dialog write the last position back, so that the caller can
mfranz [Sun, 29 Jan 2006 21:21:56 +0000 (21:21 +0000)]
on closing a dialog write the last position back, so that the caller can
open it there again next time

18 years agomake it consistent
mfranz [Sun, 29 Jan 2006 20:33:33 +0000 (20:33 +0000)]
make it consistent

18 years agoanother signal for listener callback functions:
mfranz [Sun, 29 Jan 2006 20:11:10 +0000 (20:11 +0000)]
another signal for listener callback functions:
setlistener("/sim/signals/quit", func { print("I'm called on exit.")});

18 years agoset the /sim/signals/reinit property on reinit, so that aircraft code can
mfranz [Sun, 29 Jan 2006 16:20:31 +0000 (16:20 +0000)]
set the /sim/signals/reinit property on reinit, so that aircraft code can
attach listeners. Other relevant stuff should go here, too. (/sim/crashed
could be one.)

18 years agodon't set zero width or height
mfranz [Sun, 29 Jan 2006 14:58:52 +0000 (14:58 +0000)]
don't set zero width or height

18 years agothe const-ness wasn't such a good idea :-)
mfranz [Sat, 28 Jan 2006 22:41:20 +0000 (22:41 +0000)]
the const-ness wasn't such a good idea :-)

18 years agorestore the uglyness: call the resize callback every frame again, even
mfranz [Sat, 28 Jan 2006 16:06:03 +0000 (16:06 +0000)]
restore the uglyness: call the resize callback every frame again, even
though the window wasn't resized. I'm just not adventurous enough for
a cleaner solution in the light of the upcoming release. At least the
xsize/ysize properties aren't set each frame any more, so listeners
work now properly.

18 years agonow that the renderer doesn't resize the window for every frame, we need
mfranz [Sat, 28 Jan 2006 13:38:28 +0000 (13:38 +0000)]
now that the renderer doesn't resize the window for every frame, we need
to do it manually here

18 years agowrite to /sim/startup/xsize only if the size really changed; likewise for
mfranz [Sat, 28 Jan 2006 10:40:47 +0000 (10:40 +0000)]
write to /sim/startup/xsize only if the size really changed; likewise for
ysize. This is useful for listeners, e.g. the one that will re-position the
FPS dialog.

18 years agoResizing is done when fg_os{,_sdl}.cxx says so, not in every single frame.
mfranz [Sat, 28 Jan 2006 00:10:39 +0000 (00:10 +0000)]
Resizing is done when fg_os{,_sdl}.cxx says so, not in every single frame.
(This defeats the purpose of the resize callback and makes "listening" to
the xsize useless.)

18 years agomake it const
mfranz [Fri, 27 Jan 2006 22:26:47 +0000 (22:26 +0000)]
make it const

18 years agoIgnore generated files
fredb [Fri, 27 Jan 2006 21:20:58 +0000 (21:20 +0000)]
Ignore generated files

18 years agoexport scenery click geo coords
mfranz [Fri, 27 Jan 2006 21:14:40 +0000 (21:14 +0000)]
export scenery click geo coords

18 years agogetChild(): add optional "create" flag (like in the c++ version)
mfranz [Fri, 27 Jan 2006 20:43:40 +0000 (20:43 +0000)]
getChild(): add optional "create" flag (like in the c++ version)

18 years agocosmetics: subc isn't really a sub-context here (it was where I copied it
mfranz [Fri, 27 Jan 2006 19:59:25 +0000 (19:59 +0000)]
cosmetics: subc isn't really a sub-context here (it was where I copied it
from)  -->  s/subc/c/

18 years agowe need a way to specify other contexts than _context; idea & OK by Andy
mfranz [Fri, 27 Jan 2006 19:51:25 +0000 (19:51 +0000)]
we need a way to specify other contexts than _context; idea & OK by Andy

18 years agoMelchior FRANZ:
ehofman [Fri, 27 Jan 2006 13:54:35 +0000 (13:54 +0000)]
Melchior FRANZ:

gui colors and material animation use "red", "green", "blue".
Would be nice if these light colors would use the same standard.

18 years agoallow top level <font> definition that applies to text in the whole dialog
mfranz [Fri, 27 Jan 2006 12:57:13 +0000 (12:57 +0000)]
allow top level <font> definition that applies to text in the whole dialog
and makes the layouter use this font's metrics

18 years agoTie the scene ambient,diffuse end specular red,green and blue components to a propert...
ehofman [Fri, 27 Jan 2006 12:35:15 +0000 (12:35 +0000)]
Tie the scene ambient,diffuse end specular red,green and blue components to a property under /rendering/scene

18 years agofonts are now delivered by FGFontCache (except for the splash screen,
mfranz [Fri, 27 Jan 2006 08:01:31 +0000 (08:01 +0000)]
fonts are now delivered by FGFontCache (except for the splash screen,
because this comes way before the gui subsystem is ready)

18 years agocosmetics
mfranz [Thu, 26 Jan 2006 22:22:37 +0000 (22:22 +0000)]
cosmetics

18 years agofix FGFontCache and make use of it (this allows to assign bitmap fonts
mfranz [Thu, 26 Jan 2006 21:40:37 +0000 (21:40 +0000)]
fix FGFontCache and make use of it (this allows to assign bitmap fonts
in dialogs)

18 years agoAdd support for point sprites.
ehofman [Thu, 26 Jan 2006 09:37:32 +0000 (09:37 +0000)]
Add support for point sprites.

18 years agocreate a new context for listener functions to avoid context corruption
mfranz [Thu, 26 Jan 2006 00:18:27 +0000 (00:18 +0000)]
create a new context for listener functions to avoid context corruption
(caused a crash when a timer triggered a listener) ... Idea and OK by Andy

18 years agoUpdates for newest scenery build.
curt [Wed, 25 Jan 2006 02:08:39 +0000 (02:08 +0000)]
Updates for newest scenery build.

18 years agoWhite space updates.
curt [Tue, 24 Jan 2006 17:13:48 +0000 (17:13 +0000)]
White space updates.

18 years agoNewest data file format includes range for each transmitter. Load that data
curt [Tue, 24 Jan 2006 17:13:28 +0000 (17:13 +0000)]
Newest data file format includes range for each transmitter.  Load that data
and use it instead of our own hard coded defaults.

18 years agoMathias Fröhlich:
ehofman [Tue, 24 Jan 2006 14:45:13 +0000 (14:45 +0000)]
Mathias Fröhlich:

Incorporating the shared ptr code:
- All scenegraph references from SimGear
- SGMaterial which already had a reference counter uses now that common
  infrastructure.
- SGMatModel is now counted.
- SGSoundSample from SimGear
- And the corresponding change for the sound samples in flightgear which fixes
  a latent crash if FGBeacon would evern be deleted.

18 years agoUse smooth (aliased) lines and points when available in hardware.
ehofman [Sat, 21 Jan 2006 12:07:40 +0000 (12:07 +0000)]
Use smooth (aliased) lines and points when available in hardware.

18 years agoOlaf Flebbe: MSVC fix.
ehofman [Sat, 21 Jan 2006 10:11:08 +0000 (10:11 +0000)]
Olaf Flebbe: MSVC fix.

18 years agoJim Wilson:
ehofman [Sat, 21 Jan 2006 10:06:09 +0000 (10:06 +0000)]
Jim Wilson:

This patch fixes configure.ac so that when an alternate prefix location for simgear is specified, configure tests for the header file jpegfactory.hxx in that specific location only.

18 years agoRemove a leftover from a previous patch.
ehofman [Sat, 21 Jan 2006 10:03:36 +0000 (10:03 +0000)]
Remove a leftover from a previous patch.

18 years agoSync. w. JSBSim
ehofman [Fri, 20 Jan 2006 17:50:45 +0000 (17:50 +0000)]
Sync. w. JSBSim

18 years agoIntegrate a large part of John Wojnaroski's 747 cockpit project.
ehofman [Fri, 20 Jan 2006 17:19:01 +0000 (17:19 +0000)]
Integrate a large part of John Wojnaroski's 747 cockpit project.

Erik Hofman:
This patch contains an update to net_ctrls.hxx that adds an extra 100 bytes
(or an equivalent of 25 (u)int32_t types) of reserved space. This could be
used to make the protocol forward and backward compatibel within a certain
scope. Be sure to read the instructions at the begining of the header file
when addinf new variables.

18 years agoMake sure the path to simgear and plib get added to the include path before anything...
ehofman [Sat, 14 Jan 2006 10:16:28 +0000 (10:16 +0000)]
Make sure the path to simgear and plib get added to the include path before anything else.

18 years agoFix the crashes on mini-panel and panel-reload with the KLN89, by removing the multip...
daveluff [Fri, 13 Jan 2006 22:07:50 +0000 (22:07 +0000)]
Fix the crashes on mini-panel and panel-reload with the KLN89, by removing the multiple inheritance and using an empty shell instrument to forward draw commands from the panel to the complex gps subsystem.

18 years agoRemove the filtersjb left over
fredb [Fri, 13 Jan 2006 14:33:07 +0000 (14:33 +0000)]
Remove the filtersjb left over

18 years agoRemove the filtersjb left over
fredb [Fri, 13 Jan 2006 14:00:13 +0000 (14:00 +0000)]
Remove the filtersjb left over

18 years agoHide some user generated files.
ehofman [Fri, 13 Jan 2006 09:50:58 +0000 (09:50 +0000)]
Hide some user generated files.

18 years agoRemove a bunch of unneede files.
ehofman [Fri, 13 Jan 2006 09:49:09 +0000 (09:49 +0000)]
Remove a bunch of unneede files.

18 years agoTypo
fredb [Thu, 12 Jan 2006 21:35:59 +0000 (21:35 +0000)]
Typo

18 years agoVariable 'thrust' used before being initialized
fredb [Thu, 12 Jan 2006 21:35:32 +0000 (21:35 +0000)]
Variable 'thrust' used before being initialized

18 years agoRedundant loop variable declaration
fredb [Thu, 12 Jan 2006 21:35:08 +0000 (21:35 +0000)]
Redundant loop variable declaration

18 years agoDon't declare functions that are not implemented
fredb [Thu, 12 Jan 2006 21:34:36 +0000 (21:34 +0000)]
Don't declare functions that are not implemented

18 years agosync. with JSBSim v. 2.0
ehofman [Thu, 12 Jan 2006 15:04:22 +0000 (15:04 +0000)]
sync. with JSBSim v. 2.0

18 years agoDon't display the menubar in e highres screenshot.
ehofman [Wed, 11 Jan 2006 10:53:04 +0000 (10:53 +0000)]
Don't display the menubar in e highres screenshot.

18 years agoAnother attempt at fixing a cygwin compiler error.
curt [Tue, 10 Jan 2006 15:23:10 +0000 (15:23 +0000)]
Another attempt at fixing a cygwin compiler error.

18 years agoFix a compiler warning (member initialization order in constructor.)
curt [Tue, 10 Jan 2006 15:06:17 +0000 (15:06 +0000)]
Fix a compiler warning (member initialization order in constructor.)

18 years agoSmall change to avoid dangling audio resources.
curt [Tue, 10 Jan 2006 14:51:12 +0000 (14:51 +0000)]
Small change to avoid dangling audio resources.

18 years agoCygwin fix.
curt [Tue, 10 Jan 2006 02:49:52 +0000 (02:49 +0000)]
Cygwin fix.

18 years agoAdd an srand() function to nasal (hooked into sg_srandom_time()).
curt [Mon, 9 Jan 2006 15:29:24 +0000 (15:29 +0000)]
Add an srand() function to nasal (hooked into sg_srandom_time()).
This helps avoid the problem of always getting the same random number
sequence every time you run a script.

18 years agoAdd a directory listing command to nasal. Given a directory name, return
curt [Mon, 9 Jan 2006 03:48:14 +0000 (03:48 +0000)]
Add a directory listing command to nasal.  Given a directory name, return
the file list as an array.

18 years agoWrite 'fg_root' in the property tree at /sim/fg-root
curt [Mon, 9 Jan 2006 03:47:33 +0000 (03:47 +0000)]
Write 'fg_root' in the property tree at /sim/fg-root

18 years agoExpose the "play-audio-command" through the props/telnet interface.
curt [Mon, 9 Jan 2006 02:23:50 +0000 (02:23 +0000)]
Expose the "play-audio-command" through the props/telnet interface.

18 years agoAdd a do_play_audio_message() function which is tied to the
curt [Mon, 9 Jan 2006 02:23:02 +0000 (02:23 +0000)]
Add a do_play_audio_message() function which is tied to the
"play-audio-function" command.  This function can be called from internal
code, from nasal scripts, or from external scripts to play a single one-off
wav file.  File/audio data is loaded and unloaded on the fly automatically.

18 years agoAdditional functionality added to the FlightGear sound fx manager.
curt [Mon, 9 Jan 2006 02:21:04 +0000 (02:21 +0000)]
Additional functionality added to the FlightGear sound fx manager.

Create a queue of one-off wav files.  Calling layer can request the system
to play any wav file.  The request is thrown at the end of a play queue.
The queue is serviced sequentially so that queued wav files will no longer
overlap.  When a sample is finished playing. It is removed from the queue
and deleted from memory.  The next remaining request in the queue is then
played.

18 years agoStuart Buchanan:
ehofman [Fri, 6 Jan 2006 09:50:58 +0000 (09:50 +0000)]
Stuart Buchanan:

- Provide a Nasal interface to display simple text messages on the screen
  like the ATC display. In fact, I copied the code from the ATCDisplay.cxx
  and simply shifted it further down the screen.

Erik:

TODO: Integrate the two pieces of code.

18 years agoTry to prevent a potential buffer overflow for WIN33.
ehofman [Wed, 4 Jan 2006 13:25:56 +0000 (13:25 +0000)]
Try to prevent a potential buffer overflow for WIN33.

18 years agoCygwin fixes.
ehofman [Wed, 4 Jan 2006 12:49:07 +0000 (12:49 +0000)]
Cygwin fixes.

18 years agoMinor oops, committed this file before saving the final version in my
curt [Mon, 2 Jan 2006 23:08:45 +0000 (23:08 +0000)]
Minor oops, committed this file before saving the final version in my
editor.

18 years ago- Add an "is-valid" property node so other modules can do a quick check if
curt [Mon, 2 Jan 2006 23:01:45 +0000 (23:01 +0000)]
- Add an "is-valid" property node so other modules can do a quick check if
  anything in the nav tree is valid or not.
- Fix an order problem between caching data values and searching for a new
  station that could cause odd and unexpected and hard to reproduce results.

18 years agoOlaf Flebbe: Make sure the model gets deleted when it's not referenced anymore.
ehofman [Fri, 30 Dec 2005 09:34:29 +0000 (09:34 +0000)]
Olaf Flebbe: Make sure the model gets deleted when it's not referenced anymore.

18 years agoRemove some left over debugging output.
curt [Thu, 29 Dec 2005 16:50:03 +0000 (16:50 +0000)]
Remove some left over debugging output.

18 years agoJohn Ellson:
curt [Thu, 29 Dec 2005 16:22:38 +0000 (16:22 +0000)]
John Ellson:

Changes require to compile with a Fedora-Core Development system running
on an X86_64 platform with gcc-4.1.0.

18 years agoDurk Talsma, Olaf Flebbe & Mathias Fröhlich:
ehofman [Thu, 29 Dec 2005 13:58:21 +0000 (13:58 +0000)]
Durk Talsma, Olaf Flebbe & Mathias Fröhlich:

Split up simple.cxx

18 years agoBetter estimate of cross track heading error. Also, make sure it is
curt [Thu, 29 Dec 2005 04:07:49 +0000 (04:07 +0000)]
Better estimate of cross track heading error.  Also, make sure it is
computed correctly when flying outbound from a vor.