]> git.mxchange.org Git - flightgear.git/log
flightgear.git
19 years agoVarious tweaks for v0.9.5-pre1.
curt [Thu, 15 Jul 2004 18:07:02 +0000 (18:07 +0000)]
Various tweaks for v0.9.5-pre1.

19 years agoAdd a section about subsystems.
ehofman [Thu, 15 Jul 2004 17:09:29 +0000 (17:09 +0000)]
Add a section about subsystems.

19 years agoAdd some introduction documentation.
ehofman [Mon, 12 Jul 2004 12:11:07 +0000 (12:11 +0000)]
Add some introduction documentation.

19 years agoFrederic Bouvier:
ehofman [Sat, 3 Jul 2004 10:59:34 +0000 (10:59 +0000)]
Frederic Bouvier:

I have 3 issues that are fixed by this set of patches.

1. In extensions.cxx
  #else if !defined( WIN32 )  must be changed by
  #elif !defined( WIN32 ) because the text after #else
  seems to be ignored

2. banner is not available on windows, only cygwin

3. ANSI escape sequences are not interpreted on the
   windows console. We just have garbage that is hard
   to read.

19 years agoSync. with JSBSim CVS again. This fixes some small issues.
ehofman [Sun, 27 Jun 2004 19:35:54 +0000 (19:35 +0000)]
Sync. with JSBSim CVS again. This fixes some small issues.

20 years agoOn a reinit() subsystem event, reconfigure the joystick and other
david [Thu, 24 Jun 2004 02:08:38 +0000 (02:08 +0000)]
On a reinit() subsystem event, reconfigure the joystick and other
inputs.  This makes it possible to hot-plug a new joystick while
FlightGear is running (as long as the user explicitly reinits the
input subsystem).

20 years agoWhen searching for nav records ignore stations > 300nm away.
curt [Sun, 20 Jun 2004 18:58:44 +0000 (18:58 +0000)]
When searching for nav records ignore stations > 300nm away.

20 years agoIgnore generated files.
david [Tue, 15 Jun 2004 12:48:22 +0000 (12:48 +0000)]
Ignore generated files.

20 years agoAlternator should still put out some output when RPM < 800. (Yes I know this
curt [Mon, 14 Jun 2004 21:36:10 +0000 (21:36 +0000)]
Alternator should still put out some output when RPM < 800.  (Yes I know this
is hard coded for the C172, but so far no one has asked to do this more
generically.)

20 years agoPending white space tweaks I forgot to commit the other day.
curt [Mon, 14 Jun 2004 18:47:21 +0000 (18:47 +0000)]
Pending white space tweaks I forgot to commit the other day.

20 years ago- Track a saved mouse pointer location from the previous frame so we can detect
curt [Mon, 14 Jun 2004 18:46:58 +0000 (18:46 +0000)]
- Track a saved mouse pointer location from the previous frame so we can detect
  motion.

- Track a timeout value so we can optionally hide the mouse pointer after
  some user specified timeout period.

- in doMouseMotion() always update m.x and m.y even if we return early because
  pui wanted the event.  Without this, we can't reliably detect motion vs.
  inactivity.

- in _update_mouse() add a dt parameter so we can decriment the timeout value
  in "real" time.

- in _update_mouse() optionally hide the mouse pointer if m.timeout goes to
  zero.  Restore the pointer (and the timeout counter) if the mouse is moved.

20 years agoChange the fps toggle property name a bit and add a toggle switch in the
curt [Mon, 14 Jun 2004 16:38:44 +0000 (16:38 +0000)]
Change the fps toggle property name a bit and add a toggle switch in the
gui/dialog/rendering box.

20 years agoMake the onscreen fps display toggleable via a boolean property.
curt [Mon, 14 Jun 2004 16:25:32 +0000 (16:25 +0000)]
Make the onscreen fps display toggleable via a boolean property.

20 years agoSync w. JSBSim CVS
ehofman [Mon, 14 Jun 2004 11:40:45 +0000 (11:40 +0000)]
Sync w. JSBSim CVS

20 years agoMelchior FRANZ:
ehofman [Sun, 13 Jun 2004 18:47:55 +0000 (18:47 +0000)]
Melchior FRANZ:

Remove an unused test, counter_hack is initialized to 0 and the test never reaches the increment section but instead will always call loader.update();

20 years agoWork around a potential never ending loop problem. A proper fix will be in the next...
ehofman [Sun, 13 Jun 2004 18:44:21 +0000 (18:44 +0000)]
Work around a potential never ending loop problem. A proper fix will be in the next JSBSim update.

20 years agoPut the sqrt() back in. That was part of another attempt to optimize the code, but...
ehofman [Sat, 12 Jun 2004 11:37:47 +0000 (11:37 +0000)]
Put the sqrt() back in. That was part of another attempt to optimize the code, but it hadn't fully matured yet.

20 years agoNo need to do he calculations twice. FGAIBAse::update() handles them already.
ehofman [Sat, 12 Jun 2004 11:34:05 +0000 (11:34 +0000)]
No need to do he calculations twice. FGAIBAse::update() handles them already.

20 years agoMove the radar update code to the AIBase class. There seems to be a problem where...
ehofman [Fri, 11 Jun 2004 13:49:07 +0000 (13:49 +0000)]
Move the radar update code to the AIBase class. There seems to be a problem where all targets disappear whenever one of them disappears, but that was present in the previous update also.

20 years agoRemove an unnecessary property set.
ehofman [Fri, 11 Jun 2004 08:09:55 +0000 (08:09 +0000)]
Remove an unnecessary property set.

20 years agoOops, fix a bug (introduced recently) that could cause the glide slope
curt [Thu, 10 Jun 2004 19:51:27 +0000 (19:51 +0000)]
Oops, fix a bug (introduced recently) that could cause the glide slope
indicator to read incorrectly for many approaches.

20 years agoIntroduce a radar/in-range boolean property, end check this to see whether the CPU...
ehofman [Thu, 10 Jun 2004 19:14:19 +0000 (19:14 +0000)]
Introduce a radar/in-range boolean property, end check this to see whether the CPU intensive radar calculations need to be done. If not, skip them.

20 years agoDME units report a distance based on the assumption that the ground station
curt [Wed, 9 Jun 2004 20:21:18 +0000 (20:21 +0000)]
DME units report a distance based on the assumption that the ground station
will delay it's reply by 50ms.  The ground station can change it's reply delay
to trick the airborn dme unit into reporting a distance that is offset from
the true distance by some constant value.  In FG we model this by subtracting
a fixed distance from the actual distance.

It is thus possible in our implimentation for the displayed distance to become
negative.  This patch clamp DME distance to a minimum value of 0.00 so it can
never go negative.

20 years agoOften, the elevation of an ILS component is not listed in our nav database.
curt [Wed, 9 Jun 2004 03:13:13 +0000 (03:13 +0000)]
Often, the elevation of an ILS component is not listed in our nav database.
A good elevation is critical for proper glide slope modeling.  This patch
assigns the average field elevation to any ILS component that doesn't have
a valid elevation.

Also, for an ILS approach, use the GS transmitter elevation for glide slope
calculations rather than the localizer elevation, in some cases this can
make a big difference.

20 years agoiSwitch position changes.
curt [Wed, 9 Jun 2004 00:49:04 +0000 (00:49 +0000)]
iSwitch position changes.

20 years agoMelchior FRANZ:
ehofman [Tue, 8 Jun 2004 15:32:09 +0000 (15:32 +0000)]
Melchior FRANZ:

Wouldn't it be better to prepare the whole list of paths (or two
separate ones for Terrain/Objects if necessary) in FGGlobals::set_fg_scenery,
and to pass the vector<string>s to FGTileEntry::load? It doesn't seem to make
a lot of sense to split the path up, modify it, mount it together to one string
again, and then let FGTileEntry::load split it up again.

Here we go:

Main/globals.cxx
================
As fg_scenery is now a string_list, we don't need initialization. Furthermore,
this list is cleared with every set_fg_scenery() call.

ctor: create default dir from fg_root if necessary. Otherwise check all paths
of --fg-scenery/FG_SCENERY: If the path doesn't exist, ignore it. If it contains
a dir Terrain and/or Objects, then only add that to the list. If it contains
neither, then use the path as is.

Scenery/tileentry.cxx
=====================
Trivial: don't split a "base path", but use the given path_list as is.
(I considered a variable name "path_list" better suited than "search".)

Scenery/FGTileLoader.cxx
========================
No more fiddling with sub-paths. This has to be delivered by get_fg_scenery
already.

20 years agoRemove some debugging code.
ehofman [Mon, 7 Jun 2004 09:55:55 +0000 (09:55 +0000)]
Remove some debugging code.

20 years agoMake use of the new SGPath::add() function and automatically append 'Terren' and...
ehofman [Mon, 7 Jun 2004 09:52:11 +0000 (09:52 +0000)]
Make use of the new SGPath::add() function and automatically append 'Terren' and 'Objects' the the scenery path when one or both of those subdirectories exsist.

20 years agoWindows uses ';' instead of ':' as a path separator.
ehofman [Sun, 6 Jun 2004 19:34:31 +0000 (19:34 +0000)]
Windows uses ';' instead of ':' as a path separator.

20 years agoSearch in share/FlightGear instead of lib/FlightGear
ehofman [Sun, 6 Jun 2004 19:31:42 +0000 (19:31 +0000)]
Search in share/FlightGear instead of lib/FlightGear

20 years agoMake it possible to split the Scenery into Scenery/Terrain and Scenery/Objects in...
ehofman [Sun, 6 Jun 2004 19:15:04 +0000 (19:15 +0000)]
Make it possible to split the Scenery into Scenery/Terrain and Scenery/Objects in preparation for the next scenery release.

20 years agoIncorporate some of the changes from the Linspire diff.
ehofman [Sun, 6 Jun 2004 14:28:45 +0000 (14:28 +0000)]
Incorporate some of the changes from the Linspire diff.

20 years agoDavid Culp:
ehofman [Sun, 6 Jun 2004 08:50:17 +0000 (08:50 +0000)]
David Culp:

Here's some additions to AI that allow refueling from an AI tanker (the actual
onload of fuel must be handled by the user's FDM of course, this just lets
the FDM know that the user is in position to refuel).

I've added a new class of AIAircraft called "tanker".  It uses the same
performance struct as a jet transport.  An AI tanker is just like an AI jet
transport, except it uses the already-existing radar data to control the
boolean property systems/refuel/contact.  The code change was minimal.

An AI tanker can be created like this:

  <entry>
   <callsign>Esso 1</callsign>
   <type>aircraft</type>
   <class>tanker</class>
   <model>Aircraft/737/Models/boeing733.xml</model>
   <latitude>37.61633</latitude>
   <longitude>-122.38334</longitude>
   <altitude>3000</altitude>
   <heading>020</heading>
   <speed>280</speed>
   <roll>-15</roll>
  </entry>

This puts a tanker over KSFO at 3000 feet, in a left-hand orbit.  When the
user gets within refueling range (contact position) then the property
systems/refuel/contact will be true.  Otherwise it is false.

The dimensions of the refueling envelope are pretty rough right now, but still
usable.  The user must be behind the tanker (ie. radar y_offset > 0).  The
user must be at or below the tanker's altitude (ie. radar elevation > 0).
The user's lat/lon must be within 250 feet of the tanker's lat/lon (ie. radar
range_ft < 250).  This last requirement is loose because the radar data is
only updated every 100 ms, which is accurate enough for radar use, but
which is sloppy for air refueling.  This could be tightened up by increasing
the radar update rate to once every sim cycle.

I'm going to add a light to the T-38 instrument panel that will monitor the
property systems/refuel/contact.  This will make it easier to explore the
boundaries of the refueling envelope.

20 years agoMSVC fixes
ehofman [Sun, 6 Jun 2004 08:35:09 +0000 (08:35 +0000)]
MSVC fixes

20 years agoChange the default rsync module to Scenery-0.9.5
curt [Sat, 5 Jun 2004 18:46:36 +0000 (18:46 +0000)]
Change the default rsync module to Scenery-0.9.5

20 years agoIngnore some generated files."
ehofman [Sat, 5 Jun 2004 08:51:29 +0000 (08:51 +0000)]
Ingnore some generated files."

20 years agoAdd the AIModel based air traffic subsystem from Durk Talsma.
ehofman [Thu, 3 Jun 2004 17:59:14 +0000 (17:59 +0000)]
Add the AIModel based air traffic subsystem from Durk Talsma.

20 years agoUpdates ias requested by Melchior.
ehofman [Tue, 1 Jun 2004 13:29:04 +0000 (13:29 +0000)]
Updates ias requested by Melchior.

20 years agoSome (small) modifications.
ehofman [Tue, 1 Jun 2004 11:42:05 +0000 (11:42 +0000)]
Some (small) modifications.

20 years agoUpdates.
ehofman [Tue, 1 Jun 2004 08:36:27 +0000 (08:36 +0000)]
Updates.

20 years agoDavid Culp:
ehofman [Sat, 29 May 2004 11:39:10 +0000 (11:39 +0000)]
David Culp:

Here's some new AI stuff.

1)  AI objects must now be defined in a scenario file, not in preferences.xml
or a *-set file.  (Of course this doesn't prevent objects from being created
dynamically, as with Durk's traffic manager).

2)  A new demo_scenario file is attached.  It creates 3 aircraft, a sailboat,
and a thunderstorm.

3)  Objects without flightplans live forever.

4)  FGAIShip::ProcessFlightplan() is not yet implemented.

5)  preferences.xml should now define only <enabled> and <scenario>

20 years agoRemove some left over debugging output.
curt [Fri, 28 May 2004 21:47:11 +0000 (21:47 +0000)]
Remove some left over debugging output.

20 years agoAllow a "threshold" value to determine which localizers to snap to the
curt [Fri, 28 May 2004 20:57:05 +0000 (20:57 +0000)]
Allow a "threshold" value to determine which localizers to snap to the
runway heading or not.

20 years agoFix a gxx 3.3 compiler problem.
ehofman [Fri, 28 May 2004 19:03:55 +0000 (19:03 +0000)]
Fix a gxx 3.3 compiler problem.

20 years agoCurt Olson:
curt [Fri, 28 May 2004 16:24:43 +0000 (16:24 +0000)]
Curt Olson:

These change add some code that at initialization time will snap all
localizers into perfect alignment with their runways.  It's my experience
that the DAFIF/FAA data reports runway and localizer headings to a level
of precision that is great for making charts, or adjusting your OBS, etc.
But the level of precision of this data can be far enough off to make you
visibly *un*aligned with the runway when the CDI needle is centered.

There are probably cases where the localizer isn't really perfectly
aligned with the runway, or intentionally misaligned to avoid obstacles
or terrain.  So I have made this configurable for those that trust the
data more than I do.  Just set "/sim/navdb/auto-align-localizers" to
true/false in the preferences file to turn this feature on or off in the
code.

20 years agoUse SGRawValueMethods rather than SGRawValueFunctions because that allows one to...
ehofman [Fri, 28 May 2004 08:46:33 +0000 (08:46 +0000)]
Use SGRawValueMethods rather than SGRawValueFunctions because that allows one to pass the 'this' pointer. Remove some debug code.

20 years agoAnd again ...
curt [Fri, 28 May 2004 05:33:10 +0000 (05:33 +0000)]
And again ...

20 years agoOne more ...
curt [Fri, 28 May 2004 05:31:11 +0000 (05:31 +0000)]
One more ...

20 years agoOops, and another similar one.
curt [Fri, 28 May 2004 05:29:04 +0000 (05:29 +0000)]
Oops, and another similar one.

20 years agoFix a compile error I missed in the first round.
curt [Fri, 28 May 2004 05:27:40 +0000 (05:27 +0000)]
Fix a compile error I missed in the first round.

20 years agoThis set of changes impliments the following:
curt [Fri, 28 May 2004 05:24:54 +0000 (05:24 +0000)]
This set of changes impliments the following:

- FG now directly supports Robin's native nav database file format.
- His latest data now separates out dme, gs, loc, and marker beacon
  transmitters rather than lumping them all into a single "ILS" record.
- These new data structure changes prompted me to do some code restructuring
  so that internally these different types of navaids are all kept as
  separate lists and searched and handled separately.
- This structural change had a cascading affect on any code that
  references or uses the nav databases.  I've gone and "touched" a lot of
  nav related code in a lot of places.
- As an added bonus, the new data (and code) adds DME bias so these will
  all now read as they do in real life.

- Added Navaids/navdb.cxx and Navaids/navdb.hxx which provide a front
  end loaders for the nav data.
- Added Navaids/navrecord.hxx which is a new "generic" nav data record.
- Removed Navaids/ils.hxx, Navaids/ilslist.cxx, Navaids/ilslist.hxx,
  Navaids/mkrbeacons.cxx, and Navaids/mkrbeacons.hxx which are all now
  depricated.

20 years agoAdd the this pointer to the tied function calls. This makes it possible to make a...
ehofman [Thu, 27 May 2004 13:16:53 +0000 (13:16 +0000)]
Add the this pointer to the tied function calls. This makes it possible to make a distinction between the different aircraft models.

20 years agoMove navaids and fixes out of "global" name space into the FGGlobals
curt [Wed, 26 May 2004 18:15:19 +0000 (18:15 +0000)]
Move navaids and fixes out of "global" name space into the FGGlobals
structure.

20 years agoAdd simple icing state.
curt [Wed, 26 May 2004 18:13:34 +0000 (18:13 +0000)]
Add simple icing state.

20 years agoUpdate fix management code to read Robin's native fix.dat format.
curt [Wed, 26 May 2004 16:40:27 +0000 (16:40 +0000)]
Update fix management code to read Robin's native fix.dat format.

20 years agoDon't inline static functions.
ehofman [Tue, 25 May 2004 08:58:36 +0000 (08:58 +0000)]
Don't inline static functions.

20 years agoIn the situation where a class gets deleted, the _slef pointer is set to 0. Make...
ehofman [Mon, 24 May 2004 14:36:35 +0000 (14:36 +0000)]
In the situation where a class gets deleted, the _slef pointer is set to 0. Make sure we check for that situation before using it.

20 years agoDavid Culp:
ehofman [Fri, 21 May 2004 16:50:19 +0000 (16:50 +0000)]
David Culp:

1.  Removed aircraft roll on ground.
2.  Decreased descent pitch angle.
3.  Updated flightplans to include <on-ground>
4.  Fixed property indexing, so all AI aircraft have their own property branch

The default value of <on-ground> is false, so you only need to specify it when
on the ground.  For takeoff you need to specify <on-ground>true</on-ground>
for the first waypoint, and for the acceleration waypoint.  For landing you
need to specify it for the touchdown point and any taxi points.

One problem.  WARNING **** There is a bug in the way the property system
works, which causes a segfault, but I don't know if the problem is in the
property code, or in how I'm using it.  After an AI object terminates, if you
access the property tree through the property browser the sim will segfault.

20 years agoUpdates for the bump-mapped 2d cloud code.
ehofman [Fri, 21 May 2004 14:57:42 +0000 (14:57 +0000)]
Updates for the bump-mapped 2d cloud code.

20 years agoInitialize the bump-mapped 3d clouds to true.
ehofman [Thu, 20 May 2004 14:13:36 +0000 (14:13 +0000)]
Initialize the bump-mapped 3d clouds to true.

20 years agoAdd a property to be able to turn the bump-mapped cloud code on and off at runtime.
ehofman [Thu, 20 May 2004 13:27:40 +0000 (13:27 +0000)]
Add a property to be able to turn the bump-mapped cloud code on and off at runtime.

20 years agoOops, change type.
ehofman [Wed, 19 May 2004 14:10:31 +0000 (14:10 +0000)]
Oops, change type.

20 years agoAdd AI models enableing/disableing command line option and support code.
ehofman [Wed, 19 May 2004 13:55:49 +0000 (13:55 +0000)]
Add AI models enableing/disableing command line option and support code.

20 years agoReduce the update delay from 500 ms to 100 ms.
ehofman [Wed, 19 May 2004 08:12:21 +0000 (08:12 +0000)]
Reduce the update delay from 500 ms to 100 ms.

20 years agoFred found some iterated related problems. With a bit of luck this solves the segment...
ehofman [Wed, 19 May 2004 08:01:02 +0000 (08:01 +0000)]
Fred found some iterated related problems. With a bit of luck this solves the segmentation fault problem.

20 years agoSome small updates that hopefully solve the segmentation fault at the end of the...
ehofman [Tue, 18 May 2004 09:09:08 +0000 (09:09 +0000)]
Some small updates that hopefully solve the segmentation fault at the end of the flightplan.

20 years agoFix a bug with propeller gearing. The torque computations weren't
andy [Tue, 18 May 2004 01:46:36 +0000 (01:46 +0000)]
Fix a bug with propeller gearing.  The torque computations weren't
properly correcting for the gear ratio.  This is the source of the
problems Vivian and Jim were having with the Spitfire and Mustang.

20 years agoDavid Culp:
ehofman [Mon, 17 May 2004 08:45:33 +0000 (08:45 +0000)]
David Culp:

First, preferences.xml will define the scenario filename.

For now, the other way of defining ai objects still works, so the sailboat
stays in preferences.xml.  Later, I'll move the sailboat into the demo
scenario.  If no scenario filename is given, then no scenario will be
processed.

I changed the demo scenario to create two 737's, one takes off on runway 01L,
and the other takes off on runway 01R.  This will make a good demo for the ai
system.  One problem, if you takeoff on 28L/R right away, you might run into
the taking-off 737's, or be scared.

20 years agoDon't set values on "live" text objects to prevent confusion when a
andy [Sat, 15 May 2004 21:41:42 +0000 (21:41 +0000)]
Don't set values on "live" text objects to prevent confusion when a
dialog-apply happens.

20 years agoFix a bug where the interpreter could read a garbage buffer from the stack.
andy [Sat, 15 May 2004 21:40:58 +0000 (21:40 +0000)]
Fix a bug where the interpreter could read a garbage buffer from the stack.

20 years agoUse lower case letters instead.
ehofman [Sat, 15 May 2004 12:46:25 +0000 (12:46 +0000)]
Use lower case letters instead.

20 years agoDavid Culp:
ehofman [Sat, 15 May 2004 09:07:55 +0000 (09:07 +0000)]
David Culp:

Here's the newest AI stuff.

The AIManager at init() creates a new scenario.  Right now the
default_scenario is hard coded in, but eventually the AIManager should get
the scenario filename from preferences.xml.

The scenario defines which AI objects will be created.  Right now it only
creates AIAircraft, but this is easily extended.  The scenario also defines
which flightplan will be assigned to the airplane.  Scenario config files go
in data/Data/AI.

The Airplane gets a pointer to a FlightPlan object.  Each airplane should get
its own flightplan object, even if two airplanes have the same flight plan.
This is because  the flightplan maintains the iterator pointing to the
current waypoint, and two airplanes might be at different locations (for
instance if they were created at different times).  The flight plan files go
in data/Data/AI/FlightPlans.

When the airplane gets to the waypoint named "END" it vanishes.  The
AIAircraft destructor deletes its flight plan (if it has one).

The last waypoint is a place holder only.  I called mine
<WPT><NAME>"EOF"</NAME></WPT>.

20 years agoFix the slider to request a non-zero length, and make its width a
andy [Fri, 14 May 2004 17:16:35 +0000 (17:16 +0000)]
Fix the slider to request a non-zero length, and make its width a
little larger.

The text widget can now be meaningfully associated with a property; in
PUI, it's "value" isn't the same thing as its label, but we can hack
things to treat them symmetrically.

Commit an experimental "live" property that can be set on widgets to
cause them to update their values every frame.  This works great for
text widgets, as above.  Note that this synchronization is input-only:
no support is provided (or needed -- the GUI only changes when the
user does something) for writing those properties out every frame.

20 years agoBetter support for pausing/unpausing sound.
curt [Fri, 14 May 2004 15:51:43 +0000 (15:51 +0000)]
Better support for pausing/unpausing sound.
Add support for controlling global volume of FlightGear.

20 years agomain.cxx - small tweak to sound updating logic.
curt [Fri, 14 May 2004 15:50:28 +0000 (15:50 +0000)]
main.cxx - small tweak to sound updating logic.
options.cxx - change property name /sim/sound/audible to /sim/sound/pause

20 years agoCode at this level shouldn't care if the sound manager is paused or not.
curt [Fri, 14 May 2004 15:49:10 +0000 (15:49 +0000)]
Code at this level shouldn't care if the sound manager is paused or not.
Just forge ahead as if sound is playing, let the sound manger worry about
the details.

20 years agoAdd some floating point (exception) optimizations for IRIX machines.
ehofman [Fri, 14 May 2004 11:58:57 +0000 (11:58 +0000)]
Add some floating point (exception) optimizations for IRIX machines.

20 years agoAdd a tool to test the integrity of the animation files by Josh Babcock.
ehofman [Fri, 14 May 2004 11:54:06 +0000 (11:54 +0000)]
Add a tool to test the integrity of the animation files by Josh Babcock.

20 years agoAdd -framework AGL which is required after recent gui code changes.
curt [Fri, 14 May 2004 02:05:51 +0000 (02:05 +0000)]
Add -framework AGL which is required after recent gui code changes.

20 years agoOops, need config.h before pu.h or else the build will fail on
andy [Wed, 12 May 2004 20:51:47 +0000 (20:51 +0000)]
Oops, need config.h before pu.h or else the build will fail on
glut-less systems.

20 years agoGUI layout management and a few visual/eye-candy modifications. See
andy [Wed, 12 May 2004 15:36:07 +0000 (15:36 +0000)]
GUI layout management and a few visual/eye-candy modifications.  See
DOCS/README.layout in the base package for details, along with the
modified dialog files.

20 years agoClean up some sound buffer allocation/deallocation issues.
curt [Mon, 10 May 2004 21:24:30 +0000 (21:24 +0000)]
Clean up some sound buffer allocation/deallocation issues.

20 years agoFix from Melchior: It replaces the ridiculous 5 seconds by
andy [Thu, 6 May 2004 16:28:08 +0000 (16:28 +0000)]
Fix from Melchior: It replaces the ridiculous 5 seconds by
the 30 seconds that Maik had originally intended, according to the comment.
This is important for the pending sound and rotor disc changes (and of course
for realism).

20 years agoAdd a macintosh identifier for joystick axis numbering support.
ehofman [Thu, 6 May 2004 16:07:30 +0000 (16:07 +0000)]
Add a macintosh identifier for joystick axis numbering support.

20 years agoRoy Vegard Ovesen:
ehofman [Thu, 6 May 2004 09:29:26 +0000 (09:29 +0000)]
Roy Vegard Ovesen:
Fix the leg distance calculation to display nautical miles instead of meters.

It turns out that Simgear already has a range normalize function, so I use
that one instead.

20 years agoAdd /sim/initialised which is set to true when we're ready to go.
ehofman [Wed, 5 May 2004 09:01:21 +0000 (09:01 +0000)]
Add /sim/initialised which is set to true when we're ready to go.

20 years agoAdd calls to jsInit() which is a noop on all platforms but Mac.
curt [Tue, 4 May 2004 19:02:19 +0000 (19:02 +0000)]
Add calls to jsInit() which is a noop on all platforms but Mac.

20 years agoGUI windows are now draggable. This missing feature has annoyed me
andy [Mon, 3 May 2004 00:40:50 +0000 (00:40 +0000)]
GUI windows are now draggable.  This missing feature has annoyed me
for a while, it turned out to be pretty easy to implement.  Also, the
property picker is now non-modal, I presume the modality wasn't an
intentional feature.

20 years agoReverse the sense of manual propellers. Low numbers == fast
andy [Sat, 1 May 2004 15:18:27 +0000 (15:18 +0000)]
Reverse the sense of manual propellers.  Low numbers == fast
propeller, silly.

20 years agoOops, I broke engine starting.
andy [Sat, 1 May 2004 14:30:00 +0000 (14:30 +0000)]
Oops, I broke engine starting.

20 years agoRoy Vegard Ovesen:
ehofman [Sat, 1 May 2004 09:40:09 +0000 (09:40 +0000)]
Roy Vegard Ovesen:

I've added a vertical navigation capability to the GPS module. One can input
two waypoints, wp[0] and wp[1], with altitude. If the altitudes differ, then
the altitude deviation from a "straigth" line from wp[0] to wp[1] is
calculated. The true course and course deviation from wp[0] to wp[1] is also
calculated. All this can be found in the wp subdir where one also finds the
wp[0] and wp[1] subdirs.

All this has to be done through the property browser. Maybe I should make a
gui window for the GPS!

20 years agoIt can't be turned off, so force the running flag to true to avoid an
andy [Sat, 1 May 2004 04:59:09 +0000 (04:59 +0000)]
It can't be turned off, so force the running flag to true to avoid an
uninitialization stupidity.

20 years agoInitial checkin of a TurbineEngine implementation. This hasn't been
andy [Sat, 1 May 2004 00:26:33 +0000 (00:26 +0000)]
Initial checkin of a TurbineEngine implementation.  This hasn't been
tested at all yet, but it doesn't seem to have broken anything so it
should be safe.  See the README in the base package for docs.

20 years agoAttempt at a fix for the propeller torque problems in the
andy [Sat, 1 May 2004 00:25:56 +0000 (00:25 +0000)]
Attempt at a fix for the propeller torque problems in the
"slow/windmilling propeller" regime.  I'm happy with the foundations
of the solution, but this hasn't been complete tested yet.  The
solution behavior seems fine on the planes I tried.

20 years agoRefactoring in preparation to add a turbine engine to YASim. The
andy [Fri, 30 Apr 2004 19:06:29 +0000 (19:06 +0000)]
Refactoring in preparation to add a turbine engine to YASim.  The
PistonEngine class has grown an "Engine" superclass.  Some other stuff
moved around too, and I cleaned up some property naming while I was in
there.  This hasn't been tested very thorougly, hopefully I didn't
break anything.

20 years agoMSVC fix.
ehofman [Fri, 30 Apr 2004 08:44:59 +0000 (08:44 +0000)]
MSVC fix.

20 years agoRemove the references to fg_os.hxx since the UL_USE_XXX variable is located in the...
ehofman [Fri, 30 Apr 2004 08:44:35 +0000 (08:44 +0000)]
Remove the references to fg_os.hxx since the UL_USE_XXX variable is located in the config.h file since the changes for confgure.ac.

20 years agoChanges to get FlightGear (well, the src directory at least) to
andy [Fri, 30 Apr 2004 00:52:11 +0000 (00:52 +0000)]
Changes to get FlightGear (well, the src directory at least) to
configure and compile out-of-the-box on a MinGW target:

Use -lSDL instead of -lglut32 on windows builds when --enable-sdl
is set.

Link against alut.dll in addition to openal32.dll.

Replace BSD bcopy() with ANSI C memmove() in a few places.  This is
simpler than trying to abstract it out as a platform dependency in a
header file; bcopy() has never been standard.

The ENABLE_THREADS handling has changed to be set to 0 when threads
are not in use.  This breaks expressions like #ifdef ENABLE_THREADS.
Replace with a slightly more complicated expression.  It might have
been better to fix the configure.ac script, but I didn't know how and
this whole setting is likely to go away soon anyway.

The MinGW C runtime actually does include snprintf, so only MSVC
builds (and not all WIN32 ones) need _snprintf in JSBSim/FGState.cpp

Building on a platform with no glut at all exposed some spots where
plib/pu.h was being included without a toolkit setting (it defaults to
glut).  Include fg_os.hxx first.

And when still using glut, glut.h has a bizarre dependency on a
_WCHAR_T_DEFINED symbol.  It it's not defined, it tries to redefine
(!!) wchar_t to disasterous effect.

20 years agoUpdated to document the new 3d positional tags that are available for
curt [Thu, 29 Apr 2004 19:32:03 +0000 (19:32 +0000)]
Updated to document the new 3d positional tags that are available for
sound configurations.

20 years agoAdd a description on how to setup surround sound when using ALSA.
ehofman [Thu, 29 Apr 2004 19:18:48 +0000 (19:18 +0000)]
Add a description on how to setup surround sound when using ALSA.