valkyrie-developers — Developer mailing list

Re: [Valkyrie-developers] VSIM update

[Jason R N. <jne...@co...>]

Well, the one reason I would still hesitate to go with VS solely is that not
everyone can run out and buy it. With CMake, other developers still have a
choice of a free IDE and free compilers to go with and can still build
everything. While I agree that setting up CMake wasn't my most favorite
task, it's not so bad now that I'm familiar with it, and it beats the hell
out of autoconf.

Anyone else with a voice on this? Is anyone still on this list, lol :)

    Jason

PS
I've uploaded the latest files to get everything to work using CMake to CVS,
here are the notes, check the documents directory for more notes, I try to
keep them up to date, lol.
2008-01-21: Jason Neuhaus (Max...@us...)*
automake/autoconf files removed (Makefile.am, Makefile.in).* CMake files
added CMakeLists.txt* Added a build directory to keep the source directories
clean from files used forbuilding the executive (see the developer
documentation in the documents directoryfor more information).*
graphicdata/*removed obsolete files* graphics/...removed obsolete files*
graphics/models/generic/GenericHud.cppreplaced rint rounding function with
static_cast<int>(#+0.5) since rint is notstandard.*
graphics/models/generic/VectorFont.*Fixed some warnings.Added a #include
"Windows.h" required when using the MSVC compiler. It is includedonly when
MSVC is defined.* models/generic/EngineSim.cppAdded cfloat includeChanged
finite() to _finite().* models/toolbox/devices/Clock.cppAdded include for
ul.h when using MSVC since PLIB does not consistently name theincludes
between VS solutions and MinGW/MSYS builds.*
models/toolbox/devices/Joystick.cppAdded include for js.h when using MSVC
since PLIB does not consistently name theincludes between VS solutions and
MinGW/MSYS builds.* models/toolbox/utilities/Constants.hreplaced definition
of PI to work across platforms.* main.cppRemoved obsolete headers.Added
usage of new #defines for the program version number.-----Original
Message-----
  From: val...@li...
[mailto:val...@li...]On Behalf Of Tod
Filer
  Sent: Friday, January 18, 2008 11:58 PM
  To: val...@li...
  Subject: Re: [Valkyrie-developers] VSIM update


  You're asking a person who develops applications in Microsoft development
systems for a living. Dump Linux and go pure Microsoft. Of course I have no
financial interest in Microsoft development platforms, not. It would be nice
to find a conversion program that will go from Visual Studio back to CMake.



  Tod




----------------------------------------------------------------------------
--

  From: val...@li...
[mailto:val...@li...] On Behalf Of
Jason R Neuhaus
  Sent: Sunday, January 13, 2008 6:21 PM
  To: Valkyrie-Developers
  Subject: [Valkyrie-developers] VSIM update



  Hey guys,

    Just a heads up, I spent a large part of my weekend working on the code
again, and trying to upgrade us to CMake instead of that Automake/autoconf
stuff which was always a huge headache. Well, I've gotten it to work just
fine under Eclipse and MinGW/MSYS, and I have gotten it to compile under
Visual Studio 2005, however, I still have a link problem in VS where it
can't find the libraries after it's been built *Sigh*. I have the .exe, but
when I try to run it, it can't find it's dlls. So hopefully this week I'll
figure out how to get this fixed and we'll have support for Visual Studio
added to hopefully make it easier for more people to help out.



    Now, after I say all that, does anyone have anything to say regarding
continuing support for Linux, etc.? I would like to, but even CMake can be a
bit of a hassle, although I must say, so far, it is much friendlier than
automake/autoconf ! The problem with switching to CMake of course is that
you can't edit the project in Visual Studio, i.e. if you want to add source
files, change compile lines, etc. you have to do it in the CMake files. You
can still build, debug, etc. though. Any input?



      Jason

Re: [Valkyrie-developers] VSIM update

[Tod F. <tf...@fi...>]

You're asking a person who develops applications in Microsoft development
systems for a living. Dump Linux and go pure Microsoft. Of course I have no
financial interest in Microsoft development platforms, not. It would be nice
to find a conversion program that will go from Visual Studio back to CMake.

 

Tod

 

  _____  

From: val...@li...
[mailto:val...@li...] On Behalf Of
Jason R Neuhaus
Sent: Sunday, January 13, 2008 6:21 PM
To: Valkyrie-Developers
Subject: [Valkyrie-developers] VSIM update

 

Hey guys,

  Just a heads up, I spent a large part of my weekend working on the code
again, and trying to upgrade us to CMake instead of that Automake/autoconf
stuff which was always a huge headache. Well, I've gotten it to work just
fine under Eclipse and MinGW/MSYS, and I have gotten it to compile under
Visual Studio 2005, however, I still have a link problem in VS where it
can't find the libraries after it's been built *Sigh*. I have the .exe, but
when I try to run it, it can't find it's dlls. So hopefully this week I'll
figure out how to get this fixed and we'll have support for Visual Studio
added to hopefully make it easier for more people to help out.

 

  Now, after I say all that, does anyone have anything to say regarding
continuing support for Linux, etc.? I would like to, but even CMake can be a
bit of a hassle, although I must say, so far, it is much friendlier than
automake/autoconf ! The problem with switching to CMake of course is that
you can't edit the project in Visual Studio, i.e. if you want to add source
files, change compile lines, etc. you have to do it in the CMake files. You
can still build, debug, etc. though. Any input?

 

    Jason

[Valkyrie-developers] VSIM update

[Jason R N. <jne...@co...>]

Hey guys,
  Just a heads up, I spent a large part of my weekend working on the code
again, and trying to upgrade us to CMake instead of that Automake/autoconf
stuff which was always a huge headache. Well, I've gotten it to work just
fine under Eclipse and MinGW/MSYS, and I have gotten it to compile under
Visual Studio 2005, however, I still have a link problem in VS where it
can't find the libraries after it's been built *Sigh*. I have the .exe, but
when I try to run it, it can't find it's dlls. So hopefully this week I'll
figure out how to get this fixed and we'll have support for Visual Studio
added to hopefully make it easier for more people to help out.

  Now, after I say all that, does anyone have anything to say regarding
continuing support for Linux, etc.? I would like to, but even CMake can be a
bit of a hassle, although I must say, so far, it is much friendlier than
automake/autoconf ! The problem with switching to CMake of course is that
you can't edit the project in Visual Studio, i.e. if you want to add source
files, change compile lines, etc. you have to do it in the CMake files. You
can still build, debug, etc. though. Any input?

    Jason

[Valkyrie-developers] CMake in progress

[Jason R N. <jne...@co...>]

Hey Guys,
  I'm in the process of switching us over to using cmake
(http://www.cmake.org) instead of automake/autoconf. It will likely take me
a good while to switch it over because I have no idea what I'm doing and I
want to maintain all the options that were in the automake rules. But, after
I am done, we will have Visual Studio projects, so it should make it pretty
easy for everyone to build. I will be testing it using mingw, msys, and
visual studio 2005 when I am done.
  THEN, I will move on to switching us from CVS to subversion, so if you
have changes you are working on, get them in before I switch, or let me know
so I know to hold off for a bit! :).
  Once all that is done, I'll try and update our documentation on how to
build/install everything needed again, and update us to the latest version
of Open Scene Graph, which has changed quite a bit since the last release.

  Jason

[Valkyrie-developers] Getting the ball rolling again

[Jason R N. <jne...@co...>]

Hey guys, it's been a long time since my last update, but I want to get back
into the swing of things again soon, so just a few quick notes:

1) Nope, I haven't given up (never will, lol)
2) I may look into switching from automake/autoconf to cmake
(www.cmake.org). If anyone has ANY experience with this, let me know, but it
sounds great. It supports Linux, etc. AND can produce project files for
Visual Studio..which leads too #3.
3) I may be getting a copy of Visual Studio 2005 soon, we'll see, Microsoft
had a promo running, and I'm hoping to get my free copy to use :). So it
would be a great time for me to check out cmake under mingw and visual
studio and get it all working well I hope. I'm sure Tod would be glad to see
us work with Visual Studio well.

I signed up for my free copy at Microsoft by attending two of their online
Visual Basic classes. It is only good though June 30th, so do it now if you
want it! All you have to do is do two of their training classes and leave
feedback for both. The link is here:
http://www.microsoft.com/business/vb2005upgrade/default.mspx

It's probably too good to be true, I guess I'll find out in 6-8 weeks.

    Jason

[Valkyrie-developers] Checkup

[Jason R N. <jne...@co...>]

Hey guys,
  Just a check to see who all is still out there. I haven't had much time to
work on the project lately, new computer, Ernesto, etc. But I'm checking to
see who all is still interested in actively working on the project so I can
cleanup the developer list on the website. So please email me back if you
are still interested in helping out and we'll try and get everything
reorganized :)

    Jason

[Valkyrie-developers] Vsim 0.1.2

[Jason R N. <jne...@co...>]

0.1.2 is up.

NEWS:
05/20/2006 - Jason
 Version 0.1.2 changes include:
 - Nz (G loading) is now displayed on the VF-1 HUD.
 - Mass Balance adjustments so that the VF-1 stays more neutrally balanced
as the wings
   sweep aft. This allows the VF-1 to operate at lower throttle settings.
 - A bug was fixed in the rate limit code in the filters. The surface
actuators for all
   aircraft will now be correctly limited to the specified maximum rate.

ChangeLog:
2006-05-20: Jason Neuhaus (Max...@us...) (vsim-0_1_2)
    * graphics/models/generic/VectorFont.cpp
    Added additional characters.
    * graphics/models/Macross/VF1/VF1Graphics
    Added Nz to pass to VF1 HUD.
    * graphics/models/Macross/VF1/VF1Hud
    Added display of Nz.
    * models/generic/sensors/Accelerometer.h
    Added additional const qualifies to accessors.
    * models/interfaces/Macross/VF1/VF1ModelInterface.cpp
    Passes Nz to graphics.
    * models/Macross/VF1/VF1.cpp
    Added recording of Nz at pilot station. Center of mass location was
balanced to
    more closely follow the aerodynamic center movement to attempt to
minimize the
    pitching moment required as the wings sweep.
    * models/Macross/VF1/VF1ControlSystem.cpp
    Increase the rate limit speed of the engine thrust vectoring to 160
deg/sec.
    Minimum throttle decreased to 1% from 10%.
    Speed brakes/spoilers no longer autodeploy at low throttle settings.
    Maximum control authority increased to 15 for longitudinal control, was
previously
    10.
    Converted back to the old z thrust command system due to difficulty
getting the
    direct TV drive working with the engine rate limits now working
properly.
    Engine thrust will now properly increase when a c* command is given as
the engines
    vector from zero. Pitch command has priority over thrust command.
    Fixed a integrator type wind up problem in the spoiler commands, they
should no longer
    wind up past their maximum positions.
    * models/toolbox/Filter.h
    Fixed a bug in the calculation of the maximum delta value when using
rate limits.
    The rate limits should now work properly.
--
No virus found in this outgoing message.
Checked by AVG Free Edition.
Version: 7.1.392 / Virus Database: 268.6.1/344 - Release Date: 5/19/2006

[Valkyrie-developers] VSim 0.1.1

[Jason R N. <jne...@co...>]

New set of files up. The VF-1 fighter mode is now MUCH more well behaved, at
least longitudinally. Lateral control still needs work, but it's a big
improvement. Other big changes
- Thrust vectoring angles now appear on the HUD
- ESC now exits the program correctly and does not leave the program
hanging, i.e. hard to quit.
- VF-1 now much more controlable.

    Jason

2006-04-30: Jason Neuhaus (Max...@us...) (vsim-0_1_1) *
.cdtproject New targets added for the VF-1 FF2001 engine test/table buildup.
* configure.in Version updated to 0.1.1 * main.cpp Added code to change for
SceneManager termination as an exit criteria. Previously, if ESC was
pressed, the SceneManager/OSG code would quit running but the main program
continued running. ESC will now properly exit the main program. *
graphics/SceneManager.h Added isDone() method that returns if the graphics
have received a command to exit (currently the ESC key). *
graphics/SceneManager.cpp Added code that does not use the default set of
OSG viewer settings. It has been commented out for now, however, until a
keyboard handler is added. * graphics/models/generic/GenericHud.h Moved the
VectorFont object to the protected section so it can be used to draw vector
fonts in derived graphic objects. * graphics/models/Macross/VF1/Makefile.am
Added the VF1Hud class * graphics/models/Macross/VF1/VF1Graphics.* Switched
to a VF-1 specific HUD instead of the generic HUD *
graphics/models/Macross/VF1/VF1Hud.* Class containing modification for the
VF-1 specific HUD. This class derives from Generic Hud and displays the left
and right thrust vectoring values in the upper right corner. *
models/F14A/F14AAero.h Added mutator for Cm due to Mach. putCmMach(). *
models/F14A/F14AAero.cpp Added recording of aerodynamic forces/moments Added
a temporary limit to induced drag. Updated the pitching moment due to wing
sweep. * models/generic/Test/.cvsignore Added test executables to ignore
list. * models/generic/Test/FF2001Test.cpp New test to create a table of
thrust and drag values for the FF2001 engine. *
models/generic/Test/Makefile.am Added FF2001_test as a extra program. *
models/interfaces/Macross/VF1/VF1ModelInterface.cpp Added left and right
thrust vector values. * models/Macross/VF1/FF2001.* New class. This class
performs a table lookup to determine the thrust output and ram drag of the
FF2001 engine. * models/Macross/VF1/Makefile.am Added the FF2001 class. *
models/Macross/VF1/VF1.* Switched to the FF2001 class for engine modeling.
Added accessors to the left and right thrust vector positions. Added
rotations for the left at right engine thrust vectoring. Added data
recording of engine forces/moments. Modified CM location as a function of
wing sweep. * models/Macross/VF1/VF1Aero.cpp Added a term to reduce the
pitching moment coefficient due to Mach number. This greatly reduces the
Mach tuck problem for the VF-1 since it does not have glove vanes. *
models/Macross/VF1/VF1ControlSystem.h Added mutator for Qbar. Modified
accessors to PLA to return a normalized throttle instead of in degrees. *
models/Macross/VF1/VF1ControlSystem.cpp Added limits to the C* error
integrator to attempt to prevent wind-up. Modified the longitudinal control
law to command C* through more speed ranges. Previously, the gains were only
tuned for lower speed. The VF-1 should have a much harder time flipping end
over end as in the previous version. The longitudinal law now commands the
vectoring angles directly instead of through the z thrust value. The
previous version had problems where the Z thrust command was not high enough
to obtain the value of C* commanded. A section was added to the longitudinal
law to allow the engines to automatically throttle up if the law calls for
more torque than is available at the current throttle setting. This should
help the aircraft to behave when the throttle is brought back to idle.
Lateral gains were modified to vary with Qbar and should no longer jitter at
high speed due to too high a gain. * models/toolbox/math/Integrator.* Added
limits to the integrator. The limits may be set through the putLimits()
method. * models/toolbox/utilities/Constants.h Fixed a comment for
temperature constants.
--
No virus found in this outgoing message.
Checked by AVG Free Edition.
Version: 7.1.385 / Virus Database: 268.5.1/327 - Release Date: 4/28/2006

[Valkyrie-developers] Still here

[Jason R N. <jne...@co...>]

Been a while since I've sent an email out. I just got back to working on the
game again a few weeks ago and I've been deep in the control law and
aerodynamics for the VF-1 again. I have no idea when I'll be done as it's a
non-trivial problem. I'm hoping to be able to get the aircraft to fly better
at all speeds soon. I'm also about to rework the engine model so that the
"afterburner" works in a more continuous fashion.

Anyone else got anything good coming?

    Jason
--
No virus found in this outgoing message.
Checked by AVG Free Edition.
Version: 7.1.375 / Virus Database: 268.2.1/279 - Release Date: 3/10/2006

[Valkyrie-developers] VSIM-0.1.0

[Jason R N. <jne...@co...>]

Okay, I have uploaded my changes to CVS and will upload source tarballs,
win32 executables to the project page later tonight probably, or tomorrow if
I don't get done today. I have given the current CVS version the branch name
vsim-0_1_0.

Feel free to try it out and tell me what you think when I get the files
uploaded. I will be uploading 0.1.0 source tarballs, and a zip file
containing the files needed to run under Windows (MinGW/MSYS build).

	Enjoy!
	Jason

Changelog:
2005-11-24: Jason Neuhaus (Max...@us...) (vsim-0_1_0)
	* main.cpp
	Increased run time to 5 minutes.
	Lowered number of trim iterations from 1500 to 100.
	* documents/developers_notes.html, release_notes.html,
	  vf1_overview.html
	Added documentation on the overall program, development notes, and
	an overview of the VF-1 aircraft.
	* graphics/models/F14A/F14AGraphics.cpp,
	  graphics/models/F16/F16Graphics.cpp
	Updated the HUD rotation so that the HUD is no longer reversed (left to
	right)
	* graphics/models/Macross/VF1/VF1Graphics.h,cpp
	Adding mach, EAS, and Altitude to the graphics (HUD)
	Updated the HUD rotation so that the HUD is no longer reversed (left to
	right)
	* graphics/models/generic/VectorFont.h, cpp
	New class for drawing a vector font in OpenGL (see the GenericHud class
	for example usage). Note that now all characters have yet been
	implemented.
	* graphics/models/generic/GenericHud.h, cpp
	Added Mach, Speed, and Altitude to the HUD graphics.
	Added heading to the display.
	* models/F14A/F14AAero.h, cpp
	Removed Cy beta min and max.
	Changed Alpha0 from -2.0 to -0.5.
	Approximated speedbrake effects better approximate the size of the
	speed brake. Drag was estimated using flat plate drag assuming the
	speed brake extends to 60 degrees.
	Corrected various beta terms to properly limit themselves when Cybeta
	is limited.
	* models/interfaces/Macross/VF1/VF1ModelInterface.cpp
	Added pass through for Mach, EAS (speed), and altitude for HUD.
	* models/Macross/VF1/VF1.h, cpp
	Added EAS/Mach accessors and calculations.
	Re-enabled afterburners.
	Added VS_DEBUG directives around couts.
	* models/Macross/VF1/VF1ControlSystem.h, cpp
	Increased thrust vectoring limits from 45 degrees to 60 degrees.
	This was done to give more control authority at low speeds.
	Speed brake was increased from 50 degrees to 60 degrees.
	Re-implemented the C* error integration limits. The error integration
	is now limited to 10.
	A minimum thrust limit was implemented to prevent the throttle from
	being retarded so low that the thrust vectoring can no longer work.
	Retarding the throttles past 0.20 will now cause the speed brake and
	spoilers to be deployed automatically.
	Roll gain was decreased from 1.0 to 0.5.
	* models/toolbox/dynamic/EarthAtmosphere.h
	Added an accessor to the density at sea-level.
	* models/toolbox/dynamic/Universe.cpp
	  models/toolbox/math/Integrator.cpp
	  models/toolbox/math/Table.h, cpp
	Added VS_DEBUG directives around all cout and cerr lines.
	Changed include math.h to include cmath.
	* models/toolbox/utilities/Constants.h
	Fixed an initialization error with the m_to_nmi conversion.
	* .cdtproject
	Added new Make targets for Eclipse
	* aclocal.m4
	Autoconf changes.
	* configure_debug
	SH script for configuring with DEBUG
	* configure_optimized
	SH script for configuring with Optimization turned on
	* configure.in
	updated version number to 0.1.0
	commented out default compiler options.
	* graphics/models/generic/Makefile.am
	Added new class VectorFont.
	* graphics/models/Makefile.am
	Corrected typo HEADEERS -> HEADERS
	* models/interfaces/F14A/Makefile.am
	  models/interfaces/F16/Makefile.am
	  models/interfaces/Macross/VF1/Makefile.am
	Removed extraneous / from end of paths. This was causing errors in
	automake.
--
No virus found in this outgoing message.
Checked by AVG Free Edition.
Version: 7.1.362 / Virus Database: 267.13.7/182 - Release Date: 11/24/2005

RE: [Valkyrie-developers] 0.1.0 Final tasks

[Jason R N. <jne...@co...>]

Since I haven't sent out an update recently, lol.

Final tasks for 0.1.0:
done 		- Add Heading value to HUD (Jason)
done(EAS instead of IAS) - Add Airspeed (IAS) to HUD - Mach is already
complete (Jason)
done 		- Add pitch indications to pitch ladder (Jason)
mostly done - Add a fix to prevent 0 throttle from losing control of the
VF-1. Either an
aft throttle limit, or balance the aircraft better (Jason)
done 		- Remove couts (Jason)
		- Extend runtime past 2 minutes (Jason)
		- Make a quick document describing how it works/what's going on (Jason)

So basically, I've got a few tweaks to do and documentation. A lot of files
have been modified for this change, I'll try and get it done ASAP and get us
a release! :)

	Jason

-----Original Message-----
From: val...@li...
[mailto:val...@li...]On Behalf Of
Jason R Neuhaus
Sent: Monday, October 10, 2005 4:39 PM
To: Valkyrie-Developers
Subject: [Valkyrie-developers] 0.1.0 Final tasks



I'm working on the final changes for the 0.1.0 release, if anyone else has
anything to add, let me know, otherwise I will complete my tasks, get
binaries built and put them on the site :). I've been working on the HUD for
the last few weeks and have the ability to display some text now. I plan to
have all of these tasks done within 2 weeks, most of them are relatively
simple, I just need time to do them.

Final tasks for 0.1.0:
- Add Heading value to HUD (Jason)
- Add Airspeed (IAS) to HUD - Mach is already complete (Jason)
- Add pitch indications to pitch ladder (Jason)
- Add a fix to prevent 0 throttle from losing control of the VF-1. Either an
aft throttle limit, or balance the aircraft better (Jason)
- Remove couts (Jason)
- Extend runtime past 2 minutes (Jason)
- Make a quick document describing how it works/what's going on (Jason)

	Jason
--
No virus found in this outgoing message.
Checked by AVG Anti-Virus.
Version: 7.0.344 / Virus Database: 267.11.13/126 - Release Date: 10/9/2005



-------------------------------------------------------
This SF.Net email is sponsored by:
Power Architecture Resource Center: Free content, downloads, discussions,
and more. http://solutions.newsforge.com/ibmarch.tmpl
_______________________________________________
Valkyrie-developers mailing list
Val...@li...
https://lists.sourceforge.net/lists/listinfo/valkyrie-developers
--
No virus found in this incoming message.
Checked by AVG Anti-Virus.
Version: 7.0.344 / Virus Database: 267.11.13/126 - Release Date: 10/9/2005

--
No virus found in this outgoing message.
Checked by AVG Free Edition.
Version: 7.1.362 / Virus Database: 267.13.3/173 - Release Date: 11/16/2005

[Valkyrie-developers] 0.1.0 Final tasks

[Jason R N. <jne...@co...>]

I'm working on the final changes for the 0.1.0 release, if anyone else has
anything to add, let me know, otherwise I will complete my tasks, get
binaries built and put them on the site :). I've been working on the HUD for
the last few weeks and have the ability to display some text now. I plan to
have all of these tasks done within 2 weeks, most of them are relatively
simple, I just need time to do them.

Final tasks for 0.1.0:
- Add Heading value to HUD (Jason)
- Add Airspeed (IAS) to HUD - Mach is already complete (Jason)
- Add pitch indications to pitch ladder (Jason)
- Add a fix to prevent 0 throttle from losing control of the VF-1. Either an
aft throttle limit, or balance the aircraft better (Jason)
- Remove couts (Jason)
- Extend runtime past 2 minutes (Jason)
- Make a quick document describing how it works/what's going on (Jason)

	Jason
--
No virus found in this outgoing message.
Checked by AVG Anti-Virus.
Version: 7.0.344 / Virus Database: 267.11.13/126 - Release Date: 10/9/2005

[Valkyrie-developers] CVS Update

[Jason R N. <jne...@co...>]

Several bugs/annoyances have been fixed:
1) HUD no longer jitters, this seems to have been cause by a difference in
the way the rotation matricies were computed, the camera used
MatrixTransform and the HUD used PositionAttitudeTransform. The HUD
transform has been changed to use MatrixTransform. This appears to have
fixed the problem.
2) If the joystick used does not have a throttle or rudder, the inputs into
the models will not be zeroed out as was done previously. This will be most
noticable in the VF-1 where setting the throttle to zero caused an
uncontrollable pitch.
3) The roll feedback proportional gain in the VF-1 control system has been
reduced to eliminate the jitter problem I was seeing. The gain still needs
tuning, but it behaves much better now (and rolls faster).
4) The sideslip on the HUD is now drawn in the correct direction.
5) The rudder on the VF-1 now works in the correct direction.

	Jason

2005-09-11: Jason Neuhaus (Max...@us...)
	* models/toolbox/devices/Joystick.h:
	  added accessor to the number of axes on the joystick.
	* models/interfaces/Macross/VF1/VF1ModelInterface.cpp,
	  models/interfaces/F14A/F14AModelInterface.cpp,
	  models/interfaces/F16/F16ModelInterface.cpp:
	  Adding checking for number of joystick axes to prevent
	  throttle from zeroing out if no throttle is present.
	* graphics/models/generic/GenericHud.cpp:
	  Corrected sign on sideslip term (beta). Sideslip should now be
	  shown in the correct direction.
	* models/Macross/VF1/VF1ControlSystem.cpp: rudder fix and roll jitter fix
	  Reimplemented rudder terms. Rudder works again.
	  Corrected sign on rudder command based on pedal input.
	  Reduced the roll proportional feedback gain (k_p) to prevent roll jitter
	  at the default initial condition.
	* graphics/models/Macross/VF1Graphics.cpp, .h,
	  graphics/models/F14A/F14AGraphics.cpp, .h,
	  graphics/models/F16/F16Graphics.cpp, .h:
	  Converted the main vehicle transformation from a
PositionAttitudeTransform
	  to a MatrixTransform. This eliminates the "jumping" HUD that sometimes
	  occurs. Apparently, this was caused because the camera matrix calculation
	  comes out with a slightly different transform than the PAT does, Matrix
	  Transform does not have this problem.
	* models/toolbox/utilities/Constants.h:
	  Changed the 'pi' variable to use M_PI instead of a specified value.
	* models/F14A/F14AAero.cpp: data recording
	  Added data recording for all the internal coefficients.
--
No virus found in this outgoing message.
Checked by AVG Anti-Virus.
Version: 7.0.344 / Virus Database: 267.10.21/96 - Release Date: 9/10/2005

[Valkyrie-developers] CVS update

[Jason R N. <jne...@co...>]

The HUD now has a pitch ladder! No numbers yet, but the ladder is there and
is properly aligned with the out the window scene so taht the 0 pitch line
stays on the horizon. The files have been posted to CVS.

I will work on the VF1 aerodynamics package next to try and address some
flight issues there. There is a fluttering that occurs at maximum speed near
the ground that needs to be looked into. Also, the VF-1 needs a minimum
throttle, cutting them to zero causes the aircraft to lose all pitch
authority and causes loss of control.

2005-09-05: Jason Neuhaus (Max...@us...)
	* graphics/models/generic/GenericHud.h:
	  Added new constructor arguments and variables to support specification of
	  the HUD location and size so that the pitch grid can be properly aligned
	  with the out the window scene.
	* graphics/models/generic/GenericHud.cpp:
	  Added a pitch grid. The HUD now has a pitch ladder, a correct flight
	  path indicator, and circles at +/- 90.0 degrees pitch.
	* graphics/models/F14A/F14AGraphics.cpp:
	  Configured HUD to specify HUD location and size
	* graphics/models/F16/F16Graphics.cpp:
	  Configured HUD to specify HUD location and size
	* graphics/models/Macross/BF1/VF1Graphics.cpp:
	  Configured HUD to specify HUD location and size
	* graphics/SceneManager.cpp:
	  Fixed a bug with the specification of the fields of view and clipping
	  planes. Horizontal and vertical fields of view were increased to 60
	  degrees.
	* configure, configure.in:
	  Updated configure.in file format to newer version of autoconf/automake.
	* main.cpp:
	  Corrected a missing call to the graphics interface to process outputs
	  during trim that was preventing the HUD from being visible during the
	  trim phase. The HUD is now visible at all times.
--
No virus found in this outgoing message.
Checked by AVG Anti-Virus.
Version: 7.0.344 / Virus Database: 267.10.18/89 - Release Date: 9/2/2005

[Valkyrie-developers] CVS update

[Jason R N. <jne...@co...>]

More changes, again, mostly minor. I have added a few files that will be of
use when running Eclipse, they define the standard make options. In Eclipse
when you check out a new project, select check-out from CVS, then use the
regular CVS settings "cvs.sourceforge.net" "/cvsroot/valkyrie", login id,
etc. extssh, and checkout the "vsim" directory, then you should be golden,
let me know if not, I'm still learning Eclipse.

Tod, there is one change that will change for windows systems, and while I'm
thinking about it, one other thing you need to do for VC++ builds:
1) I changed main.cpp to not calibrate the joystick on windows systems, it
is unnecessary (I think). So go ahead and define "PLATFORM_WINDOWS" in your
builds, that will bypass joystick calibration.
2) The automake makefiles that I use define a VERSION preprocessor value
that I use when starting up the program to display the version number. You
should define "VERSION" equal to whatever version we are currently at
(currently 0.1.0RC1). If you look in the "configure.in" file, the version
number is specified on line 3 "AM_INIT_AUTOMAKE(vsim, 0.1.0RC1)".

Up next, real fixes for the 0.1.0 release! (imagine that)

	Jason

2005-08-28: Jason Neuhaus (Max...@us...)
    * configure.in, configure:
      Added a define for windows systems (PLATFORM_WINDOWS)
    * ChangeLog:
      Updated format to match what is specified on the GNU site.
    * .cvsignore:
      Added *.dll
      Added record_list.dat (data recording input file)
    * main.cpp:
      Windows systems do not need to calibrate the joystick, so if
      PLATFORM_WINDOWS is defined, the joystick calibration routine will
      not be called.
    * .cdtproject:
      Added new file for Eclipse development, this file will setup the
standard build
      options under Eclipse.
    * .project:
      Another new file for Eclipse. Sets up some options for the vsim
project.
--
No virus found in this outgoing message.
Checked by AVG Anti-Virus.
Version: 7.0.344 / Virus Database: 267.10.16/83 - Release Date: 8/26/2005

[Valkyrie-developers] CVS update

[Jason R N. <jne...@co...>]

New updates on CVS, all minor, i.e. if you are already working on something,
don't feel the need to pull these changes in. A lot of them were to help me
work using Eclipse, which by the way I am finding totally useful right now,
once I learned how to use it. Most of the changes were the addition of
.cvsignore files all over the place so that CVS knows what files it can
safely ignore instead of questioning every little file. Other changes
included a few comment fixes. The only functional changes were made to
main.cpp, GenericHud.cpp, and a minor warning fix to GraphicsInterface.h.

	Jason


08/20/2005 jason
    * main.cpp:
      Added code to update the model graphics during trim.
    * configure.in, configure:
      Update revision number to 0.1.0RC1.
      Added linking in of opengl32 library to allow compilation under
Cygwin.
      Removed extraneous '\' from end of line as it was confusing the Cygwin
Automake
      tools and didn't need to be there.
    * graphics/models/GraphicsInterface.h:
      Made destructor virtual.
    * graphics/models/F16/F16Graphics.cpp:
      Updated comments to say "F-16" instead of "VF-1"
    * graphics/models/F14A/F14AGraphics.cpp:
      Updated comments to say "F-14A" instead of "VF-1"
    * graphics/models/generic/GenericHud.cpp:
      Initialized some previously uninitialized values.
    * .cvsignore (various locations:
      Added new .cvsignores all over to attempt to clean up CVS trees. This
makes
      working with Eclipse much easier.
--
No virus found in this outgoing message.
Checked by AVG Anti-Virus.
Version: 7.0.338 / Virus Database: 267.10.13/78 - Release Date: 8/19/2005

[Valkyrie-developers] vsim-0.1.0 RC1 on CVS

[Jason R N. <jne...@co...>]

Release candidate one is up on CVS. No new files will be coming in unless
something is broken, so let's try to keep everything to bug fixes until we
get this out. I have no plans to add any new classes in the near future. If
you have problems, let me know, I'm really good at fixing all compiler
errors/warnings.

	Jason

Change Log additions:
08/01/2005 jason (vsim-0_1_0_RC1)
	* main.cpp:
	  Reimplemented code to allow creation of the F14A or F16 models again,
	  now that base classes are available to generalize the interface and
	  the F14A and F16 have classes to provide inputs to the models.
	* models/F14A/F14AAero.cpp:
	  Fixed an Unitialized Memory Read that may have been causing erratic
	  behavior on startup at times (span variable).
	* models/interfaces/F14AModelInterface.cpp, .h:
	  New class, model interface to F14A model.
	* models/interfaces/F16ModelInterface.cpp, .h:
	  New class, model interface to F16 model.
	* graphics/models/F14A/F14AGraphics.cpp, .h:
	  New class, graphics interface for the F14A model.
	* graphics/models/F16/F16Graphics.cpp, .h:
	  New class, graphics interface for the F16 model.
	* graphics/models/VF1/VF1Graphics.cpp, .h:
	  Modifed class to derive from GraphicsInterface.
	* graphics/models/GraphicsInterface.cpp, .h:
	  New base class for interfaces to graphics classes.
	* various Makefile.am, Makefile.in:
	  Added new graphics interfaces for F16 and F14A models.
--
No virus found in this outgoing message.
Checked by AVG Anti-Virus.
Version: 7.0.338 / Virus Database: 267.9.7/60 - Release Date: 7/28/2005

[Valkyrie-developers] New files coming, when CVS gets back

[Jason R N. <jne...@co...>]

Okay, I got motivated enough today to repeat what I did for the VF1 for the
F16 and F-14A models, so soon those models will be available to run again
(mostly me, it will still default to VF-1). The CVS server is not talking to
me and hasn't been for a few hours now, I'll send out a more official e-mail
when it goes. I also found a memory problem in my F-14 aerodynamics package
which affects both the F-14A and VF-1 models and may occasionally cause the
model to not trim properly (Tod, I'm hoping this will fix the problem you
were seeing with the vehicle going nuts from the get-go).

If anyone doesn't know "Valgrind" is a wonderful free memory analysis tool.
I was suprised it worked relatively well (I use Purify at work, nicer but
$$$$$).

This will be the last set of "NEW" files before a 0.1.0 release. I would
like to get these in, mark it as 0.1.0_RC1 (Release Candidate 1) and we'll
get Tod busy on the Windows version and maybe harass Katron to try a Mac
version for us.  Jase, are you still available to make sure we're still good
on FreeBSD?

I'm hoping to get all this in the 0.1.0 release:
- Windows support! (VS, I will try to make sure MinGW and/or Cygwin work)
- Mac support nice, but optional
- FreeBSD should still work
- Linux of course
- A small amount of HUD work to make it a little more usable
- NO new classes once RC1 is out to make everyones life easier
- Small mods will be okay and I will attempt to keep them very localized.
The only ones I forsee are the new GerericHud class and maybe F14AAero or
VF1Aero. I think the rest is good enough for 0.1.0

So another e-mail will come when I have added the last of the new files.
I've got them compiled and running, just can't check them in. grrr.

	Cheers,
	Jason
--
No virus found in this outgoing message.
Checked by AVG Anti-Virus.
Version: 7.0.338 / Virus Database: 267.9.7/60 - Release Date: 7/28/2005

RE: [Valkyrie-developers] Update

[Jason R N. <jne...@co...>]

Ohhh, he lives! hehe. *watches quietly as the busy beaver pops his head out
of the water for a breath* hehe.

But seriously, yeah, I agree. I'm dodging that area at the moment, but only
just barely.

	Jason

-----Original Message-----
From: val...@li...
[mailto:val...@li...]On Behalf Of
Vernard Martin
Sent: Friday, July 29, 2005 10:05 AM
To: val...@li...
Subject: Re: [Valkyrie-developers] Update


Jason R Neuhaus wrote:
> Now, the most obvious question is why did I split models/interfaces and
> graphics/models. The reason for the split is so that network data can
later
> be used to drive graphics/models without the need for the dynamic model
> itself.  So "player 2" can drive network data on his position,
orientation,
> etc. to "player 1" and the graphics/models will be the only clases
required
> to drive the display.

That is a good design decision. And we should probably start working on
the networking part of this to ensure that we arent over- or
under-designing this thing.

I'm done teaching after the 1st week of August. I'll have much more time
available after that since I'll only have one full time job instead of
two. I'll be able to try thing out myself.




-------------------------------------------------------
SF.Net email is Sponsored by the Better Software Conference & EXPO September
19-22, 2005 * San Francisco, CA * Development Lifecycle Practices
Agile & Plan-Driven Development * Managing Projects & Teams * Testing & QA
Security * Process Improvement & Measurement * http://www.sqe.com/bsce5sf
_______________________________________________
Valkyrie-developers mailing list
Val...@li...
https://lists.sourceforge.net/lists/listinfo/valkyrie-developers
--
No virus found in this incoming message.
Checked by AVG Anti-Virus.
Version: 7.0.338 / Virus Database: 267.9.7/60 - Release Date: 7/28/2005

--
No virus found in this outgoing message.
Checked by AVG Anti-Virus.
Version: 7.0.338 / Virus Database: 267.9.7/60 - Release Date: 7/28/2005

Re: [Valkyrie-developers] Update

[Vernard M. <ve...@cc...>]

Jason R Neuhaus wrote:
> Now, the most obvious question is why did I split models/interfaces and
> graphics/models. The reason for the split is so that network data can later
> be used to drive graphics/models without the need for the dynamic model
> itself.  So "player 2" can drive network data on his position, orientation,
> etc. to "player 1" and the graphics/models will be the only clases required
> to drive the display.

That is a good design decision. And we should probably start working on 
the networking part of this to ensure that we arent over- or 
under-designing this thing.

I'm done teaching after the 1st week of August. I'll have much more time 
available after that since I'll only have one full time job instead of 
two. I'll be able to try thing out myself.

[Valkyrie-developers] Update

[Jason R N. <jne...@co...>]

Okay, I've uploaded the files that outline the creation of the VF1 HUD. A
lot of things are missing or not working yet, but the class layout is there.
The changes relating to the HUD shouldn't result in any new files, just
updates to the files to get more and more of the HUD working.

I've also uploaded a screenshot to the website, you can see it here:
https://sourceforge.net/project/screenshots.php?group_id=3429

The VF-1 model still needs work, and only part of the HUD is driven
correctly. It's not much, but it's a start.

I will try to begin documenting the framework as it is becoming more and
more complicated and I'm sure hard for you guys to figure out why in the
heck I did things a certain way. I tried to put a lot of thought into each
decision to gives us all the flexibility to do everything with this game I
want to see later. That doesn't mean I didn't miss something, so feel free
to ask if something looks strange and I'll try and either explain, or go,
whoops, I didn't think of that.

Here are the basics.
Most everything in the models directory relates to driving the dynamics
models (VF-1) based on joystick, etc. input.
The SceneManager in graphics is just there to setup the (limited) graphics.
The models/interfaces files are there to provide an interface between a
dynamic models, it's inputs (joystick, etc.) and it's outputs (graphics)
The graphics/models files are intended to provide an interface to the
specific graphics that are drawn for each dynamic model.

Now, the most obvious question is why did I split models/interfaces and
graphics/models. The reason for the split is so that network data can later
be used to drive graphics/models without the need for the dynamic model
itself.  So "player 2" can drive network data on his position, orientation,
etc. to "player 1" and the graphics/models will be the only clases required
to drive the display.

	Enjoy,
	Jason

Here are the changes:
07/24/2005 jason
	* various Makefile.am and Makefile.in, configure, configure.in:
	  Update a few library names libdyntool->libdynamic, libtoolboxmath->
	  libmath.
	  Fixed some order dependencies and added new libraries relating to
	  the generic HUD being worked on.
	* main.cpp:
	  Major code cleanup. Removed most of the necessity to have pointers
	  to each aircraft type in the file. Joystick operations have also
	  been moved out. There are now base classes that the main executive
	  can operate on so that the tie in code for each specific vehicle
	  does not need to be in the main loop.
	* graphics/SceneManager.cpp:
	  Added code to set the lens perspective and remove the near clipping
	  plane so that the HUD displays when it is close to the camera.
	* graphics/SceneManager.h:
	  Added a method for graphics objects to add their node to the
	  scene.
	* graphics/models/Macross/VF1/VF1Graphics.cpp, .h:
	  New class. This class provides an interface to the graphics used
	  to display the VF-1. Currently it just passes along HUD data and
	  the position/orientation of the vehicle.
	* graphics/models/generic/GenericHud.cpp, .h:
	  New class. This class is the build up of a generic HUD capability
	  for aircraft. It is implemented using OpenGL calls.
	* graphics/models/generic/Hud.cpp, .h:
	  New class. Baseline class for HUD objects. Turns off lighting and
	  provides a means by which to setup display lists since OSG makes
	  this a pain.
	* models/interfaces/ModelInterface.cpp, .h:
	  New class. This class is the base class for the interface between
	  a model and it's input devices (joystick) and outputs (graphics).
	* models/interfaces/Macross/VF1/VF1ModelInterface.cpp, .h:
	  New class. This class provides the interface between the VF1 model
	  and the joystick and HUD.
--
No virus found in this outgoing message.
Checked by AVG Anti-Virus.
Version: 7.0.338 / Virus Database: 267.9.4/57 - Release Date: 7/22/2005

[Valkyrie-developers] almost done

[Jason R N. <jne...@co...>]

I've finally got some code written this weekend to clean up a lot of the
mess in the main program. I've got the VF-1 modified so that a new
VF1ModelInterface class pulls data from the joystick and puts it into the
VF1 aircraft model itself. After that it takes the model outputs and passes
them along to the graphics so that the graphics can be updated based on
various model states. I've still got a few issues to work out, and then I
need to copy the new VF1 classes and make similar versions for the F-16 and
F-14A models. I was going to put out a screenshot, but I shut down the
machine before I rememebered to take one, lol.

Also, the HUD is currently placed as an actual object in the 3D scene, i.e.
not just pasted on top of the scene like most HUDs. And I finally got it to
stay in front of the airplane when it flies! (As opposed to moving around in
all sorts of weird directions, anyone who has ever has to switch between
right and left handed coordinate systems is all too familiar with this I'm
sure).

So thinks are looking pretty good, and I have a fair idea of how the
graphics models will tie in now as well as have some ideas on how the
network data will work.

I'm also thinking of switching to Eclipse (www.eclipse.org) for an IDE.
(Visual studio support would still be great, but for those of us without it,
Eclipse if free). Does anyone have any experience with it? Will it make my
life easier from a Makefile perspective? Automake is far nicer than coding
Makefiles by hand, but with my efforts this weekend, I spent about half my
time setting up paths in the makefiles so that the compiler could find the
source files for various little parts. Then again, I probably need to
consolidate libraries some, that would more than likely help.

I'm hoping to get the HUD working and then move on to seeing how Eclipse
works and possibly switching to it.

	Jason
--
No virus found in this outgoing message.
Checked by AVG Anti-Virus.
Version: 7.0.323 / Virus Database: 267.8.11/45 - Release Date: 7/9/2005

[Valkyrie-developers] 3DS format support

[Jason R N. <jne...@co...>]

I think there was some question about whether or not OSG supports 3D Studio
or not. I believe it does and that their website says it supports lib3ds. My
guess is they forgot to pull the lib off the front of the extension since
that is likely the name of the library, so I believe it will support the 3D
Studio file format.

Also, I will be on vacation for about a week, so I won't be around to answer
e-mails. I'm taking my design papers with me and I'll be harassing Vernard
on this trip, ;) so hopefully I'll have it all worked out on paper before I
get back.

Have a good 4th everyone!
	Jason
--
No virus found in this outgoing message.
Checked by AVG Anti-Virus.
Version: 7.0.323 / Virus Database: 267.8.6/33 - Release Date: 6/28/2005

[Valkyrie-developers] Welcome Katron

[Jason R N. <jne...@co...>]

Hey guys,
  Just a heads up, I've snagged another anime fan to help with ideas and
such. Katron works just down the hall from me and has some ideas for AI and
general ideas about how things might work.

In other news, I spent the better part of the day trying ot figure out how
to tie all the code together, i.e. class diagrams and such. I've struggled
to come up with a design which will handle multiplayer, and in particular
the case where 1 vehicle can have 2 people in control, or even just 1 along
for the ride. I've got an idea sketched out, but right now I'm not 100%
confident it's the best solution. I don't have any diagrams on the computer
yet, so nothing to look at. But the basic problem is that, e.g. with a VT-1,
someone has to be flying it, but what happens if the pilot goes link dead.
The pilots client machine can't be the only computer that knows the exact
state of the aircraft. What I'm thinking is a minimal amount of data (I'm
hoping just the graphics data will be enough to do this since they are
similar parameters) will be sent continuously to the server, if the pilot
goes LD, the server either hands off the state to the other person in the
vehicle, or internally continues calculations roughly where they left off.

Now, I'm sure you are thinking that sounds easy, sure, just compute
everything from the last know position and orientation. The problem is there
are a lot of states maintained in the aircraft, and in order to be able to
continuing like nothing had ever happened, everything would have to be
copied to the server. For example, in order for the integration to even
proceed "exactly" like nothing had happened, all of the past integration
step values would have to be transfered, etc. Which could quickly turn into
a lot of data. I'm hoping we can just say you were "roughly" here with
"roughly" these states, client #2 do your best.

I believe I have a design that is flexible enough to do this. Really, this
came up when I was thinking about an F-14 with a pilot and radar signal
officer (I think that's right, lol). The RSO is in the vehicle with the
pilot and would probably be able to see what he was doing (i.e. could
probably see the HUD), well, if he's on another computer, he won't see squat
unless we transmit enough data to his computer to reproduce the HUD display.
It's obviously a stretch for now, but later if we get into vehicles that
have more and more people, there will have to be a way to make sure everyone
is seeing roughly the same thing.

I'm sure I'll be bugging Vernard a bunch about this in the near future to
try and make sure I'm not going to be overdoing the network bandwidth.
Either that or we can all switch to fiber! Whoohoo...who's buying my
connection ;).

So if my current design survives a few more days of me thinking about it,
I'll code up enough to get the HUD working and upload it. The HUD has just
brought in a LOT of complexities I hadn't even thought of that I'd like to
solve properly now rather than hack now and have a big headache later to
fix.

	Jason
--
No virus found in this outgoing message.
Checked by AVG Anti-Virus.
Version: 7.0.323 / Virus Database: 267.6.9 - Release Date: 6/11/2005

[Valkyrie-developers] Update notes

[Jason R N. <jne...@co...>]

Anyone else who is seeing odd behavior ever since we switched to Open Scene
Graph, go download their latest package (0.9.9, released May 25th), it says
it fixes numerous bugs, and some of the problems I was seeing have
disappeared. It also comes with new versions of Open Threads and Producer,
so all will need to be installed again.

Tod, this may also fix the problem you were seeing when running under Visual
Studio, I occasionally saw similar problems where it would not trim for no
apparent reason.

I may have FINALLY gotten OpenGL display lists semi-working under OSG...they
don't make it easy in the LEAST, so the HUD is finally underway again! I was
hoping to be able to put up a screenshot this weekend, but this display list
problem took me much longer than expected to work around. I'll try to get a
shot of it up next weekend. I'm hoping to have a working pitch ladder and
flight path indicator.

	Jason
--
No virus found in this outgoing message.
Checked by AVG Anti-Virus.
Version: 7.0.322 / Virus Database: 267.3.0 - Release Date: 5/30/2005