simuv2 wheel minor sound bugfix

simuv2/wheel.cpp
diff with r1.21
196c196
<               if (v < 2.0f) {
---
>               if (v < 2.0f && fabs(wrl) < 2.0f) {

Changed it so that skid is only turned off when both speed is small 
and wheels don't spin.

--

-- 
Christos Dimitrakakis
IDIAP (http://www.idiap.ch/~dimitrak/main.html)

-------------------------------------------------------
SF.Net email is sponsored by: Discover Easy Linux Migration Strategies
from IBM. Find simple to follow Roadmaps, straightforward articles,
informative Webcasts and more! Get everything you need to get up to
speed, fast. http://ads.osdn.com/?ad_id=7477&alloc_id=16492&op=click

compilation error in 1.2.4-test

Hello,

trying to compile Torcs 1.2.4-test, I got that :

g++ -I/home/chris/torcs-1.2.4/export/include -I/home/chris/torcs-1.2.4 -g -O2 
-Wall -fPIC -O2 -DUSE_RANDR_EXT -Wall -fPIC -O2 -DUSE_RANDR_EXT  
-D_SVID_SOURCE -D_BSD_SOURCE -DSHM -DHAVE_CONFIG_H  -c grmultitexstate.cpp
grmultitexstate.cpp: In member function `virtual void 
grMultiTexState::apply(int)':
grmultitexstate.cpp:30: error: `glActiveTexture' undeclared (first use this 
function)
grmultitexstate.cpp:30: error: (Each undeclared identifier is reported only 
once for each function it appears in.)
make[5]: *** [grmultitexstate.o] Erreur 1
make[5]: Leaving directory 
`/home/chris/torcs-1.2.4/src/modules/graphic/ssggraph'
make[4]: *** [subdirs] Erreur 1
make[4]: Leaving directory `/home/chris/torcs-1.2.4/src/modules/graphic'
make[3]: *** [subdirs] Erreur 1
make[3]: Leaving directory `/home/chris/torcs-1.2.4/src/modules'
make[2]: *** [subdirs] Erreur 1
make[2]: Leaving directory `/home/chris/torcs-1.2.4/src'
make[1]: *** [subdirs] Erreur 1
make[1]: Leaving directory `/home/chris/torcs-1.2.4'
make: *** [restart] Erreur 2

Regards,
--

-- 
Christophe BAEGERT

Re: compilation error in 1.2.4-test

Hi Christophe

Look at the TORCS site ("Update"). Does it help? If not:
- What OpenGL header do you use?
- config.log.
- which version exactly (package, CVS)?

Bye, Bernhard.

> Hello,
> 
> trying to compile Torcs 1.2.4-test, I got that :

Which one, test1 package or CVS?

--

-- 

Visit my homepage http://www.berniw.org
Official TORCS racing: The TORCS Racing Board, http://www.berniw.org/trb

-------------------------------------------------------
SF.Net email is sponsored by: Discover Easy Linux Migration Strategies
from IBM. Find simple to follow Roadmaps, straightforward articles,
informative Webcasts and more! Get everything you need to get up to
speed, fast. http://ads.osdn.com/?ad_id=7477&alloc_id=16492&op=click

dead zone

Re: [Torcs-devel] dead zone

> For TDC <http://speedy.chonchon.free.fr/indextdc.html>, I've tried to drive 
> with my wheel (Microso$t $idewinder FFB). I would like to remove the dead 
> zone.
> I try several value in */home/eliam/.torcs/drivers/human/preferences.xml* (0, 
> 1, -5) and I have the same result in the game. :-\

I have had the same problem with my logitech wheel. It is not a 
problem with torcs, but rather with the wheel, which has a built-in 
dead zone. It is possible that this is configurable by sending 
appropriate messages through the USB port.

--

-- 
Christos Dimitrakakis
IDIAP (http://www.idiap.ch/~dimitrak/main.html)

-------------------------------------------------------
SF.Net email is sponsored by: Discover Easy Linux Migration Strategies
from IBM. Find simple to follow Roadmaps, straightforward articles,
informative Webcasts and more! Get everything you need to get up to
speed, fast. http://ads.osdn.com/?ad_id=7477&alloc_id=16492&op=click

Re: [Torcs-users] Re: dead zone

Christos Dimitrakakis wrote:
>> For TDC <http://speedy.chonchon.free.fr/indextdc.html>, I've tried to 
>> drive with my wheel (Microso$t $idewinder FFB). I would like to remove 
>> the dead zone.
>> I try several value in 
>> */home/eliam/.torcs/drivers/human/preferences.xml* (0, 1, -5) and I 
>> have the same result in the game. :-\
> 
> 
> I have had the same problem with my logitech wheel. It is not a problem 
> with torcs, but rather with the wheel, which has a built-in dead zone. 
> It is possible that this is configurable by sending appropriate messages 
> through the USB port.
> 

Setting the "Steer dead zone" parameter in the control
configuration of the driver to "0" should do that for you...

Bye,
Eric.

Re: [Torcs-users] Re: dead zone

Hello Eric!

On Wed, 3 Aug 2005, Eric Espie wrote:

> Christos Dimitrakakis wrote:
>>> For TDC <http://speedy.chonchon.free.fr/indextdc.html>, I've tried to 
>>> drive with my wheel (Microso$t $idewinder FFB). I would like to remove the 
>>> dead zone.
>>> I try several value in */home/eliam/.torcs/drivers/human/preferences.xml* 
>>> (0, 1, -5) and I have the same result in the game. :-\
>> 
>> 
>> I have had the same problem with my logitech wheel. It is not a problem 
>> with torcs, but rather with the wheel, which has a built-in dead zone. It 
>> is possible that this is configurable by sending appropriate messages 
>> through the USB port.
>> 
>
> Setting the "Steer dead zone" parameter in the control
> configuration of the driver to "0" should do that for you...
>

That would be neat. So I thought I'd check the code. It appears that 
the only case in which deadzone is taken into account is when the 
mouse is used.

Correct me if I am wrong, but the only thing I see in drivers/human 
is instances of:
     case GFCTRL_TYPE_MOUSE_AXIS:
 	ax0 = mouseInfo->ax[cmd[CMD_THROTTLE].val] - cmd[CMD_THROTTLE].deadZone;

in driver.cpp, which I guess applies a dead zone in software to mouse 
inputs, while the following in pref.cpp:

 	if (cmdCtrl[cmd].deadZoneName) {
 	    cmdCtrl[cmd].deadZone = (float)GfParmGetNum(PrefHdle, defaultSettings,
cmdCtrl[cmd].deadZoneName, (char*)NULL, (tdble)cmdCtrl[cmd].deadZone);
 	    cmdCtrl[cmd].deadZone = (float)GfParmGetNum(PrefHdle, sstring, cmdCtrl[cmd].deadZoneName,
(char*)NULL, (tdble)cmdCtrl[cmd].deadZone);
 	}

seems to set the deadZone variable from the prefs file.

Conclusion: The wheel uses a dead zone of 0, at least on the TORCS 
side, unless there is some code elsewhere which passes this info to 
the OS and/or controller.

Suggested solution: I guess you should configure your wheel using some 
kind of system-wide prefs utility, if possible. I think the official 
software should do for windows machines, and for linux there is 
jscalibrator.

Further references: google for 'plib joystick dead zone'

--

-- 
Christos Dimitrakakis
IDIAP (http://www.idiap.ch/~dimitrak/main.html)

-------------------------------------------------------
SF.Net email is sponsored by: Discover Easy Linux Migration Strategies
from IBM. Find simple to follow Roadmaps, straightforward articles,
informative Webcasts and more! Get everything you need to get up to
speed, fast. http://ads.osdn.com/?ad_id=7477&alloc_id=16492&op=click

Re: compilation error in 1.2.4-test

Hi Bernhard,

Le Mercredi 3 Août 2005 11:07, Bernhard Wymann a écrit :
> Hi Christophe
>
> Look at the TORCS site ("Update").

sorry for my stupid bug report... That did it !

Regards,
--

-- 
Christophe BAEGERT

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

Re: compilation error in 1.2.4-test

Hi Christophe

>>Look at the TORCS site ("Update").
> 
> sorry for my stupid bug report... That did it !

Was not stupid, it is a problem... The configure script of test2 will 
set that variable by default, there will be as well an option to switch 
it off it.

Bye, Bernhard.

--

-- 

Visit my homepage http://www.berniw.org
Official TORCS racing: The TORCS Racing Board, http://www.berniw.org/trb

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

1.2.4-test2 status report (almost done)

Hi all

Test2 makes good progress, I guess I will start commiting my changes
tomorrow, so expect test2 within the next 7 days. The major change is
that we will have now a working debug build for Windows, I think this
will help a lot.
The only thing left to do is a final verification hop to Windows and
the Windows installer, and finally some shakedown tests.

Be aware that the CVS is still frozen otherwise, I will overwrite any
changes to avoid picking up new trouble Test2 will be a candidate
for the final release, we will see.

Bye, Bernhard.

Changes since last report
-------------------------
- Ported learning to Windows (Bernhard).
- Ported Olethros to Windows (Bernhard).
- Windows version result generation now correct if either started with full
   path or with wtorcs[.exe], current working directory must be the directory
   containing wtorcs.exe, there is no reason to improve this more (people
   start via desktop icon/menu entry anyway) (Bernhard).
- Added GfCreateDir in tgf (Christos, Bernhard).
- Fixed olethros clkdtm texture (Christos).
- Changed some code according to mail of Steven Gutknecht, did not look at the
   sqrt problem (if you want to have TORCS compile for a certain non free
   compiler send me patches or the money to buy it) (Steven Gutknecht,
   Bernhard).
- Kicked out SmartAssert.h for 2 reasons: 1. I want to get rid of this "clever"
   preprocessor tricks (because I think it is really silly to do such things),
   2. Assertions simply do never belong into any code which a third person can
   look at (if the user runs the software to code must deal with its problems by
   itself, the messages will be meaningless for the user anyway) (Bernhard).
- Removed assertions from Olethros (Bernhard).
- Fixed various memory leaks in Olethros caused by sloppy use of make_message
   (a comment regarding that: if possible do not allocate heap memory inside
   a called function, it is better to allocate it before calling and pass the
   buffer and its size as parameter, because the allocation is then visible on
   the first glimpse) (Bernhard).
- Fixed crash bug in Olethros (the proposed singleton does not work, reverted
   to working version) (Bernhard).
- Fixed a possible crash bug in gui.cpp, added deactivation of
   glutKeyboardUpFunc and glutSpecialUpFunc in GfuiScreenDeactivate (Bernhard).
- removed make_message and replaced it (Bernhard).
- Created a debug project for Windows, needed to remove make_message, enhanced
   custom realloc for debug versions of cutom malloc/free, added a workaround
   for clean allocator/deallocator pair in img.cpp, etc, huh... that was a lot
   of work (Bernhard).
- Fixed one more olethros bug, linker (in Linux) was confused about which
   CheckMatchingToken to use, added static to those in ANN.cpp -> file scope
   (Bernhard).
- Tested, build and debug build work with: gcc 3.3.1 (0 warnings), gcc 3.4.3
   (0 warnings), gcc 4.0.1 (a lot of warnings, but just in trackgen, otherwise
   0 warnings), vc++ 6.0sp6 (release build 0 warnings, debug 131 warnings which
   one can safely ignore) (Bernhard).

Changes since 1.2.3-test1
-------------------------
- Tried to fix problem with GL headers differences in Linux (Bernhard).
- Car sound adjustments and new samples (Christos).
- Moved learning and math headers in subdirectories to stop the upcoming
   chaos, updated dependent files (Bernhard).
- Updated Windows project preparation bat files (Bernhard).
- Ported the project to Windows, except olethros and learning (Bernhard).
- Fixed resident texture problem in Windows (Bernhard).
- Updated gl headers for Windows, oss.sgi.com (Bernhard).
- Resolved some OpenAL compatibility issues (Bernhard).
- Fixed XSLT problem in Windows, was a combination of wrong dtd filename,
   missing GetLocalDir and GetDataDir setup (Bernhard).
- Fixed weird bug in screen.cpp, execlp had wrong arguments, therefore
   the XML results headers become formatted wrong when you changed the
   display properties... yikes. Currently it does now work when the full path
   is given in argv[0], so a bit more work is needed (Bernhard).
- Make gcc 4.0.1 compile the code (warnings in trackgen are scheduled for
   1.2.5) (Bernhard).
- Removed unused code in various places (Bernhard).
- Fixed virtual constructor in olethros (Bernhard).
- Removed volume setting from sound menu (Bernhard).

--

-- 
Christos Dimitrakakis
IDIAP (http://www.idiap.ch/~dimitrak/main.html)

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