torcs-devel Mailing List for TORCS - The Open Racing Car Simulator

Hi,
   thanks for the help with getting current torcs running,

I'm driving the human with mouse and find it a bit very sensitive
to control for me of the car is running at high speed.

So my idea was to make the mouse sensivity a function of the speed
of the car and reduce sensivity when the speed is high to compensate
for the stronger effect of turning in the steer and thus make it
easy to control the car at high speeds with the mouse. Thus you
can use the resulution of the mouse move to to finer grained
steer control at high speed, which is good to avoid not intented
turns or so and made me able to reach a much better lap time
on the default track(which is great!) with mouse.

I'm using the trackpad from the notebook as mouse, but I think
what I found with the trackpad should apply to a normal mouse
and trackballs(all mouse pointer-based controlled steering) as well.

This is the patch I tried and helped me much to control the
steer fine enough at high speeds:

--- src/drivers/human/human.cpp	30 Dec 2002 11:30:20 -0000	1.15
+++ src/drivers/human/human.cpp	22 Feb 2003 21:24:15 -0000
@@ -393,9 +393,13 @@

     if (mouseMove < 0) {
 	mouseAxis[1] = -mouseMove;
+	mouseAxis[1] -= car->_speed_x*2;
 	mouseAxis[0] = 0;
     } else {
 	mouseAxis[0] = mouseMove;
+	mouseAxis[0] -= car->_speed_x*2;
 	mouseAxis[1] = 0;
     }
     mouseMove = (float)(refMouse.Y - mouse->Y);

Maybe the change should be applied to another place instead and
make it only one line, but I just want the people know of this
idea and give an example how it looks like that is could work
(or works by luck most of the time).

I think Eric should best if it is a useable generally, I was
not reading the driver howto in detail so far and did not much
with torcs, this is just my first patch which seemed to work
for me for the track and the car I've tested it with(the default
track and car)

Best regards and many thanks for torcs, it's great fun!
Bernd

On Sun, 9 Feb 2003, Eric Espie wrote:

> Hi,
>
> As CVS do not remember versions on directories, it is
> necessary to use the -P option with the cvs co or
> the cvs export functions to avoid getting old empty directories.
>
> So, you have to remove all the empty directories by yourself now,
> and run again make datainstall
>
> Without the data, the programs crashes badly as it is really not
> robust when the data files are not present.
>
> Eric.

Hi all,

I have put the 1.2.0-rc1 in limited diffusion for tests.
The download information are <http://torcs.org/tmp/download.html>

Please report any problem/remarks to the mailing lists
and don't forget to list the environment (OS/version/video...)
generally it eases the debugging process.

Bye,
Eric.
-- 
=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=
                        TORCS
             The Open Racing Car Simulator
                  http://torcs.org
=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=

G.Galli wrote:

>Hi,
>
>i need some help making my first torcs track.
>i generated a simple track model with blender (only a road circle with
>texture), but i don't know how i can make a real torcs track from the
>blender file....
>i've read the (only) track-editing posting on the torcs forum. eric
>wrote that you have to make an xml file to describe the track,but i need
>some documentation about what i have to write in this xml file...
>Or could someone write a short howto about track editing in general?;)
>
>Gernot
>

Hello,

Making tracks (and cars) are not really easy with TORCS.

Some one is currently trying to make a track editor. ( Eric, could tell 
us more about that).
I tryed, months ago, to do an XML track editor in GTK, but I  didn't 
finish it ....

If you want to create a track, you have to follow the following 
"undocumented" steps:
    - First, you have to know that tracks are described in XML, and the 
XML file
is more important than the 3D geometry. The XML file contains every 
thing needed
for the simulation. It contains the full descriptions of the track's 
segments (curve, banking,
surfaces, ...).
    - Before making any 3D geometry you have to describe your track in 
XML. The best way to
learn the (very proprietary) XML format is to take a simple existing 
track, such as a basic oval.
    - in the src/tools directory of TORCS, there is a program called 
"trackgen". This program is used
to transform the XML decription into AC3D files. Trackgen is able, using 
the XML file, to do
automatically many things, such as making the inner and the outer 3D 
geometry of the track, placing
trees or objects, putting textures  .....
But it is very tricky to use it. I suppose that only the author of 
trackgen is able to use the best of trackgen. :-) ....
    - but trackgen is very  useful to determine if your track "loops" 
(or joins, sorry I don't know an
english word for that ), and also for creating the basic 3D file in AC3D 
format.
    - Once you have the "basic generated" AC3D file, you can modify it 
as you want, but you MUST
respect the track path and geometry. If you don't respect it the cars 
will follow the track described
in the XML file, but your 3D geometry will describe something else, and 
the result will be completly
absurd under TORCS.
    - There is also a very "useful" tool, it is called "accc", you may 
have to use it for compiling the AC3D
files (for STRIPE optimization), or for assembling different AC3D files 
for making multi-texturing.
An example of multi-texturing can be found in g-track-3 (only in CVS 
right now) for making the
shadows on the tracks.

Ok,Ok, now,:-, I know why no body made tracks :-)

I personnaly spend days for making g-track-3.

And now, you know what the (*)  means in the TORCS video, in the
"build you own cars and tracks (*)" sentence.

christophe

On Mon, 2003-02-10 at 19:36, tor...@li...
wrote:
> Torcs-devel -- confirmation of subscription -- request 832076
> 
> We have received a request from 217.235.33.66 for subscription of your
> email address, <za...@we...>, to the
> tor...@li... mailing list.  To confirm the
> request, please send a message to
> tor...@li..., and either:
> 
> - maintain the subject line as is (the reply's additional "Re:" is
> ok),
> 
> - or include the following line - and only the following line - in the
> message body: 
> 
> confirm 832076
> 
> (Simply sending a 'reply' to this message should work from most email
> interfaces, since that usually leaves the subject line in the right
> form.)
> 
> If you do not wish to subscribe to this list, please simply disregard
> this message.  Send questions to
> tor...@li....
> 

Hi,

i need some help making my first torcs track.
i generated a simple track model with blender (only a road circle with
texture), but i don't know how i can make a real torcs track from the
blender file....
i've read the (only) track-editing posting on the torcs forum. eric
wrote that you have to make an xml file to describe the track,but i need
some documentation about what i have to write in this xml file...
Or could someone write a short howto about track editing in general?;)

Gernot

On Mon, 2003-02-10 at 19:36, tor...@li...
wrote:
> Torcs-devel -- confirmation of subscription -- request 832076
> 
> We have received a request from 217.235.33.66 for subscription of your
> email address, <za...@we...>, to the
> tor...@li... mailing list.  To confirm the
> request, please send a message to
> tor...@li..., and either:
> 
> - maintain the subject line as is (the reply's additional "Re:" is
> ok),
> 
> - or include the following line - and only the following line - in the
> message body: 
> 
> confirm 832076
> 
> (Simply sending a 'reply' to this message should work from most email
> interfaces, since that usually leaves the subject line in the right
> form.)
> 
> If you do not wish to subscribe to this list, please simply disregard
> this message.  Send questions to
> tor...@li....
> 

On Mon, 2003-02-03 at 14:54, tor...@li...
wrote:
> Torcs-devel -- confirmation of subscription -- request 201931
> 
> We have received a request from 217.235.39.5 for subscription of your
> email address, <ger...@on...>, to the
> tor...@li... mailing list.  To confirm the
> request, please send a message to
> tor...@li..., and either:
> 
> - maintain the subject line as is (the reply's additional "Re:" is
> ok),
> 
> - or include the following line - and only the following line - in the
> message body: 
> 
> confirm 201931
> 
> (Simply sending a 'reply' to this message should work from most email
> interfaces, since that usually leaves the subject line in the right
> form.)
> 
> If you do not wish to subscribe to this list, please simply disregard
> this message.  Send questions to
> tor...@li....

On Sun, 9 Feb 2003, Eric Espie wrote:

> Hi,
>
> As CVS do not remember versions on directories, it is
> necessary to use the -P option with the cvs co or
> the cvs export functions to avoid getting old empty directories.

Thanks, sic, another lession learnt from cvs, lots of thanks ;-)

> So, you have to remove all the empty directories by yourself now,
> and run again make datainstall

Thanks, did cvs up -dP and it's gone.

> Without the data, the programs crashes badly as it is really not
> robust when the data files are not present.

Understood, but I think it would be good to have at least some
short 'file xxx missing, check make datainstall' message, to
accelerate the work of people looking at torcs and not scare
them so that they never come back because of installing problems.

I think the cvs tip to use -P should also be noted somewhere on
a torcs cvs download / development starter page,
Bernd

> Eric.
>
> Bernhard Kaindl wrote:
> > Hi,
> >    I checkout the CVS today (installed gcc-3.3 also), and noticed
> > that make datainstall fails in data/cars/categories/offroad-r-a
> > because it does not include a Makefile(I guess it should be removed
> > or disabled in data/cars/categories/Makefile)
> >
> > So I went to do it this way:
> >
> > make install
> > (cd data;LD_LIBRARY_PATH=/usr/local/lib gdb ../src/linux/torcs )
> >
> > Is this ok?
> >
> > Program received signal SIGFPE, Arithmetic exception.
> > [Switching to Thread 16384 (LWP 19343)]
> > disabling TCL support
> > No race manager available
> >
> > Program received signal SIGSEGV, Segmentation fault.
> > 0x40344c55 in GfRlstGetFirst(tRingListHead*) (head=0x10) at ringlist.cpp:134
> > 134         if (head->next == (tRingList*)head) {
> > (gdb) bt
> > #0  0x40344c55 in GfRlstGetFirst(tRingListHead*) (head=0x10) at ringlist.cpp:134
> > #1  0x40327027 in GenCarsInfo () at driverconfig.cpp:241
> > #2  0x4032805d in TorcsDriverMenuInit(void*) (prevMenu=0x8106400) at driverconfig.cpp:566
> > #3  0x40326905 in TorcsOptionOptionInit(void*) (precMenu=0x8105410) at optionmenu.cpp:50
> > #4  0x40326493 in TorcsMainMenuInit() () at mainmenu.cpp:81
> > #5  0x403263a8 in TorcsEntry() () at entry.cpp:48
> > #6  0x08048ccf in main (argc=1, argv=0xbffff574) at main.cpp:52
> > #7  0x40478917 in __libc_start_main () from /lib/libc.so.6
> > (gdb)
> >
> > As *head is 0x10, head->next is not in accessible memory.
> >
> > Same with gcc-3.2 and gcc-3.3, as hack I could add an easy check
> > for a small value but this won't fix the source of the problem.
> >
> > Does anybody know a date/tag which should work?
> >
> > (I wanted to extend the human driver to work also with kbd/mouse
> >  and didn't want to work on ancient code)
> >
> > Many thanks, btw, great looking project otherwise!
> > Bernd
> >
> >
> > -------------------------------------------------------
> > This SF.NET email is sponsored by:
> > SourceForge Enterprise Edition + IBM + LinuxWorld = Something 2 See!
> > http://www.vasoftware.com
> > _______________________________________________
> > Torcs-devel mailing list
> > Tor...@li...
> > https://lists.sourceforge.net/lists/listinfo/torcs-devel
> >
> >
>
>
> --
> =+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=
>                         TORCS
>              The Open Racing Car Simulator
>                   http://torcs.org
> =+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=
>
>
>
> -------------------------------------------------------
> This SF.NET email is sponsored by:
> SourceForge Enterprise Edition + IBM + LinuxWorld = Something 2 See!
> http://www.vasoftware.com
> _______________________________________________
> Torcs-devel mailing list
> Tor...@li...
> https://lists.sourceforge.net/lists/listinfo/torcs-devel
>

Hi,

As CVS do not remember versions on directories, it is
necessary to use the -P option with the cvs co or
the cvs export functions to avoid getting old empty directories.

So, you have to remove all the empty directories by yourself now,
and run again make datainstall

Without the data, the programs crashes badly as it is really not
robust when the data files are not present.

Eric.

Bernhard Kaindl wrote:
> Hi,
>    I checkout the CVS today (installed gcc-3.3 also), and noticed
> that make datainstall fails in data/cars/categories/offroad-r-a
> because it does not include a Makefile(I guess it should be removed
> or disabled in data/cars/categories/Makefile)
> 
> So I went to do it this way:
> 
> make install
> (cd data;LD_LIBRARY_PATH=/usr/local/lib gdb ../src/linux/torcs )
> 
> Is this ok?
> 
> Program received signal SIGFPE, Arithmetic exception.
> [Switching to Thread 16384 (LWP 19343)]
> disabling TCL support
> No race manager available
> 
> Program received signal SIGSEGV, Segmentation fault.
> 0x40344c55 in GfRlstGetFirst(tRingListHead*) (head=0x10) at ringlist.cpp:134
> 134         if (head->next == (tRingList*)head) {
> (gdb) bt
> #0  0x40344c55 in GfRlstGetFirst(tRingListHead*) (head=0x10) at ringlist.cpp:134
> #1  0x40327027 in GenCarsInfo () at driverconfig.cpp:241
> #2  0x4032805d in TorcsDriverMenuInit(void*) (prevMenu=0x8106400) at driverconfig.cpp:566
> #3  0x40326905 in TorcsOptionOptionInit(void*) (precMenu=0x8105410) at optionmenu.cpp:50
> #4  0x40326493 in TorcsMainMenuInit() () at mainmenu.cpp:81
> #5  0x403263a8 in TorcsEntry() () at entry.cpp:48
> #6  0x08048ccf in main (argc=1, argv=0xbffff574) at main.cpp:52
> #7  0x40478917 in __libc_start_main () from /lib/libc.so.6
> (gdb)
> 
> As *head is 0x10, head->next is not in accessible memory.
> 
> Same with gcc-3.2 and gcc-3.3, as hack I could add an easy check
> for a small value but this won't fix the source of the problem.
> 
> Does anybody know a date/tag which should work?
> 
> (I wanted to extend the human driver to work also with kbd/mouse
>  and didn't want to work on ancient code)
> 
> Many thanks, btw, great looking project otherwise!
> Bernd
> 
> 
> -------------------------------------------------------
> This SF.NET email is sponsored by:
> SourceForge Enterprise Edition + IBM + LinuxWorld = Something 2 See!
> http://www.vasoftware.com
> _______________________________________________
> Torcs-devel mailing list
> Tor...@li...
> https://lists.sourceforge.net/lists/listinfo/torcs-devel
> 
> 


-- 
=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=
                        TORCS
             The Open Racing Car Simulator
                  http://torcs.org
=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=

Hi,
   I checkout the CVS today (installed gcc-3.3 also), and noticed
that make datainstall fails in data/cars/categories/offroad-r-a
because it does not include a Makefile(I guess it should be removed
or disabled in data/cars/categories/Makefile)

So I went to do it this way:

make install
(cd data;LD_LIBRARY_PATH=/usr/local/lib gdb ../src/linux/torcs )

Is this ok?

Program received signal SIGFPE, Arithmetic exception.
[Switching to Thread 16384 (LWP 19343)]
disabling TCL support
No race manager available

Program received signal SIGSEGV, Segmentation fault.
0x40344c55 in GfRlstGetFirst(tRingListHead*) (head=0x10) at ringlist.cpp:134
134         if (head->next == (tRingList*)head) {
(gdb) bt
#0  0x40344c55 in GfRlstGetFirst(tRingListHead*) (head=0x10) at ringlist.cpp:134
#1  0x40327027 in GenCarsInfo () at driverconfig.cpp:241
#2  0x4032805d in TorcsDriverMenuInit(void*) (prevMenu=0x8106400) at driverconfig.cpp:566
#3  0x40326905 in TorcsOptionOptionInit(void*) (precMenu=0x8105410) at optionmenu.cpp:50
#4  0x40326493 in TorcsMainMenuInit() () at mainmenu.cpp:81
#5  0x403263a8 in TorcsEntry() () at entry.cpp:48
#6  0x08048ccf in main (argc=1, argv=0xbffff574) at main.cpp:52
#7  0x40478917 in __libc_start_main () from /lib/libc.so.6
(gdb)

As *head is 0x10, head->next is not in accessible memory.

Same with gcc-3.2 and gcc-3.3, as hack I could add an easy check
for a small value but this won't fix the source of the problem.

Does anybody know a date/tag which should work?

(I wanted to extend the human driver to work also with kbd/mouse
 and didn't want to work on ancient code)

Many thanks, btw, great looking project otherwise!
Bernd

Hi all,

If you did not notice, There is a new version of the
AC3D modeler (the one I used for the cars and tracks).

There are plenty of new features like texture UV editor,
LOD generation, near snap of points...

The announce was:
-----------------------
New version of AC3D now available

A new version of this popular and easy to use 3D object/scene modeler is now
available for Windows and Linux.

The latest version includes faster rendering, improved 3DS support and a
graphical texture coordinate editor.

AC3D supports many file formats including POV-Ray, VRML (1 and 2), RenderMan
,3D Studio 3DS, Lightwave, DXF, Alias  triangle, Wavefront OBJ, Direct X,
Milkshape, MD2, Quake 3 BSP.

For more information and downloads, see http://www.ac3d.org

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

Bye,
Eric.
-- 
=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=
                        TORCS
             The Open Racing Car Simulator
                  http://torcs.org
=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=