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

Strangely enough, the engine revs up faster with first gear than when
on neutral on some cars. This happens although in transmission.cpp the
total drivetrain inertia is taken to be the sum of engine inertia +
gear inertia. Any idea what is causing that?



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

Hi All

> Couldn't get kde to run properly so have gone back to X 4.2 and instead 
> have reverted the full screen changes.
> Works great, love the new sounds, car damage etc.

I will have a look into the build system and make it also compile if 
Xrandr is not available (in fact it is prepared, the missing bits are 
the creation of a config.h, include it and fix the name of the #ifdef 
guards according to config.h).

bye, Bernhard.


-- 

visit my homepage http://www.berniw.org

Geoff Reidy wrote:
> 
>>
>> Anyway I updated by cvs up -dPA to see the newest developments and get
>> this undefined reference to `XRenderSetSubpixelOrder' error.
>> I am running mandrake 9.0 with XFree86-4.2.1-3.5.90mdk and nvidia driver
>> 5336.
>>
> 
> Progress report:
> 
> I've updated to XFree86-4.3-8.7.91mdk and now current torcs compiles 
> cleanly. Only catch is my kde is not happy but at least icewm still works.
> Looks like XFree86 > 4.2.1 will be needed to run newer torcs versions.
> 

Couldn't get kde to run properly so have gone back to X 4.2 and instead 
have reverted the full screen changes.
Works great, love the new sounds, car damage etc.

Geoff

Christos Dimitrakakis wrote:
> I have submitted the new textures.. I am not sure if they are on the
> default branch or on the main trunk.. how can I put them on the main
> CVS trunk?
> 
It's ok, they are on the main branch.

Eric.
-- 
=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=
               TORCS - http://torcs.org
             The Open Racing Car Simulator
         AKA The Other Release Coming Soon (Skin'r)
            How soon is soon ? (RaceBlizter)
=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=

CVS: Added Files:
CVS:    ancient-concrete-bump.rgb basalt1.rgb basalt1_2.rgb
CVS:    basalt2.rgb basalt2_2.rgb basalt2_2b.rgb bushes_side.png
CVS:    cut_red_earth_side.png earth_and_grass.png fields_small.png
CVS:    forest_far_above.png grey_rock_side.png hard-dirt.rgb
CVS:    hard-dirt1-side.rgb hard-dirt1.rgb mud-gravel-side.rgb
CVS:    mud-gravel1.rgb mud-hard-dirt.rgb mud-river.rgb red_dirt.png
CVS:    red_earth.png red_rock_side.png road_concrete_boundary.png
CVS:    road_earth_boundary.png road_gravel_boundary.png
CVS:    rusted_barrier.png trans-mud-hard-dirt.rgb
CVS:    yellow_rock_side.png
CVS:
--------------------------------------------------------------------

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

I have submitted the new textures.. I am not sure if they are on the
default branch or on the main trunk.. how can I put them on the main
CVS trunk?

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

> 
> Anyway I updated by cvs up -dPA to see the newest developments and get
> this undefined reference to `XRenderSetSubpixelOrder' error.
> I am running mandrake 9.0 with XFree86-4.2.1-3.5.90mdk and nvidia driver
> 5336.
> 

Progress report:

I've updated to XFree86-4.3-8.7.91mdk and now current torcs compiles 
cleanly. Only catch is my kde is not happy but at least icewm still works.
Looks like XFree86 > 4.2.1 will be needed to run newer torcs versions.

Regards,
Geoff

Bernhard Wymann wrote:
> Hi Geoff
> 
>> Only by updating the whole distro, which is a hit or miss process with 
>> mandrake, bit tempted to try Debian or Gentoo instead, in the meantime 
>> will try and rebuild a source rpm from mandrake 9.1...
> 
> 
> Are there not kind of "inofficial" Mandrake downloads with KDE, XFreee, 
> etc updates? I just ask because SuSE makes available such "inofficial" 
> unsupported updates. If you compile XFree for the first time yourself it 
> might cause some trouble...
> 
> bye, Bernhard.
> 

I'm running an unofficial version of KDE but haven't found one for X. 
Rebuilding the source rpm is as easy as
rpm --rebuild XFree86-4.3-8.7.91mdk.src.rpm
though it doesn't always work for packages from newer versions but it's 
worth a try.

Geoff

Hi Geoff

> Only by updating the whole distro, which is a hit or miss process with 
> mandrake, bit tempted to try Debian or Gentoo instead, in the meantime 
> will try and rebuild a source rpm from mandrake 9.1...

Are there not kind of "inofficial" Mandrake downloads with KDE, XFreee, 
etc updates? I just ask because SuSE makes available such "inofficial" 
unsupported updates. If you compile XFree for the first time yourself it 
might cause some trouble...

bye, Bernhard.

-- 

visit my homepage http://www.berniw.org

Bernhard Wymann wrote:
> Hi Geoff
> 
>> cvs up -r1-2-2 (or -r 1-2-2)gives me:
>> cvs [server aborted]: Numeric tag 1-2-2 contains characters other than 
>> digits and '.'
> 
> 
> What if you try -r "r1-2-2"?

OK, that's working, and I know that version builds ok.

> 
>> I poked around in my version of Xrender.h and there is no 
>> "XRenderSetSubpixelOrder" there. I found a version of this file on the 
>> net which does have it so I think the problem is with my version of X.
>> Looks like I've fallen behind again :(
>> I might try and get hold of X 4.3 and build it.
> 
> 
> Is there no binary update available for you distro?
> 

Only by updating the whole distro, which is a hit or miss process with 
mandrake, bit tempted to try Debian or Gentoo instead, in the meantime 
will try and rebuild a source rpm from mandrake 9.1...

> bye, Bernhard.
> 

Thanks again,
Geoff

Hi Geoff

> cvs up -r1-2-2 (or -r 1-2-2)gives me:
> cvs [server aborted]: Numeric tag 1-2-2 contains characters other than 
> digits and '.'

What if you try -r "r1-2-2"?

> I poked around in my version of Xrender.h and there is no 
> "XRenderSetSubpixelOrder" there. I found a version of this file on the 
> net which does have it so I think the problem is with my version of X.
> Looks like I've fallen behind again :(
> I might try and get hold of X 4.3 and build it.

Is there no binary update available for you distro?

bye, Bernhard.

-- 

visit my homepage http://www.berniw.org

Oops, sorry about the posting mix up, should really be asleep :)

Bernhard Wymann wrote:
> Hi all
> 
>> For the CVS tree, I forgot to make the symbolic tag for the 1.2.2 
>> version :-(
> 
> 
> But there is a r1-2-2 visible in the ViewCVS?
> 
>> For your compilation problem, I don't know where the problem is,
>> so I forward this to the devel-list, if someone has an idea...
>>
>> Eric.
>>
>>
>> -------- Original Message --------
>> Subject: torcs compilation problem
>> Date: Fri, 16 Apr 2004 21:19:56 +1000
>> From: Geoff Reidy <gh...@pa...>
>> To: eri...@fr...
>>
>> Hi Eric,
>>
>> I am currently running torcs 1.2.2 (I think) which I downloaded via cvs
>> -r 1.2.2 (from memory) and this version compiled and runs nicely, it has
>> the new Kcendra historic cars.
>> Sorry for the confusion here but I am now unable to get a good source
>> tree using -r 1.2.2 (no configure or configure.in?). However -D '6 weeks
> 
> 
> Perhaps you need "r1-2-2" instead of "1.2.2" ?

cvs up -r1-2-2 (or -r 1-2-2)gives me:
cvs [server aborted]: Numeric tag 1-2-2 contains characters other than 
digits and '.'

> 
>> ago' does give me a good source tree which compiles cleanly.
>>
>> Anyway I updated by cvs up -dPA to see the newest developments and get
>> this undefined reference to `XRenderSetSubpixelOrder' error.
>> I am running mandrake 9.0 with XFree86-4.2.1-3.5.90mdk and nvidia driver
>> 5336.
>>
>> If I remove the test from configure, leaving only:
>>
>> #define HAVE_LIBXRENDER 1
>>
>>    LIBS="-lXrender $LIBS"
>>
>> configure will pass but make fails with:
>>
>> g++ -I/usr/local/src/torcs-new/torcs/torcs//export/include  -g -O2 -Wall
>> -Wstrict-prototypes -O3 -mieee-fp -Wall -Wstrict-prototypes -O3
>> -mieee-fp  -D_SVID_SOURCE -D_BSD_SOURCE -DSHM  -c screen.cpp
>> screen.cpp: In function `void gfScreenInit()':
>> screen.cpp:123: `XRRConfigSizes' undeclared (first use this function)
>> screen.cpp:123: (Each undeclared identifier is reported only once for
>> each function it appears in.)
>> screen.cpp:161: `XRRFreeScreenConfigInfo' undeclared (first use this
>> function)
>> make[3]: *** [screen.o] Error 1
>> make[3]: Leaving directory
>> `/usr/local/src/torcs-new/torcs/torcs/src/libs/tgfclient'
>> make[2]: *** [subdirs] Error 1
>> make[2]: Leaving directory 
>> `/usr/local/src/torcs-new/torcs/torcs/src/libs'
>> make[1]: *** [subdirs] Error 1
>> make[1]: Leaving directory `/usr/local/src/torcs-new/torcs/torcs/src'
>> make: *** [subdirs] Error 1
>> [
> 
> 
> I have modified the method for the full screen switch with GLUT to use 
> the Xrandr extension, it worked with the CVS main branch when I checked 
> it in. I recommend a clean new checkout of the cvs main branch or "r1-2-2".
> 
> bye, Bernhard.
> 

I poked around in my version of Xrender.h and there is no 
"XRenderSetSubpixelOrder" there. I found a version of this file on the 
net which does have it so I think the problem is with my version of X.
Looks like I've fallen behind again
I might try and get hold of X 4.3 and build it.

Thanks for getting back to me.

Regards,
Geoff

Bernhard Wymann wrote:
> Hi all
> 
>> For the CVS tree, I forgot to make the symbolic tag for the 1.2.2 
>> version :-(
> 
> 
> But there is a r1-2-2 visible in the ViewCVS?
> 
>> For your compilation problem, I don't know where the problem is,
>> so I forward this to the devel-list, if someone has an idea...
>>
>> Eric.
>>
>>
>> -------- Original Message --------
>> Subject: torcs compilation problem
>> Date: Fri, 16 Apr 2004 21:19:56 +1000
>> From: Geoff Reidy <gh...@pa...>
>> To: eri...@fr...
>>
>> Hi Eric,
>>
>> I am currently running torcs 1.2.2 (I think) which I downloaded via cvs
>> -r 1.2.2 (from memory) and this version compiled and runs nicely, it has
>> the new Kcendra historic cars.
>> Sorry for the confusion here but I am now unable to get a good source
>> tree using -r 1.2.2 (no configure or configure.in?). However -D '6 weeks
> 
> 
> Perhaps you need "r1-2-2" instead of "1.2.2" ?
> 
>> ago' does give me a good source tree which compiles cleanly.
>>
>> Anyway I updated by cvs up -dPA to see the newest developments and get
>> this undefined reference to `XRenderSetSubpixelOrder' error.
>> I am running mandrake 9.0 with XFree86-4.2.1-3.5.90mdk and nvidia driver
>> 5336.
>>
>> If I remove the test from configure, leaving only:
>>
>> #define HAVE_LIBXRENDER 1
>>
>>    LIBS="-lXrender $LIBS"
>>
>> configure will pass but make fails with:
>>
>> g++ -I/usr/local/src/torcs-new/torcs/torcs//export/include  -g -O2 -Wall
>> -Wstrict-prototypes -O3 -mieee-fp -Wall -Wstrict-prototypes -O3
>> -mieee-fp  -D_SVID_SOURCE -D_BSD_SOURCE -DSHM  -c screen.cpp
>> screen.cpp: In function `void gfScreenInit()':
>> screen.cpp:123: `XRRConfigSizes' undeclared (first use this function)
>> screen.cpp:123: (Each undeclared identifier is reported only once for
>> each function it appears in.)
>> screen.cpp:161: `XRRFreeScreenConfigInfo' undeclared (first use this
>> function)
>> make[3]: *** [screen.o] Error 1
>> make[3]: Leaving directory
>> `/usr/local/src/torcs-new/torcs/torcs/src/libs/tgfclient'
>> make[2]: *** [subdirs] Error 1
>> make[2]: Leaving directory 
>> `/usr/local/src/torcs-new/torcs/torcs/src/libs'
>> make[1]: *** [subdirs] Error 1
>> make[1]: Leaving directory `/usr/local/src/torcs-new/torcs/torcs/src'
>> make: *** [subdirs] Error 1
>> [
> 
> 
> I have modified the method for the full screen switch with GLUT to use 
> the Xrandr extension, it worked with the CVS main branch when I checked 
> it in. I recommend a clean new checkout of the cvs main branch or "r1-2-2".
> 
> bye, Bernhard.
> 

Hi all

> For the CVS tree, I forgot to make the symbolic tag for the 1.2.2 
> version :-(

But there is a r1-2-2 visible in the ViewCVS?

> For your compilation problem, I don't know where the problem is,
> so I forward this to the devel-list, if someone has an idea...
> 
> Eric.
> 
> 
> -------- Original Message --------
> Subject: torcs compilation problem
> Date: Fri, 16 Apr 2004 21:19:56 +1000
> From: Geoff Reidy <gh...@pa...>
> To: eri...@fr...
> 
> Hi Eric,
> 
> I am currently running torcs 1.2.2 (I think) which I downloaded via cvs
> -r 1.2.2 (from memory) and this version compiled and runs nicely, it has
> the new Kcendra historic cars.
> Sorry for the confusion here but I am now unable to get a good source
> tree using -r 1.2.2 (no configure or configure.in?). However -D '6 weeks

Perhaps you need "r1-2-2" instead of "1.2.2" ?

> ago' does give me a good source tree which compiles cleanly.
> 
> Anyway I updated by cvs up -dPA to see the newest developments and get
> this undefined reference to `XRenderSetSubpixelOrder' error.
> I am running mandrake 9.0 with XFree86-4.2.1-3.5.90mdk and nvidia driver
> 5336.
> 
> If I remove the test from configure, leaving only:
> 
> #define HAVE_LIBXRENDER 1
> 
>    LIBS="-lXrender $LIBS"
> 
> configure will pass but make fails with:
> 
> g++ -I/usr/local/src/torcs-new/torcs/torcs//export/include  -g -O2 -Wall
> -Wstrict-prototypes -O3 -mieee-fp -Wall -Wstrict-prototypes -O3
> -mieee-fp  -D_SVID_SOURCE -D_BSD_SOURCE -DSHM  -c screen.cpp
> screen.cpp: In function `void gfScreenInit()':
> screen.cpp:123: `XRRConfigSizes' undeclared (first use this function)
> screen.cpp:123: (Each undeclared identifier is reported only once for
> each function it appears in.)
> screen.cpp:161: `XRRFreeScreenConfigInfo' undeclared (first use this
> function)
> make[3]: *** [screen.o] Error 1
> make[3]: Leaving directory
> `/usr/local/src/torcs-new/torcs/torcs/src/libs/tgfclient'
> make[2]: *** [subdirs] Error 1
> make[2]: Leaving directory `/usr/local/src/torcs-new/torcs/torcs/src/libs'
> make[1]: *** [subdirs] Error 1
> make[1]: Leaving directory `/usr/local/src/torcs-new/torcs/torcs/src'
> make: *** [subdirs] Error 1
> [

I have modified the method for the full screen switch with GLUT to use 
the Xrandr extension, it worked with the CVS main branch when I checked 
it in. I recommend a clean new checkout of the cvs main branch or "r1-2-2".

bye, Bernhard.

-- 

visit my homepage http://www.berniw.org

For the CVS tree, I forgot to make the symbolic tag for the 1.2.2 version :-(
For your compilation problem, I don't know where the problem is,
so I forward this to the devel-list, if someone has an idea...

Eric.


-------- Original Message --------
Subject: torcs compilation problem
Date: Fri, 16 Apr 2004 21:19:56 +1000
From: Geoff Reidy <gh...@pa...>
To: eri...@fr...

Hi Eric,

I am currently running torcs 1.2.2 (I think) which I downloaded via cvs
-r 1.2.2 (from memory) and this version compiled and runs nicely, it has
the new Kcendra historic cars.
Sorry for the confusion here but I am now unable to get a good source
tree using -r 1.2.2 (no configure or configure.in?). However -D '6 weeks
ago' does give me a good source tree which compiles cleanly.

Anyway I updated by cvs up -dPA to see the newest developments and get
this undefined reference to `XRenderSetSubpixelOrder' error.
I am running mandrake 9.0 with XFree86-4.2.1-3.5.90mdk and nvidia driver
5336.

If I remove the test from configure, leaving only:

#define HAVE_LIBXRENDER 1

    LIBS="-lXrender $LIBS"

configure will pass but make fails with:

g++ -I/usr/local/src/torcs-new/torcs/torcs//export/include  -g -O2 -Wall
-Wstrict-prototypes -O3 -mieee-fp -Wall -Wstrict-prototypes -O3
-mieee-fp  -D_SVID_SOURCE -D_BSD_SOURCE -DSHM  -c screen.cpp
screen.cpp: In function `void gfScreenInit()':
screen.cpp:123: `XRRConfigSizes' undeclared (first use this function)
screen.cpp:123: (Each undeclared identifier is reported only once for
each function it appears in.)
screen.cpp:161: `XRRFreeScreenConfigInfo' undeclared (first use this
function)
make[3]: *** [screen.o] Error 1
make[3]: Leaving directory
`/usr/local/src/torcs-new/torcs/torcs/src/libs/tgfclient'
make[2]: *** [subdirs] Error 1
make[2]: Leaving directory `/usr/local/src/torcs-new/torcs/torcs/src/libs'
make[1]: *** [subdirs] Error 1
make[1]: Leaving directory `/usr/local/src/torcs-new/torcs/torcs/src'
make: *** [subdirs] Error 1
[

I get the same error if I use:

#define HAVE_LIBXRENDER 0

and remove the line:

    LIBS="-lXrender $LIBS"

I am using the cvs version of plib so I had to change the version test
for plib in configure but this does work ok for the version I am
currently running.

It seems that the addition of -lXrandr breaks things for me, perhaps my
xfree verson is now too old?

Great program by the way.

Regards,
Geoff


-- 
=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=
               TORCS - http://torcs.org
             The Open Racing Car Simulator
         AKA The Other Release Coming Soon (Skin'r)
            How soon is soon ? (RaceBlizter)
=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=

Hi Christos

> I went on a photo expedition while in greece to grab some possible
> texture material. I don't have enough space to keep them on my
> website. They will require some processing to make them useable for
> the game. They include mud, dirt, rocks, treadmarks on dirt, roads,

Is it legal to use these pictures without permisson of the photographers 
(I'm a bit paranoid about such stuff, we live in a strange world, at 
least strange for me...). Apart from that it would be very nice to have 
some additional textures.

bye, Bernhard.

-- 

visit my homepage http://www.berniw.org

I went on a photo expedition while in greece to grab some possible
texture material. I don't have enough space to keep them on my
website. They will require some processing to make them useable for
the game. They include mud, dirt, rocks, treadmarks on dirt, roads,
road boundaries, bushes, some views of hills from medium distance and
a few shots that have been taken with fixed camera settings and which
might possible to construct into a panoramic view.

If anyone is interest to work with the data, tell me and we can
arrange to place them somewhere.

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

thanks for the info. Can you pass me his contact?
I desperately need to integrate this :)

Thanks
Ashwani
On Wednesday, April 14, 2004, at 02:16  PM, Eric Espie wrote:

> Bernhard Wymann wrote:
>> Hi Ashwani
>>> I was wondering what it will take to adapt the current simulation to 
>>> work on real-city 3d rendered maps? Here is what I hope to do.. take 
>>> the GIS data..convert it to AC3D file.. use this ac3d file into the 
>>> simulator. Will this work?
>> No. The simulator takes the track model described in the XML file for 
>> the whole simulation (at least for now), the AC3D file is just a 
>> "visual" object which is not considered for the simulation.
>> bye, Bernhard.
>
> Adam Courchesne is currently on a similar project,
> may be efforts can be shared on the topic.
>
> Bye,
> Eric.
> -- 
> =+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=
>               TORCS - http://torcs.org
>             The Open Racing Car Simulator
>         AKA The Other Release Coming Soon (Skin'r)
>            How soon is soon ? (RaceBlizter)
> =+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=

Bernhard Wymann wrote:
> Hi Ashwani
> 
>> I was wondering what it will take to adapt the current simulation to 
>> work on real-city 3d rendered maps? Here is what I hope to do.. take 
>> the GIS data..convert it to AC3D file.. use this ac3d file into the 
>> simulator. Will this work?
> 
> 
> No. The simulator takes the track model described in the XML file for 
> the whole simulation (at least for now), the AC3D file is just a 
> "visual" object which is not considered for the simulation.
> 
> bye, Bernhard.
> 
> 

Adam Courchesne is currently on a similar project,
may be efforts can be shared on the topic.

Bye,
Eric.
-- 
=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=
               TORCS - http://torcs.org
             The Open Racing Car Simulator
         AKA The Other Release Coming Soon (Skin'r)
            How soon is soon ? (RaceBlizter)
=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=

Hi Ashwani

> I was wondering what it will take to adapt the current simulation to 
> work on real-city 3d rendered maps? Here is what I hope to do.. take the 
> GIS data..convert it to AC3D file.. use this ac3d file into the 
> simulator. Will this work?

No. The simulator takes the track model described in the XML file for 
the whole simulation (at least for now), the AC3D file is just a 
"visual" object which is not considered for the simulation.

bye, Bernhard.


-- 

visit my homepage http://www.berniw.org

Hi,

I was wondering what it will take to adapt the current simulation to 
work on real-city 3d rendered maps? Here is what I hope to do.. take 
the GIS data..convert it to AC3D file.. use this ac3d file into the 
simulator. Will this work?

Thanks for help

Ashwani

Steve Baker is going to try and integrate my 1.6.0 fixes into 1.8.3,
so we should have that soon. Things of note in 1.8.3 that are cool:

The new particle system.
The new sky system. (so we can replace the cylinder). If we don't
replace the cylinder, someone has to fix the way it does not respond
to zoom in/out and close the hole at the top :)


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

Christos Dimitrakakis wrote:
>>>Right-o... right now I have done a first implementation of the damage,
>>>by adding a bendCar to the graphics interface. However I have a

I should say that the current damage implementation is very impressive :-)

>>>question.. where do I get a pointer to the module interface from?
>>>In raceinit.cpp, ReInit() the ReInfo structure is initialised. How can
>>>I give access to ReInfo to a module? Will declaring it as extern work?
>>>Currently I am passing a pointer to my simu module from the call to
>>>simConfig().
>>
>>This is the best way IMHO.
>>
>>But, can't the damages stay in the car structure ? why do you need the ReInfo ?
>>This structure should stay private to the race engine module, and the car
>>structure should be the place where to define the car modifications.
>>
> 
> 
> I'm calling a graphics function from within simu. Alternatively I
> could save the changes in a list of collisions and the graphics module
> could look at the list and apply them. Slightly more overhead
> involved. A third option is to include a dedicated simu/graphics
> interface.

If you read the (always in progress) network spec:
http://torcs.free.fr/specs/TORCSNetArch2.pdf

you'll see that the simu and the graphic modules won't be on the
same system, so in order to get some modularity in the architecture,
it would better use the car structure instead of direct calls between
modules ;-)

Also the cars deformations and the wheels damages should be reversible
when the cars are repaired in the pits.

> 
> 
>>>Lastly, later today I will update the sound modules, with a #define
>>>option to use the modified plib or not. That should solve any problems
>>>at least with people that use 1.6.0, while making sure that the main
>>>code is still the same. Currently it's a bit of a nightmare to
>>>maintain 2 different branches.
>>>
>>
>>How considering switching to OpenAL ? I'd like to add
>>sound support to the GUI too so it would be a little
>>more user friendly to have a sound feedback on actions.
>>
>>
> 
> 
> I will finish testnig open al next week
> 

Thanks,
Eric.
-- 
=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=
               TORCS - http://torcs.org
             The Open Racing Car Simulator
         AKA The Other Release Coming Soon (Skin'r)
            How soon is soon ? (RaceBlizter)
=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=

Christos Dimitrakakis wrote:
> I am now putting the updates on CVS in a couple of hours. Please try
> and test compilation soon as things might have broken. It seems to
> work on my system but you never know.
> 
> 
> Changes:
> 
> -- note that the changes break simuv2 because of changes in the
> interface and module hooks --.

fixed.

> 
> ssggraph:
> - grsound: New sounds are used. There is a #define OLETHROS_PLIB that
> switches between vanilla 1.6.0 and the one that uses my patches.

Not working with PLIB 1.8.0 (not using the #ifdef OLETHROS_PLIB)
got tons of:

WARNING: slScheduler: Too many pending callback events!

and the screen stays black...

> - grskidmarks: skidmarks now work have colours, texture and width that
> depend on the surface.
> - grsmoke: smoke now has speed, however there is no twirling effect
> cause by the vortex of spinning tires. The latter effect should be
> apparent for very light smoke particles.
> - grcar: there is a routine that warps the frame of the car according
> to a simple field equation. This hooks in from the simulation.

nice :-)

> 
> simuv3:
> - slightly more stable model
> - collide: it now hooks up to the graphic interface
> 
> interfaces:
> - added some more interfaces between simuv3 and ssggraph
> 
> libs/raceengineclient/raceinit:
> - passes ReInfo to the simuv3 module. Is it possible to get ReInfo in
> the simu module in some other way? If yes we can revert to that
> without breaking simuv2.
> 
> data/textures:
> new grey-tracks.rgb
> 
> data/sound:
> all the new sounds

not found, are you sure you added them to CVS ?

> 

Thanks,
Eric.
-- 
=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=
               TORCS - http://torcs.org
             The Open Racing Car Simulator
         AKA The Other Release Coming Soon (Skin'r)
            How soon is soon ? (RaceBlizter)
=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=

Christos Dimitrakakis wrote:
> Right-o... right now I have done a first implementation of the damage,
> by adding a bendCar to the graphics interface. However I have a
> question.. where do I get a pointer to the module interface from?
> In raceinit.cpp, ReInit() the ReInfo structure is initialised. How can
> I give access to ReInfo to a module? Will declaring it as extern work?
> Currently I am passing a pointer to my simu module from the call to
> simConfig().

This is the best way IMHO.

But, can't the damages stay in the car structure ? why do you need the ReInfo ?
This structure should stay private to the race engine module, and the car
structure should be the place where to define the car modifications.

> 
> What else? Oh.. it seems that interfaces/car.h misses the following
> definition, but I am not sure whether this is true or not..
> 
> #define PRM_SPEEDO_DIGITAL "speedometer digital"

well in mine it's located at line 545 but I'll check if I have commited
the file.

> 
> 
> 
> Lastly, later today I will update the sound modules, with a #define
> option to use the modified plib or not. That should solve any problems
> at least with people that use 1.6.0, while making sure that the main
> code is still the same. Currently it's a bit of a nightmare to
> maintain 2 different branches.
> 

How considering switching to OpenAL ? I'd like to add
sound support to the GUI too so it would be a little
more user friendly to have a sound feedback on actions.


Eric.
-- 
=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=
               TORCS - http://torcs.org
             The Open Racing Car Simulator
         AKA The Other Release Coming Soon (Skin'r)
            How soon is soon ? (RaceBlizter)
=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=