You can subscribe to this list here.
| 2000 |
Jan
|
Feb
|
Mar
(2) |
Apr
(5) |
May
(8) |
Jun
(2) |
Jul
|
Aug
|
Sep
|
Oct
|
Nov
|
Dec
|
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 2001 |
Jan
|
Feb
|
Mar
|
Apr
|
May
|
Jun
|
Jul
|
Aug
|
Sep
|
Oct
|
Nov
|
Dec
(4) |
| 2002 |
Jan
|
Feb
|
Mar
(3) |
Apr
|
May
|
Jun
(7) |
Jul
(6) |
Aug
(1) |
Sep
|
Oct
(9) |
Nov
(10) |
Dec
(6) |
| 2003 |
Jan
(3) |
Feb
(8) |
Mar
(2) |
Apr
(47) |
May
(23) |
Jun
(5) |
Jul
(6) |
Aug
(19) |
Sep
(13) |
Oct
(3) |
Nov
(29) |
Dec
(3) |
| 2004 |
Jan
(2) |
Feb
(89) |
Mar
(10) |
Apr
(3) |
May
(17) |
Jun
(6) |
Jul
(12) |
Aug
(25) |
Sep
(20) |
Oct
(28) |
Nov
(23) |
Dec
(9) |
| 2005 |
Jan
(18) |
Feb
(7) |
Mar
(36) |
Apr
(29) |
May
(10) |
Jun
(9) |
Jul
(35) |
Aug
(64) |
Sep
(40) |
Oct
(11) |
Nov
(3) |
Dec
(9) |
| 2006 |
Jan
(12) |
Feb
(13) |
Mar
(7) |
Apr
(10) |
May
(18) |
Jun
(19) |
Jul
(3) |
Aug
(5) |
Sep
(7) |
Oct
(18) |
Nov
(11) |
Dec
(10) |
| 2007 |
Jan
(15) |
Feb
(6) |
Mar
(10) |
Apr
(11) |
May
(10) |
Jun
(18) |
Jul
(10) |
Aug
(18) |
Sep
(31) |
Oct
(21) |
Nov
(13) |
Dec
(2) |
| 2008 |
Jan
(26) |
Feb
(15) |
Mar
(24) |
Apr
(23) |
May
(11) |
Jun
(5) |
Jul
(16) |
Aug
(11) |
Sep
(12) |
Oct
(10) |
Nov
(3) |
Dec
(16) |
| 2009 |
Jan
(18) |
Feb
(3) |
Mar
(3) |
Apr
(15) |
May
(5) |
Jun
(19) |
Jul
(4) |
Aug
(5) |
Sep
(16) |
Oct
(23) |
Nov
(7) |
Dec
(7) |
| 2010 |
Jan
(14) |
Feb
(27) |
Mar
(12) |
Apr
(10) |
May
(2) |
Jun
(8) |
Jul
(6) |
Aug
(3) |
Sep
(6) |
Oct
(4) |
Nov
(3) |
Dec
(1) |
| 2011 |
Jan
(3) |
Feb
(6) |
Mar
(2) |
Apr
(2) |
May
(2) |
Jun
(2) |
Jul
(4) |
Aug
(7) |
Sep
|
Oct
(1) |
Nov
(6) |
Dec
(2) |
| 2012 |
Jan
(3) |
Feb
(1) |
Mar
(6) |
Apr
(14) |
May
|
Jun
(7) |
Jul
(15) |
Aug
(9) |
Sep
(35) |
Oct
(28) |
Nov
(23) |
Dec
(10) |
| 2013 |
Jan
(8) |
Feb
(7) |
Mar
(17) |
Apr
(8) |
May
(17) |
Jun
(14) |
Jul
(3) |
Aug
(2) |
Sep
(22) |
Oct
(18) |
Nov
(31) |
Dec
(15) |
| 2014 |
Jan
|
Feb
(4) |
Mar
(1) |
Apr
(2) |
May
(19) |
Jun
(2) |
Jul
(1) |
Aug
(13) |
Sep
(4) |
Oct
(1) |
Nov
|
Dec
(3) |
| 2015 |
Jan
(8) |
Feb
(2) |
Mar
(4) |
Apr
(23) |
May
(4) |
Jun
(5) |
Jul
|
Aug
(3) |
Sep
(4) |
Oct
(3) |
Nov
(8) |
Dec
(4) |
| 2016 |
Jan
(4) |
Feb
(9) |
Mar
(12) |
Apr
(13) |
May
(5) |
Jun
(4) |
Jul
(1) |
Aug
(8) |
Sep
(3) |
Oct
(15) |
Nov
(10) |
Dec
(3) |
| 2017 |
Jan
(10) |
Feb
(6) |
Mar
(11) |
Apr
(15) |
May
(13) |
Jun
(6) |
Jul
(3) |
Aug
(7) |
Sep
(1) |
Oct
(11) |
Nov
(2) |
Dec
(3) |
| 2018 |
Jan
(4) |
Feb
(1) |
Mar
(1) |
Apr
|
May
(1) |
Jun
|
Jul
|
Aug
(1) |
Sep
(2) |
Oct
(3) |
Nov
(1) |
Dec
(4) |
| 2019 |
Jan
|
Feb
(1) |
Mar
(2) |
Apr
(1) |
May
(1) |
Jun
(2) |
Jul
|
Aug
|
Sep
(1) |
Oct
(3) |
Nov
|
Dec
(1) |
| 2020 |
Jan
|
Feb
(2) |
Mar
|
Apr
(1) |
May
(1) |
Jun
|
Jul
|
Aug
(1) |
Sep
|
Oct
|
Nov
|
Dec
|
| 2021 |
Jan
|
Feb
(2) |
Mar
|
Apr
(1) |
May
|
Jun
|
Jul
|
Aug
|
Sep
|
Oct
(5) |
Nov
|
Dec
|
| 2023 |
Jan
|
Feb
|
Mar
|
Apr
|
May
|
Jun
(1) |
Jul
(2) |
Aug
|
Sep
|
Oct
|
Nov
|
Dec
|
| 2024 |
Jan
|
Feb
|
Mar
|
Apr
|
May
|
Jun
|
Jul
|
Aug
|
Sep
|
Oct
|
Nov
|
Dec
(2) |
| 2025 |
Jan
(1) |
Feb
|
Mar
|
Apr
|
May
|
Jun
|
Jul
|
Aug
|
Sep
|
Oct
|
Nov
|
Dec
|
| S | M | T | W | T | F | S |
|---|---|---|---|---|---|---|
|
1
|
2
(3) |
3
|
4
(2) |
5
|
6
|
7
|
|
8
(1) |
9
|
10
(4) |
11
(5) |
12
(4) |
13
|
14
|
|
15
(1) |
16
(5) |
17
|
18
|
19
|
20
|
21
|
|
22
|
23
|
24
|
25
|
26
|
27
|
28
|
|
29
|
30
|
31
|
|
|
|
|
|
From: <r_g...@ya...> - 2004-08-16 22:06:44
|
Hi all, I solved my problem "segmentation fault". In fact the parser crash if a specific caracter is present in the xml file. In my case, the caracter was "û". Hope this help. Bye Richard. --- Richard Guignon <r_g...@ya...> a écrit : > Hi all, > > Indeed, it works now with robotgen V1.2. Thanks for > your help Bernhard. > Now I have an other problem (lin 54 : segmentation > fault) but I see some solutions in the archive > (torcs > -s or filename~). I will try it. > Le nouveau Yahoo! Messenger est arrivé ! Découvrez toutes les nouveautés pour dialoguer instantanément avec vos amis. A télécharger gratuitement sur http://fr.messenger.yahoo.com |
|
From: <r_g...@ya...> - 2004-08-16 17:22:32
|
Hi all, Indeed, it works now with robotgen V1.2. Thanks for your help Bernhard. Now I have an other problem (lin 54 : segmentation fault) but I see some solutions in the archive (torcs -s or filename~). I will try it. Tks again. Richard --- Bernhard Wymann <be...@bl...> a écrit : > Hi Richard > > > I used robotgen,v 1.3 2003/10/19 18:59:26 but I > > changed few lines to get it works : > > Ok, I think this was the problem, for the robot > tutorial you should have > used the file which is linked into the tutorial, not > the CVS one. The > CVS robotgen v1.3 was outdated, the bug is fixed > since two weeks or > so... but it still just works if you have the CVS > sources installed. > > If you follow the tutorial form the 1.2.2 packages > you need the robotgen > script from the tutorial. > > I hope this will solve the problem and we see you > robot soon:-) > Le nouveau Yahoo! Messenger est arrivé ! Découvrez toutes les nouveautés pour dialoguer instantanément avec vos amis. A télécharger gratuitement sur http://fr.messenger.yahoo.com |
|
From: Bernhard W. <be...@bl...> - 2004-08-16 15:45:32
|
Hi Richard > I used robotgen,v 1.3 2003/10/19 18:59:26 but I > changed few lines to get it works : Ok, I think this was the problem, for the robot tutorial you should have used the file which is linked into the tutorial, not the CVS one. The CVS robotgen v1.3 was outdated, the bug is fixed since two weeks or so... but it still just works if you have the CVS sources installed. If you follow the tutorial form the 1.2.2 packages you need the robotgen script from the tutorial. I hope this will solve the problem and we see you robot soon:-) bye, have fun, Bernhard. -- visit my homepage http://www.berniw.org coming soon: The TORCS Racing Board, http://www.berniw.org/trb |
|
From: <r_g...@ya...> - 2004-08-16 15:06:59
|
Hi all,
Find below more information :
--------------------
VARIABLES
[rico@localhost rico]$ env | egrep "(MAKE|TORCS)"
MAKE_DEFAULT=/usr/src/torcs/torcs-1.2.2/Make-default.mk
TORCS_BASE=/usr/src/torcs/torcs-1.2.2
--------------------
--------------------
INCLUDES
[rico@localhost rico]$ ls -l
/usr/src/torcs/torcs-1.2.2/export/include
total 44
drwxrwxrwx 2 rico rico 4096 aoû 15 14:13 3D/
lrwxrwxrwx 1 root root 47 aoû 15 14:13 car.h ->
/usr/src/torcs/torcs-1.2.2/src/interfaces/car.h
lrwxrwxrwx 1 root root 51 aoû 15 14:13 client.h ->
/usr/src/torcs/torcs-1.2.2/src/libs/client/client.h
lrwxrwxrwx 1 root root 61 aoû 15 14:13
confscreens.h ->
/usr/src/torcs/torcs-1.2.2/src/libs/confscreens/confscreens.h
....
...
..
[rico@localhost rico]$ ls -l
/usr/src/torcs/torcs-1.2.2/src/interfaces/car.h
-rw-r--r-- 1 rico rico 20582 fév 12 2004
/usr/src/torcs/torcs-1.2.2/src/interfaces/car.h
Also, If I rename the car.h I have an other error :
[rico@localhost carico]$ make
make: *** Pas de règle pour fabriquer la cible «
/usr/src/torcs/torcs-1.2.2/export/include/car.h »,
nécessaire pour « carico.o ». Arrêt.
So, I think that the access to include files is
correct.
--------------------
--------------------
ROBOTGEN
I used robotgen,v 1.3 2003/10/19 18:59:26 but I
changed few lines to get it works :
inital :
CATEGORY=$(grep category
data/cars/models/$CAR/$CAR.xml | sed -e
's/^.*val=\"//' -e 's/\".*$//')
cp data/cars/models/$CAR/$CAR.rgb $ROBOTDIR || exit 1
cp data/cars/models/$CAR/$CAR.xcf $ROBOTDIR
2>/dev/null
cp data/data/textures/logo.rgb $ROBOTDIR || exit 1.
now :
CATEGORY=$(grep category
/usr/local/share/games/torcs/cars/$CAR/$CAR.xml | sed
-e 's/^.*val=\"//' -e 's/\".*$//')
cp /usr/local/share/games/torcs/cars/$CAR/$CAR.rgb
$ROBOTDIR || exit 1
cp /usr/local/share/games/torcs/cars/$CAR/$CAR.xcf
$ROBOTDIR 2>/dev/null
cp /usr/local/share/games/torcs/data/textures/logo.rgb
$ROBOTDIR || exit 1
--------------------
--------------------
CODE
The errors is on line "memset". Extract from
"carico.cpp" :
/* Drive during race. */
static void
drive(int index, tCarElt* car, tSituation *s)
{
memset((void *)&car.ctrl, 0, sizeof(tCarCtrl));
car->ctrl.brakeCmd = 1.0; /* all brakes on ... */
/*
* add the driving code here to modify the
* car->ctrl.steer
* car->ctrl.accelCmd
* car->ctrl.brakeCmd
*/
}
I had a look into src/drivers/bt/bt.cpp
The code is different, because a "driver.h" is called
at the beginning and the memset is not done in bt.cpp.
/* Drive during race. */
static void drive(int index, tCarElt* car, tSituation
*s)
{
driver[index]->drive(s);
}
For memory the memset is done is the driver.h,
something like :
memset(&car.ctrl, 0, sizeof(tCarCtrl));
--------------------
--- Bernhard Wymann <be...@bl...> a écrit :
> Hi Richard
>
> > I follow the famous robot tutorial but...my first
> > robot compliation failed (step 1.2) :
> >
> > [rico@localhost carico]$ make
> > gcc -E -M -D__DEPEND__
> > -I/usr/src/torcs/torcs-1.2.2/export/include
> > -D_SVID_SOURCE -D_BSD_SOURCE -DSHM carico.cpp >
> > .depend
> > c++ -I/usr/src/torcs/torcs-1.2.2/export/include
> -g
> > -O2 -Wall -Wstrict-prototypes -O3 -mieee-fp -Wall
> > -Wstrict-prototypes -O3 -mieee-fp -D_SVID_SOURCE
> > -D_BSD_SOURCE -DSHM -c carico.cpp
> > carico.cpp: In function `void drive(int, tCarElt*,
> > tSituation*)':
> > carico.cpp:96: error: request for member `ctrl' in
> > `car', which is of non-class type `tCarElt*'
> > make: *** [carico.o] Erreur 1
> >
> > I don't change anything yet in the file
> carico.cpp.
> > So, it's a copy of the initial "bt" robot file...
>
> Hmm, not enough information, but look at this:
> - Are the include files accessible in
> /usr/src/torcs/torcs-1.2.2/export/include?
> - Which bt file are you using, can you post the
> lines where the error
> occurs, from where originates this file (perhaps the
> file is outdated or
> something, but I checked it against TORCS 1.2.2...)?
> - Is $TORCS_BASE and $MAKE_DEFAULT set?
>
> bye, Bernhard.
>
Le nouveau Yahoo! Messenger est arrivé ! Découvrez toutes les nouveautés pour dialoguer instantanément avec vos amis. A télécharger gratuitement sur http://fr.messenger.yahoo.com
|
|
From: Bernhard W. <be...@bl...> - 2004-08-16 07:54:59
|
Hi Richard > I follow the famous robot tutorial but...my first > robot compliation failed (step 1.2) : > > [rico@localhost carico]$ make > gcc -E -M -D__DEPEND__ > -I/usr/src/torcs/torcs-1.2.2/export/include > -D_SVID_SOURCE -D_BSD_SOURCE -DSHM carico.cpp > > .depend > c++ -I/usr/src/torcs/torcs-1.2.2/export/include -g > -O2 -Wall -Wstrict-prototypes -O3 -mieee-fp -Wall > -Wstrict-prototypes -O3 -mieee-fp -D_SVID_SOURCE > -D_BSD_SOURCE -DSHM -c carico.cpp > carico.cpp: In function `void drive(int, tCarElt*, > tSituation*)': > carico.cpp:96: error: request for member `ctrl' in > `car', which is of non-class type `tCarElt*' > make: *** [carico.o] Erreur 1 > > I don't change anything yet in the file carico.cpp. > So, it's a copy of the initial "bt" robot file... Hmm, not enough information, but look at this: - Are the include files accessible in /usr/src/torcs/torcs-1.2.2/export/include? - Which bt file are you using, can you post the lines where the error occurs, from where originates this file (perhaps the file is outdated or something, but I checked it against TORCS 1.2.2...)? - Is $TORCS_BASE and $MAKE_DEFAULT set? bye, Bernhard. -- visit my homepage http://www.berniw.org coming soon: The TORCS Racing Board, http://www.berniw.org/trb |
|
From: <r_g...@ya...> - 2004-08-15 23:08:49
|
Hi all, I follow the famous robot tutorial but...my first robot compliation failed (step 1.2) : [rico@localhost carico]$ make gcc -E -M -D__DEPEND__ -I/usr/src/torcs/torcs-1.2.2/export/include -D_SVID_SOURCE -D_BSD_SOURCE -DSHM carico.cpp > .depend c++ -I/usr/src/torcs/torcs-1.2.2/export/include -g -O2 -Wall -Wstrict-prototypes -O3 -mieee-fp -Wall -Wstrict-prototypes -O3 -mieee-fp -D_SVID_SOURCE -D_BSD_SOURCE -DSHM -c carico.cpp carico.cpp: In function `void drive(int, tCarElt*, tSituation*)': carico.cpp:96: error: request for member `ctrl' in `car', which is of non-class type `tCarElt*' make: *** [carico.o] Erreur 1 I don't change anything yet in the file carico.cpp. So, it's a copy of the initial "bt" robot file... Any idea ? tks Richard Vous manquez despace pour stocker vos mails ? Yahoo! Mail vous offre GRATUITEMENT 100 Mo ! Créez votre Yahoo! Mail sur http://fr.benefits.yahoo.com/ Le nouveau Yahoo! Messenger est arrivé ! Découvrez toutes les nouveautés pour dialoguer instantanément avec vos amis. A télécharger gratuitement sur http://fr.messenger.yahoo.com |
|
From: Eric E. <eri...@fr...> - 2004-08-12 16:25:29
|
Salut Richard, For robot developement I recommend the exellent robot tutorial done by Bernhard Wymann available on http://www.berniw.org/ Have fun, Eric. Richard Guignon wrote: > Hi all, > > I discovered Torcs yesterday... It's a really great > project !! > > Now, I plan to built my own robot... it will be not > easy (I didn't develop anything since a long time) ... > but I will try... > > My question is : where can I download the "Reference > manual" to print it or to read it offline ? > > Thanks all, > > > > > > > Vous manquez d’espace pour stocker vos mails ? > Yahoo! Mail vous offre GRATUITEMENT 100 Mo ! > Créez votre Yahoo! Mail sur http://fr.benefits.yahoo.com/ > > Le nouveau Yahoo! Messenger est arrivé ! Découvrez toutes les nouveautés pour dialoguer instantanément avec vos amis. A télécharger gratuitement sur http://fr.messenger.yahoo.com > > > ------------------------------------------------------- > SF.Net email is sponsored by Shop4tech.com-Lowest price on Blank Media > 100pk Sonic DVD-R 4x for only $29 -100pk Sonic DVD+R for only $33 > Save 50% off Retail on Ink & Toner - Free Shipping and Free Gift. > http://www.shop4tech.com/z/Inkjet_Cartridges/9_108_r285 > _______________________________________________ > Torcs-users mailing list > Tor...@li... > https://lists.sourceforge.net/lists/listinfo/torcs-users > > -- =+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+= TORCS - http://torcs.org The Open Racing Car Simulator AKA The Other Release Coming Soon (Skin'r) How soon is soon ? (RaceBlizter) =+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+= |
|
From: <r_g...@ya...> - 2004-08-12 09:27:13
|
Hi all, I discovered Torcs yesterday... It's a really great project !! Now, I plan to built my own robot... it will be not easy (I didn't develop anything since a long time) ... but I will try... My question is : where can I download the "Reference manual" to print it or to read it offline ? Thanks all, Vous manquez despace pour stocker vos mails ? Yahoo! Mail vous offre GRATUITEMENT 100 Mo ! Créez votre Yahoo! Mail sur http://fr.benefits.yahoo.com/ Le nouveau Yahoo! Messenger est arrivé ! Découvrez toutes les nouveautés pour dialoguer instantanément avec vos amis. A télécharger gratuitement sur http://fr.messenger.yahoo.com |
|
From: <sec...@fu...> - 2004-08-12 04:58:58
|
Quoting Bernhard Wymann <be...@bl...>: > Hi Hamster > >> How do I use simv3 rather than simv2 trying to find where I saw it mentioned > > The CVS is already configured, but you can change it in > /usr/local/share/games/torcs/config (or wherever it is...) in the file > raceengine.xml, simply change "simuv2" to "simuv3" or the other way > round, depending on what you prefer. > >> before it but not finding it >> >> I'll try the cvs version too sounds good. >> >> Okay I tried not to mention the reckeless bot, but you forced it out >> of me. It's >> the Inferno bot. It's got some good moves, but overtaking seems to involve >> nudging other players. Also it drives so far past the edge it's >> constantly off >> the circuit and then just comes back without any view to other cars. Okay I >> admit it's hard to get this sort of thing right, but in the end I don't do >> practise runs against the opposition because of this. >> >> As regards complication, I understand the complexity issue, humans >> are clever > > Are they? Hmm, must have missed something... but that is another > discussion... > >> creatures, learning stuff quickly. Robots have to be programmed for >> each of the > > I disagree. Look at real drivers, overtaking is even for humans a real > issue. Without having statistics handy, just with the feeling from > watching TV I would claim that around 50% of the overtaking on not oval > tracks (real, not lapping) ends in damages, collisions or spinning > cars... so humans perform miserable as well. > I was thinking more along the lines of. "Oh my god it's Sato, let him through or he'll just clatter me anyway." Or "Oh knoit's a Minardi, I'm alot faster so I need to push by it quickly" Humans can 'adapt' (perhaps that was the word I was looking for before) more quickly than the robots, because we can reprogramme ourselves during the race, rather than just between races and wothout outside assistance (usually). I think we can put this issue to bed now :) PS I've just think I've just realised something. Bump and Rebound refer to the dampers totally. I always had the impression before that bump was a spring setting. Am I correct? With bump referring to any depression of the damper and rebound any expansion. >> situations, hard enough keeping on the circuit let alone all the rest of it. >> However, look at the setup of the car, springs, bump, rebound, bellcranks, >> suspension course, toe in, etcs. Is that not complicating things, >> considering >> it's supposed to be a 'simple' robot racing game. >> Not trying to be arguementative, but the setup of the car comes alot >> from the >> driver and the drivers style, you can't drive in place of your bot to set up >> the car as you'll have your own style. > > That is absolutely true IMHO, it depends very on the driving style what > braking balance you need, etc. But these settings help to adopt the > robot driving style to the car. I'm just against the introduction of > "whatever wonderful real features" which just waste cpu cycles and hold > potential racers away because the stuff is too complex. I think already > it is not very easy to get started... the reason why I wrote the tutorial. > >> I really do complement you on the program, it's something I've >> thought of doing >> for years, but neither the time or the expertise. Again I know I'm >> really new >> to this, but I see it with fresh eyes and desperately want to help. >> I'm quite >> happy to work along with the developers to get some more documentation out. > > I have already a little setup section in the robot tutorial, have you > seen this? > >> The final point, okay you use old hardware, much like myself. My >> Duron 850 is >> the most powerful thing I'm likely to have for years. The point was with the >> way things are going, my processor is likely to be incredibly slow >> compared to >> most, therefore use of extra processing power most likely won't get >> screams of >> protest from most users. > > Hmm, its hard to say... I wonder sometimes myself when people come up > with R128 and Matrox G450 problems, but I think there are more "old" > machines out there than one might believe, not all people want or can > spend the money to exchange their hardware every 2 to 4 years. E.g. my > girlfriends box is a PII-266, 192MB RAM, voodoo banshee 16MB... too slow > for TORCS, but my K6-III 400 is still alive. > > Of course there is also the possibility to allow the user/system to > switch off some expensive features. > >> Anyway contact me with what I can do to help, in the meantime I'll >> mumble in the >> background and complain. I'll try to get some explainations underway of how >> stuff I understand works. > > Do not do that:-) The cool thing on such projects is that the source is > open... so if you think feature X is missing, simply ask loud if anybody > is already working on it, and if not -> start, if yes -> synchronize and > help or look for another feature you want. The most important thing is > that you have a good time doing it:-) > > There is actually a lot to do, if you are in the mood: > - Configuration screen to enable/disable telemetry, write telemetry to > $HOME/.torcs > - Write a complete track building tutorial or better, a track editor. > - Car building tutorial. > - Setup guide. > - New Tracks, Cars, Robots. > - What about a telemetry viewer in TORCS? > - Setup editor anyone (here we would need to agree about the directory > structure first). > - Automatic setup. > - SDL port. > - Cygwin/MinGW port. > - .... I think I could tell you 100 more... > > Btw. what the TODO list concerns, there are some informations: > - http://sourceforge.net/pm/?group_id=3777 > - http://sourceforge.net/developer/diary.php?diary_user=318462 > - the FAQ (look at berniw.org or the CVS for the newest one). > - search the mailing list archives for "task", "future". > > bye, have fun, > > Bernhard. > > > -- > > visit my homepage http://www.berniw.org > coming soon: The TORCS Racing Board, http://www.berniw.org/trb |
|
From: <sec...@fu...> - 2004-08-11 21:20:51
|
Quoting Eric Espie <eri...@fr...>: > sec...@fu... wrote: >> How do I use simv3 rather than simv2 trying to find where I saw it mentioned >> before it but not finding it > > It's located in the /usr/local/share/game/torcs/config/raceengine.xml > >> >> I'll try the cvs version too sounds good. >> >> Okay I tried not to mention the reckeless bot, but you forced it out >> of me. It's >> the Inferno bot. It's got some good moves, but overtaking seems to involve >> nudging other players. Also it drives so far past the edge it's >> constantly off >> the circuit and then just comes back without any view to other cars. Okay I >> admit it's hard to get this sort of thing right, but in the end I don't do >> practise runs against the opposition because of this. > > It's a question of tuning, I did not have the patience to tune > all the robots on all the tracks... > It's difficult to do a fast robot not aggressive some time ;-) > you should note that it's never deliberate, because it's always > causing a lot of dammages. > It looks like I'm complaining, I'm not :) I realise the bot is not doing it intentionally, I mentioned this element of judging other cars is hard. The Inferno bot is fast, real fast. Problem is when it gets it wrong, it seems to collect all the other cars with it. :) if only my bot was that fast and also avoided collisions >> >> As regards complication, I understand the complexity issue, humans >> are clever >> creatures, learning stuff quickly. Robots have to be programmed for >> each of the >> situations, hard enough keeping on the circuit let alone all the rest of it. >> However, look at the setup of the car, springs, bump, rebound, bellcranks, >> suspension course, toe in, etcs. Is that not complicating things, >> considering >> it's supposed to be a 'simple' robot racing game. > > yes, but using other bots car setup can be a good starting point. however setup is still critical, from an also-ran to a winner > >> Not trying to be arguementative, but the setup of the car comes alot >> from the >> driver and the drivers style, you can't drive in place of your bot to set up >> the car as you'll have your own style. >> I really do complement you on the program, it's something I've >> thought of doing >> for years, but neither the time or the expertise. Again I know I'm >> really new >> to this, but I see it with fresh eyes and desperately want to help. >> I'm quite >> happy to work along with the developers to get some more documentation out. >> >> The final point, okay you use old hardware, much like myself. My >> Duron 850 is >> the most powerful thing I'm likely to have for years. The point was with the >> way things are going, my processor is likely to be incredibly slow >> compared to >> most, therefore use of extra processing power most likely won't get >> screams of >> protest from most users. > > there are already some graphic setup available in the "options" menu > of the game, but you'll need to reduce the number of cars to get > a good frame rate. > >> >> Anyway contact me with what I can do to help, in the meantime I'll >> mumble in the >> background and complain. I'll try to get some explainations underway of how >> stuff I understand works. >> >> The Secret Hamster >> >> Quoting Bernhard Wymann <be...@bl...>: >> >>> Hi Hamster >>> >>>> Is there a 'to do' list? I was maybe looking at trying to improve >>>> the sound, at >>>> least under Linux. >>> >>> >>> Official TODO list -> no. >>> >>> There is also much improved sound in the CVS:-) You hear all cars, >>> doppler-effect, better samples, ... nice. >>> >>>> No I don't know if you've looked at this and thought it takes too >>>> much cpu, but >>>> I'm thinking my duron 850/geforce4 can do 15 cars at 30fps at the standard >>>> settings, which I find fairly smooth. Therefore I don't think it >>>> unreasonable >>>> to require more of the cpu, especailly as the slowest celeron and >>>> intel are now >>>> 2800Mhz I'm not sure of the AMD chips, but I know they are as quick. >>> >>> >>> The slowest you can buy... not the slowest which still run. I work quite >>> often still with 400 and 550MHz:-) >>> >>>> other stuff I'd like to see done, are >>>> wear and tear on the cars. e.g engines (due to high revving), >>>> gearbox etc. It >>> >>> >>> Tire wear and temperature is in simuv3, but I do not like it:-( It >>> depends on your view, but because I like more the idea of robots racing >>> I think the environment should stay simple (already now it will be hard >>> to find people for a competition I guess...). >>> >>> Even for most people I think it is more fun if it is not too realistic, >>> look at the success of e. g. NFS. >>> >>>> slightly annoys me that reckless bots can pretty mush get away >>>> with bashing >>>> into everything then getting repaired as if nothing has happened >>> >>> >>> Reckless bots? >>> >>>> also certain damage should be irrepairable, or take a long time to repair. >>>> Therefore you can change the gearbox but it'll take you 20mins. >>>> This could lend >>>> to some real endurance races (I'd like to setup a GT Endurance >>>> league in the >>>> end) >>> >>> >>> I disagree here as well... but that is my perspective, I think I'm quite >>> alone with that, so just ignore me. >>> >>>> Track/car designer (maybe something I could look to work on as a >>>> supporting >>> >>> >>> That would be cool:-) >>> >>>> role, PS I am an incredibly novice programmer, perl is more my thing) >>>> rules implemented, and teams (I believe this is already being looked at) >>> >>> >>> In pro mode some rules are already implemented (e.g. pit speeding). >>> >>> bye, Bernhard. >>> >>> >>> -- visit my homepage http://www.berniw.org >>> coming soon: The TORCS Racing Board, http://www.berniw.org/trb >>> > > Eric. > -- > =+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+= > TORCS - http://torcs.org > The Open Racing Car Simulator > AKA The Other Release Coming Soon (Skin'r) > How soon is soon ? (RaceBlizter) > =+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+= |
|
From: Bernhard W. <be...@bl...> - 2004-08-11 20:07:05
|
Hi Hamster > How do I use simv3 rather than simv2 trying to find where I saw it > mentioned The CVS is already configured, but you can change it in /usr/local/share/games/torcs/config (or wherever it is...) in the file raceengine.xml, simply change "simuv2" to "simuv3" or the other way round, depending on what you prefer. > before it but not finding it > > I'll try the cvs version too sounds good. > > Okay I tried not to mention the reckeless bot, but you forced it out of > me. It's > the Inferno bot. It's got some good moves, but overtaking seems to involve > nudging other players. Also it drives so far past the edge it's > constantly off > the circuit and then just comes back without any view to other cars. Okay I > admit it's hard to get this sort of thing right, but in the end I don't do > practise runs against the opposition because of this. > > As regards complication, I understand the complexity issue, humans are > clever Are they? Hmm, must have missed something... but that is another discussion... > creatures, learning stuff quickly. Robots have to be programmed for each > of the I disagree. Look at real drivers, overtaking is even for humans a real issue. Without having statistics handy, just with the feeling from watching TV I would claim that around 50% of the overtaking on not oval tracks (real, not lapping) ends in damages, collisions or spinning cars... so humans perform miserable as well. > situations, hard enough keeping on the circuit let alone all the rest of > it. > However, look at the setup of the car, springs, bump, rebound, bellcranks, > suspension course, toe in, etcs. Is that not complicating things, > considering > it's supposed to be a 'simple' robot racing game. > Not trying to be arguementative, but the setup of the car comes alot > from the > driver and the drivers style, you can't drive in place of your bot to > set up > the car as you'll have your own style. That is absolutely true IMHO, it depends very on the driving style what braking balance you need, etc. But these settings help to adopt the robot driving style to the car. I'm just against the introduction of "whatever wonderful real features" which just waste cpu cycles and hold potential racers away because the stuff is too complex. I think already it is not very easy to get started... the reason why I wrote the tutorial. > I really do complement you on the program, it's something I've thought > of doing > for years, but neither the time or the expertise. Again I know I'm > really new > to this, but I see it with fresh eyes and desperately want to help. I'm > quite > happy to work along with the developers to get some more documentation out. I have already a little setup section in the robot tutorial, have you seen this? > The final point, okay you use old hardware, much like myself. My Duron > 850 is > the most powerful thing I'm likely to have for years. The point was with > the > way things are going, my processor is likely to be incredibly slow > compared to > most, therefore use of extra processing power most likely won't get > screams of > protest from most users. Hmm, its hard to say... I wonder sometimes myself when people come up with R128 and Matrox G450 problems, but I think there are more "old" machines out there than one might believe, not all people want or can spend the money to exchange their hardware every 2 to 4 years. E.g. my girlfriends box is a PII-266, 192MB RAM, voodoo banshee 16MB... too slow for TORCS, but my K6-III 400 is still alive. Of course there is also the possibility to allow the user/system to switch off some expensive features. > Anyway contact me with what I can do to help, in the meantime I'll > mumble in the > background and complain. I'll try to get some explainations underway of how > stuff I understand works. Do not do that:-) The cool thing on such projects is that the source is open... so if you think feature X is missing, simply ask loud if anybody is already working on it, and if not -> start, if yes -> synchronize and help or look for another feature you want. The most important thing is that you have a good time doing it:-) There is actually a lot to do, if you are in the mood: - Configuration screen to enable/disable telemetry, write telemetry to $HOME/.torcs - Write a complete track building tutorial or better, a track editor. - Car building tutorial. - Setup guide. - New Tracks, Cars, Robots. - What about a telemetry viewer in TORCS? - Setup editor anyone (here we would need to agree about the directory structure first). - Automatic setup. - SDL port. - Cygwin/MinGW port. - .... I think I could tell you 100 more... Btw. what the TODO list concerns, there are some informations: - http://sourceforge.net/pm/?group_id=3777 - http://sourceforge.net/developer/diary.php?diary_user=318462 - the FAQ (look at berniw.org or the CVS for the newest one). - search the mailing list archives for "task", "future". bye, have fun, Bernhard. -- visit my homepage http://www.berniw.org coming soon: The TORCS Racing Board, http://www.berniw.org/trb |
|
From: Eric E. <eri...@fr...> - 2004-08-11 19:54:17
|
sec...@fu... wrote: > How do I use simv3 rather than simv2 trying to find where I saw it > mentioned > before it but not finding it It's located in the /usr/local/share/game/torcs/config/raceengine.xml > > I'll try the cvs version too sounds good. > > Okay I tried not to mention the reckeless bot, but you forced it out of > me. It's > the Inferno bot. It's got some good moves, but overtaking seems to involve > nudging other players. Also it drives so far past the edge it's > constantly off > the circuit and then just comes back without any view to other cars. Okay I > admit it's hard to get this sort of thing right, but in the end I don't do > practise runs against the opposition because of this. It's a question of tuning, I did not have the patience to tune all the robots on all the tracks... It's difficult to do a fast robot not aggressive some time ;-) you should note that it's never deliberate, because it's always causing a lot of dammages. > > As regards complication, I understand the complexity issue, humans are > clever > creatures, learning stuff quickly. Robots have to be programmed for each > of the > situations, hard enough keeping on the circuit let alone all the rest of > it. > However, look at the setup of the car, springs, bump, rebound, bellcranks, > suspension course, toe in, etcs. Is that not complicating things, > considering > it's supposed to be a 'simple' robot racing game. yes, but using other bots car setup can be a good starting point. > Not trying to be arguementative, but the setup of the car comes alot > from the > driver and the drivers style, you can't drive in place of your bot to > set up > the car as you'll have your own style. > I really do complement you on the program, it's something I've thought > of doing > for years, but neither the time or the expertise. Again I know I'm > really new > to this, but I see it with fresh eyes and desperately want to help. I'm > quite > happy to work along with the developers to get some more documentation out. > > The final point, okay you use old hardware, much like myself. My Duron > 850 is > the most powerful thing I'm likely to have for years. The point was with > the > way things are going, my processor is likely to be incredibly slow > compared to > most, therefore use of extra processing power most likely won't get > screams of > protest from most users. there are already some graphic setup available in the "options" menu of the game, but you'll need to reduce the number of cars to get a good frame rate. > > Anyway contact me with what I can do to help, in the meantime I'll > mumble in the > background and complain. I'll try to get some explainations underway of how > stuff I understand works. > > The Secret Hamster > > Quoting Bernhard Wymann <be...@bl...>: > >> Hi Hamster >> >>> Is there a 'to do' list? I was maybe looking at trying to improve the >>> sound, at >>> least under Linux. >> >> >> Official TODO list -> no. >> >> There is also much improved sound in the CVS:-) You hear all cars, >> doppler-effect, better samples, ... nice. >> >>> No I don't know if you've looked at this and thought it takes too >>> much cpu, but >>> I'm thinking my duron 850/geforce4 can do 15 cars at 30fps at the >>> standard >>> settings, which I find fairly smooth. Therefore I don't think it >>> unreasonable >>> to require more of the cpu, especailly as the slowest celeron and >>> intel are now >>> 2800Mhz I'm not sure of the AMD chips, but I know they are as quick. >> >> >> The slowest you can buy... not the slowest which still run. I work quite >> often still with 400 and 550MHz:-) >> >>> other stuff I'd like to see done, are >>> wear and tear on the cars. e.g engines (due to high revving), gearbox >>> etc. It >> >> >> Tire wear and temperature is in simuv3, but I do not like it:-( It >> depends on your view, but because I like more the idea of robots racing >> I think the environment should stay simple (already now it will be hard >> to find people for a competition I guess...). >> >> Even for most people I think it is more fun if it is not too realistic, >> look at the success of e. g. NFS. >> >>> slightly annoys me that reckless bots can pretty mush get away with >>> bashing >>> into everything then getting repaired as if nothing has happened >> >> >> Reckless bots? >> >>> also certain damage should be irrepairable, or take a long time to >>> repair. >>> Therefore you can change the gearbox but it'll take you 20mins. This >>> could lend >>> to some real endurance races (I'd like to setup a GT Endurance league >>> in the >>> end) >> >> >> I disagree here as well... but that is my perspective, I think I'm quite >> alone with that, so just ignore me. >> >>> Track/car designer (maybe something I could look to work on as a >>> supporting >> >> >> That would be cool:-) >> >>> role, PS I am an incredibly novice programmer, perl is more my thing) >>> rules implemented, and teams (I believe this is already being looked at) >> >> >> In pro mode some rules are already implemented (e.g. pit speeding). >> >> bye, Bernhard. >> >> >> -- >> >> visit my homepage http://www.berniw.org >> coming soon: The TORCS Racing Board, http://www.berniw.org/trb >> Eric. -- =+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+= TORCS - http://torcs.org The Open Racing Car Simulator AKA The Other Release Coming Soon (Skin'r) How soon is soon ? (RaceBlizter) =+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+= |
|
From: <sec...@fu...> - 2004-08-11 19:11:50
|
How do I use simv3 rather than simv2 trying to find where I saw it mentioned before it but not finding it I'll try the cvs version too sounds good. Okay I tried not to mention the reckeless bot, but you forced it out of me. It's the Inferno bot. It's got some good moves, but overtaking seems to involve nudging other players. Also it drives so far past the edge it's constantly off the circuit and then just comes back without any view to other cars. Okay I admit it's hard to get this sort of thing right, but in the end I don't do practise runs against the opposition because of this. As regards complication, I understand the complexity issue, humans are clever creatures, learning stuff quickly. Robots have to be programmed for each of the situations, hard enough keeping on the circuit let alone all the rest of it. However, look at the setup of the car, springs, bump, rebound, bellcranks, suspension course, toe in, etcs. Is that not complicating things, considering it's supposed to be a 'simple' robot racing game. Not trying to be arguementative, but the setup of the car comes alot from the driver and the drivers style, you can't drive in place of your bot to set up the car as you'll have your own style. I really do complement you on the program, it's something I've thought of doing for years, but neither the time or the expertise. Again I know I'm really new to this, but I see it with fresh eyes and desperately want to help. I'm quite happy to work along with the developers to get some more documentation out. The final point, okay you use old hardware, much like myself. My Duron 850 is the most powerful thing I'm likely to have for years. The point was with the way things are going, my processor is likely to be incredibly slow compared to most, therefore use of extra processing power most likely won't get screams of protest from most users. Anyway contact me with what I can do to help, in the meantime I'll mumble in the background and complain. I'll try to get some explainations underway of how stuff I understand works. The Secret Hamster Quoting Bernhard Wymann <be...@bl...>: > Hi Hamster > >> Is there a 'to do' list? I was maybe looking at trying to improve >> the sound, at >> least under Linux. > > Official TODO list -> no. > > There is also much improved sound in the CVS:-) You hear all cars, > doppler-effect, better samples, ... nice. > >> No I don't know if you've looked at this and thought it takes too >> much cpu, but >> I'm thinking my duron 850/geforce4 can do 15 cars at 30fps at the standard >> settings, which I find fairly smooth. Therefore I don't think it >> unreasonable >> to require more of the cpu, especailly as the slowest celeron and >> intel are now >> 2800Mhz I'm not sure of the AMD chips, but I know they are as quick. > > The slowest you can buy... not the slowest which still run. I work quite > often still with 400 and 550MHz:-) > >> other stuff I'd like to see done, are >> wear and tear on the cars. e.g engines (due to high revving), >> gearbox etc. It > > Tire wear and temperature is in simuv3, but I do not like it:-( It > depends on your view, but because I like more the idea of robots racing > I think the environment should stay simple (already now it will be hard > to find people for a competition I guess...). > > Even for most people I think it is more fun if it is not too realistic, > look at the success of e. g. NFS. > >> slightly annoys me that reckless bots can pretty mush get away with bashing >> into everything then getting repaired as if nothing has happened > > Reckless bots? > >> also certain damage should be irrepairable, or take a long time to repair. >> Therefore you can change the gearbox but it'll take you 20mins. This >> could lend >> to some real endurance races (I'd like to setup a GT Endurance league in the >> end) > > I disagree here as well... but that is my perspective, I think I'm quite > alone with that, so just ignore me. > >> Track/car designer (maybe something I could look to work on as a supporting > > That would be cool:-) > >> role, PS I am an incredibly novice programmer, perl is more my thing) >> rules implemented, and teams (I believe this is already being looked at) > > In pro mode some rules are already implemented (e.g. pit speeding). > > bye, Bernhard. > > > -- > > visit my homepage http://www.berniw.org > coming soon: The TORCS Racing Board, http://www.berniw.org/trb > > > > ------------------------------------------------------- > SF.Net email is sponsored by Shop4tech.com-Lowest price on Blank Media > 100pk Sonic DVD-R 4x for only $29 -100pk Sonic DVD+R for only $33 > Save 50% off Retail on Ink & Toner - Free Shipping and Free Gift. > http://www.shop4tech.com/z/Inkjet_Cartridges/9_108_r285 > _______________________________________________ > Torcs-users mailing list > Tor...@li... > https://lists.sourceforge.net/lists/listinfo/torcs-users |
|
From: Eric E. <eri...@fr...> - 2004-08-11 16:39:24
|
Bernhard Wymann wrote: > Hi Hamster > > The bellcrank is kind of a rotating lever to multiply or divide the > suspension course and force (law of the lever). here is a link to an image: > > http://www.mulsannescorner.com/Ferrari-MF18.jpg > > bye, Bernhard. > and the value is the multiplying coeff, e.g a bellcrank value of 2.0 will double the spring course vs the wheel course. Eric. -- =+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+= TORCS - http://torcs.org The Open Racing Car Simulator AKA The Other Release Coming Soon (Skin'r) How soon is soon ? (RaceBlizter) =+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+= |
|
From: Bernhard W. <be...@bl...> - 2004-08-10 21:00:53
|
Hi Hamster The bellcrank is kind of a rotating lever to multiply or divide the suspension course and force (law of the lever). here is a link to an image: http://www.mulsannescorner.com/Ferrari-MF18.jpg bye, Bernhard. -- visit my homepage http://www.berniw.org coming soon: The TORCS Racing Board, http://www.berniw.org/trb |
|
From: Bernhard W. <be...@bl...> - 2004-08-10 20:40:43
|
Hi Hamster > Is there a 'to do' list? I was maybe looking at trying to improve the sound, at > least under Linux. Official TODO list -> no. There is also much improved sound in the CVS:-) You hear all cars, doppler-effect, better samples, ... nice. > No I don't know if you've looked at this and thought it takes too much cpu, but > I'm thinking my duron 850/geforce4 can do 15 cars at 30fps at the standard > settings, which I find fairly smooth. Therefore I don't think it unreasonable > to require more of the cpu, especailly as the slowest celeron and intel are now > 2800Mhz I'm not sure of the AMD chips, but I know they are as quick. The slowest you can buy... not the slowest which still run. I work quite often still with 400 and 550MHz:-) > other stuff I'd like to see done, are > wear and tear on the cars. e.g engines (due to high revving), gearbox etc. It Tire wear and temperature is in simuv3, but I do not like it:-( It depends on your view, but because I like more the idea of robots racing I think the environment should stay simple (already now it will be hard to find people for a competition I guess...). Even for most people I think it is more fun if it is not too realistic, look at the success of e. g. NFS. > slightly annoys me that reckless bots can pretty mush get away with bashing > into everything then getting repaired as if nothing has happened Reckless bots? > also certain damage should be irrepairable, or take a long time to repair. > Therefore you can change the gearbox but it'll take you 20mins. This could lend > to some real endurance races (I'd like to setup a GT Endurance league in the > end) I disagree here as well... but that is my perspective, I think I'm quite alone with that, so just ignore me. > Track/car designer (maybe something I could look to work on as a supporting That would be cool:-) > role, PS I am an incredibly novice programmer, perl is more my thing) > rules implemented, and teams (I believe this is already being looked at) In pro mode some rules are already implemented (e.g. pit speeding). bye, Bernhard. -- visit my homepage http://www.berniw.org coming soon: The TORCS Racing Board, http://www.berniw.org/trb |
|
From: <sec...@fu...> - 2004-08-10 19:31:11
|
Is there a 'to do' list? I was maybe looking at trying to improve the sound, at least under Linux. Same sound for each car, as if in car. Need other sounds like bumping noises and stuff. My idea on the sound todo was Sound depending on engine, 4,6,8,10,12 cylinders Sound depending on throttle AND rpm (listen to your own car, 3000 idle sounds very different to 3000 at full throttle) Filters for exhaust (another way to make each car different) Mix in close cars Sound from viewing point (i.e further away = quieter) Doppler effects of approach No I don't know if you've looked at this and thought it takes too much cpu, but I'm thinking my duron 850/geforce4 can do 15 cars at 30fps at the standard settings, which I find fairly smooth. Therefore I don't think it unreasonable to require more of the cpu, especailly as the slowest celeron and intel are now 2800Mhz I'm not sure of the AMD chips, but I know they are as quick. other stuff I'd like to see done, are wear and tear on the cars. e.g engines (due to high revving), gearbox etc. It slightly annoys me that reckless bots can pretty mush get away with bashing into everything then getting repaired as if nothing has happened also certain damage should be irrepairable, or take a long time to repair. Therefore you can change the gearbox but it'll take you 20mins. This could lend to some real endurance races (I'd like to setup a GT Endurance league in the end) Track/car designer (maybe something I could look to work on as a supporting role, PS I am an incredibly novice programmer, perl is more my thing) rules implemented, and teams (I believe this is already being looked at) The Secret Hamster |
|
From: <sec...@fu...> - 2004-08-10 19:19:18
|
Okay so now I'm having some fun, well almost Okay what are the various settings on the car setups for gearing, I get totally wing, I get (PS no settings to front aerofoil on the GT cars?) BRAKES front rear partition, I get (balance front back of braking power) max pressure, I get (severity of brakes, too high means hard to get right braking, too little and can't brake enough) ANTI ROLL BAR spring, I get (higher strength transfers more weight to other side) suspension course, don't really get (wobble in wheels?) bellcrank, it's some other form of spring I think no idea what it does, Google has let me down. Find lots of mentions but not what it is for WHEELS ride height, I get (height of suspension, reports inversely on telemetry. lower the better for increase aerodynamic effects) toe, I get (degree in or out that the wheel points, helps with responsiveness) camber, I get (the tilt of the wheel vertically, helps with maitaining a good contact patch) SUSPENSION spring, I get (harder spring less give, more responsiveness less mechanical grip) suspension course, err that thing again bellcrank, wtf is a bellcrank please tell me, even the mechanical engineers at work don't know what one is, ahhhhhhh packers, I get (stop you bottoming out on the long fast straights, PS seem to have no effect) slow bump, err isn't this a spring? so whats the spring? fast bump, spring that takes the impact on hitting curbs and bumpy surfaces, err so what's the other spring for slow rebound, often known as dampers but not (affects the feel of the corner, can help stop car lifting up on opposite corner) fast rebound, dampner thingy, helps keep wheel on the road on bumps and kerbs etc. stops the wheel flying up under the action of the fast bump If I could somehow get an interactive meet online with one of the authors of TORCS I can get get some notes down. Hopefully then I might be able to write some meaningful documentation for you guys. The Secret Hamster Quoting Bernhard Wymann <be...@bl...>: > Hi Hamster > > I have it now working here, it took just one minute. Instructions: > (Precodition: Intallation according to the robot tutorial -> write > permissions and environment variables) > - cd into src/libs/robottools > - copy the attached rttelem.cpp into the code base > - make > - make install (I guess you did not do that before, otherwise you would > have got a segfault) > - now do a practise session with inferno 1. > - after some laps it will spit out telemetry (you need write permission > in usr/local/share/games/torcs). > > I think I know now why Eric removed that for the release, the problem is > that we need to write the telemetry data to ~/.torcs, it is not > acceptable for distros to write into /usr for users. > >> IMHO if you don't have telemetry you can't set up a car, it's also >> pretty hard >> to tell how your bot is reacting too. > > It depends, you can get a lot with simpe visual inspection, slowmotion > is also very nice tool for analysis. > >> This may mean I actually get my hands dirty and try to fix this, but >> don't hold > > Now its fixed, I will have a look at it to write into the ~/.torcs > directory, but you can work now. Have fun, bye > > Bernhard. > > -- > > visit my homepage http://www.berniw.org > coming soon: The TORCS Racing Board, http://www.berniw.org/trb |
|
From: Eugen T. <eug...@un...> - 2004-08-08 14:54:59
|
Hi Bernard, I updated my graphics card (ATI Radeon 9600TX) driver to Catalyst 4.6 and tried to make some screen shots in the game. When my desktop is configured with 32 bpp and game is running in full screen, i get a blue picture insted of a screen shot. So i still have the same problem. bye, Eugen Bernhard Wymann wrote: > Hi Eugen > >>>> The in-game function for making screen shots by pressing F12 doesn't >>>> work. I use torcs 1.2.2 in Windows XP. I only get a small blue png >>>> file with the right resolution. >>>> In the old 1.2.1 release it was possible to make screen shots only >>>> in 16 >>> >>> >>> >>> >>> Interesting, so you are sure that 24 and 32 bpp did not work? Was >>> your screen at 16bpp? >> >> >> >> Here are some detailed results: >> Desktop bpp torcs bpp Fullscreen/Windowed Screenshot-Function >> 16 32 FS OK >> 32 24,32 FS blue picture >> 32 16,24,32 W OK >> >> Other modes are not playable here. > > > Hmm, I could not reproduce this problem, in all modes where the game is > working screenshots work as well (Windows 2000, GeForce3, 6177 Driver). > What graphic card and driver are you using? > > bye, Bernhard. > |
|
From: Bernhard W. <be...@bl...> - 2004-08-04 09:48:31
|
Hi Hamster I have it now working here, it took just one minute. Instructions: (Precodition: Intallation according to the robot tutorial -> write permissions and environment variables) - cd into src/libs/robottools - copy the attached rttelem.cpp into the code base - make - make install (I guess you did not do that before, otherwise you would have got a segfault) - now do a practise session with inferno 1. - after some laps it will spit out telemetry (you need write permission in usr/local/share/games/torcs). I think I know now why Eric removed that for the release, the problem is that we need to write the telemetry data to ~/.torcs, it is not acceptable for distros to write into /usr for users. > IMHO if you don't have telemetry you can't set up a car, it's also pretty hard > to tell how your bot is reacting too. It depends, you can get a lot with simpe visual inspection, slowmotion is also very nice tool for analysis. > This may mean I actually get my hands dirty and try to fix this, but > don't hold Now its fixed, I will have a look at it to write into the ~/.torcs directory, but you can work now. Have fun, bye Bernhard. -- visit my homepage http://www.berniw.org coming soon: The TORCS Racing Board, http://www.berniw.org/trb |
|
From: <sec...@fu...> - 2004-08-04 08:51:09
|
Whoops sent this to Berni only, apologies to Berni who will receive this twice Quoting Bernhard Wymann <be...@bl...>: > Hi Hamster > >> Any of you linux users able to get telemetry to work? >> >> I have copied and enabled the telemetry script from the inferno >> robot and also >> seen the same text used elsewhere (tita is one I believe) >> >> however tracking the functions through (seems very confusing way of >> doing it to >> me, but I'm no programmer) they seem to be commented out in the >> robottools.cpp. >> So are they working or not? > > You mean src/libs/robottools/rttelem.cpp, do you? It seems like they > have been disabled by Eric (I did not hear from him for quite some time, > so I guess he is enjoying his holidays:-)). > > I don't know what the reason was, but usually there is good reason for > such a change. It is possible that the telemetry code stopped working > after the introduction of the raceengineclient, so it was simply > disabled till an new implementation is in place (just guessing). > > You could get the revision 1.6 of rttelem.cpp > (http://cvs.sourceforge.net/viewcvs.py/torcs/torcs/torcs/src/libs/robottools/rttelem.cpp?rev=1.7&view=log) > and recompile TORCS with it, perhaps you are lucky and it works (but I > guess not, what was the reason to disable it then...). > > bye, Bernhard. > > > -- > > visit my homepage http://www.berniw.org > coming soon: The TORCS Racing Board, http://www.berniw.org/trb > > > > ------------------------------------------------------- > This SF.Net email is sponsored by OSTG. Have you noticed the changes on > Linux.com, ITManagersJournal and NewsForge in the past few weeks? Now, > one more big change to announce. We are now OSTG- Open Source Technology > Group. Come see the changes on the new OSTG site. www.ostg.com > _______________________________________________ > Torcs-users mailing list > Tor...@li... > https://lists.sourceforge.net/lists/listinfo/torcs-users Okay I downloaded and compiled the version 1.6 of rttelem.cpp, still no joy. nothing seems to happen. It runs fine, just no files I presume you need to have the /usr/local/share/games/torcs/telemetry directory rw by the user running the program, however nothing appears in there. If anyone did have this running on previous versions, can they post here how it used to work before it broke. I'm very interested in getting this fixed. IMHO if you don't have telemetry you can't set up a car, it's also pretty hard to tell how your bot is reacting too. This may mean I actually get my hands dirty and try to fix this, but don't hold your breath while I do it. The Secret Hamster ----- End forwarded message ----- |
|
From: Bernhard W. <be...@bl...> - 2004-08-02 11:48:33
|
Hi Hamster > Any of you linux users able to get telemetry to work? > > I have copied and enabled the telemetry script from the inferno robot and also > seen the same text used elsewhere (tita is one I believe) > > however tracking the functions through (seems very confusing way of doing it to > me, but I'm no programmer) they seem to be commented out in the robottools.cpp. > So are they working or not? You mean src/libs/robottools/rttelem.cpp, do you? It seems like they have been disabled by Eric (I did not hear from him for quite some time, so I guess he is enjoying his holidays:-)). I don't know what the reason was, but usually there is good reason for such a change. It is possible that the telemetry code stopped working after the introduction of the raceengineclient, so it was simply disabled till an new implementation is in place (just guessing). You could get the revision 1.6 of rttelem.cpp (http://cvs.sourceforge.net/viewcvs.py/torcs/torcs/torcs/src/libs/robottools/rttelem.cpp?rev=1.7&view=log) and recompile TORCS with it, perhaps you are lucky and it works (but I guess not, what was the reason to disable it then...). bye, Bernhard. -- visit my homepage http://www.berniw.org coming soon: The TORCS Racing Board, http://www.berniw.org/trb |
|
From: <sec...@fu...> - 2004-08-02 11:04:08
|
Any of you linux users able to get telemetry to work? I have copied and enabled the telemetry script from the inferno robot and also seen the same text used elsewhere (tita is one I believe) however tracking the functions through (seems very confusing way of doing it to me, but I'm no programmer) they seem to be commented out in the robottools.cpp. So are they working or not? The Secret Hamster |
|
From: Bernhard W. <be...@bl...> - 2004-08-02 09:35:37
|
Hi Eugen >>> The in-game function for making screen shots by pressing F12 doesn't >>> work. I use torcs 1.2.2 in Windows XP. I only get a small blue png >>> file with the right resolution. >>> In the old 1.2.1 release it was possible to make screen shots only in 16 >> >> >> >> Interesting, so you are sure that 24 and 32 bpp did not work? Was your >> screen at 16bpp? > > > Here are some detailed results: > Desktop bpp torcs bpp Fullscreen/Windowed Screenshot-Function > 16 32 FS OK > 32 24,32 FS blue picture > 32 16,24,32 W OK > > Other modes are not playable here. Hmm, I could not reproduce this problem, in all modes where the game is working screenshots work as well (Windows 2000, GeForce3, 6177 Driver). What graphic card and driver are you using? bye, Bernhard. -- visit my homepage http://www.berniw.org coming soon: The TORCS Racing Board, http://www.berniw.org/trb |