Racer v0.4.5beta released

Hi all, I've just uploaded a new version of Racer (it's getting a bit complex at the moment, 3 platforms to support, hm). Get it at: http://www.marketgraph.nl/gallery/racer/ Note that work is slowing down a bit; I've been enjoying the sun and pumping through real-life work. But that's life. ;-) Racer v0.4.5 BETA ----------------- Changes from v0.4.1: - Rendering optimized a little. It now sorts by material, made some 10% difference on my O2. - Texture loading optimized. Multiple occurences are only loaded once. This was previously only used in TrackEd. Should load (much) faster now and use much less memory. If you previously tried to convert tracks like Monza, try again; this version should be much better. - Differential code has been placed in. It's quite beta, and will exhibit quite a few problems while driving (or should I say trying to drive). New features: - Linux version added! Not a fully compilable version probably (mail me your problems; it SHOULD be doable), but perhaps interesting for future serving of races. Strangely enough, the Linux version was up & displaying in 3 days. No sound yet though. Any takers? - Collision detection (car vs. track). A bit crude still; no angular impulses yet (and no angular rotation collision prediction). Thanks to Dave Eberly for his OBB-Triangle detection code. - Power oversteer. Note the Ferrari is quite heavy and requires some pushing to make it spin out from a start. The addition was possible thanks to a friction curve which specifies maximum friction based on the slip velocity. But oh gosh, ofcourse there's too much oversteer! - Lit tracks; some lighting can be added to the tracks (it's on by default). Although subtle, it does add a bit of flavour. To get lighting to work on old exported tracks, use TrackEd, select your track, and click 'Add Lighting'. This will process all track .DOFs and add normals so that lighting works correctly. - Transparent track objects (fences and such). Although the rendering order is not be correct in all cases, alpha rendering is supported. For .BMP texturemaps, any purple pixels are made transparent. The bridge rail in Carrera for example. Note that half-transparent textures (if you use .tga files) will not render in the right order. - ACT Labs wheel control sets were added (thanks to Dedlast) - Logitech wheel axes support. The Logitech wheel used slider axes and these could not be specified. Support was added. Use 'slider0' or 'slider1' as the axis name in the control set files (no Logitech ctrlset exists yet; please send one to me if you get it working) - More axes are supported; the Logitech wheel uses a slider axis, and this was not recognized. Now if someone can do a control set that I can include... (use slider0 or slider1 instead of the regular axes, X/Y/Z/RX/RY/RZ). - Full compilable source available for Win32, SGI and Linux. I might have forgotten things (I'm almost 100% sure), or getting it to work is (very) unclear. Mail me with what you had to do to get it to compile, or what your problems are in getting it up, so I can make it easier for future releases. Warning: it might be a bit of a mess right now; didn't put enough time I guess into testing and searching for omissions. - Force feedback improved. I now have friction working to get rid of the loose feel at low speeds. Look at the msff_ff.ini file in data/ctrlsets for some new parameters. TrackEd changes: - AI waypoint editing now possible (consisting of direction markers and left/right road markers; possible speeds are learned by the robot drivers). - (Track) Camera handling is now a bit where it should be; not too much funny alpha stuff. It's ok to do your own track camera's now, in fact, I'd love to see what you make of it, since I don't really put time into those much. Surfers' Paradise needs them, badly. BONUS BONUS BONUS: - The track 'Surfer's Paradise' has been uploaded. This is originally an SCGT conversion of the Microsoft Cart game. It's not really complete; adding track cams is left as an excersise for the user (and those skyboxes!). If you have created a nice track cam set, please mail the 'special.ini' file to me so I can include it in a forthcoming release. Thanks to Gary Johnson for letting me premier this track. Enjoy! Ruud van Gaal Ruud van Gaal, GPL Rank +53.25 Pencil art : http://www.marketgraph.nl/gallery/ Free car sim : http://www.marketgraph.nl/gallery/racer/

good stuff ... just have to wait til i get home to try it. Shane Ruud van Gaal <[email hidden]> wrote in message news:[email hidden]...

Quoted text (95 lines)

> Hi all, > > I've just uploaded a new version of Racer (it's getting a bit complex > at the moment, 3 platforms to support, hm). > > Get it at: http://www.marketgraph.nl/gallery/racer/ > > Note that work is slowing down a bit; I've been enjoying the sun and > pumping through real-life work. But that's life. ;-) > > Racer v0.4.5 BETA > ----------------- > > Changes from v0.4.1: > - Rendering optimized a little. > It now sorts by material, made some 10% difference on my O2. > - Texture loading optimized. > Multiple occurences are only loaded once. This was previously only > used in TrackEd. Should load (much) faster now and use much less > memory. If you previously tried to convert tracks like Monza, try > again; this version should be much better. > - Differential code has been placed in. It's quite beta, and will > exhibit quite a few problems while driving (or should I say trying to > drive). > > New features: > - Linux version added! Not a fully compilable version probably (mail > me your problems; it SHOULD be doable), but perhaps interesting for > future serving of races. Strangely enough, the Linux version was up & > displaying in 3 days. No sound yet though. Any takers? > - Collision detection (car vs. track). A bit crude still; no angular > impulses yet (and no angular rotation collision prediction). Thanks to > Dave Eberly for his OBB-Triangle detection code. > - Power oversteer. > Note the Ferrari is quite heavy and requires some pushing to make it > spin out from a start. The addition was possible thanks to a friction > curve which specifies maximum friction based on the slip velocity. But > oh gosh, ofcourse there's too much oversteer! > - Lit tracks; some lighting can be added to the tracks (it's on by > default). > Although subtle, it does add a bit of flavour. To get lighting to > work on old exported tracks, use TrackEd, select your track, and click > 'Add Lighting'. This will process all track .DOFs and add normals so > that lighting works correctly. > - Transparent track objects (fences and such). > Although the rendering order is not be correct in all cases, alpha > rendering is supported. For .BMP texturemaps, any purple pixels are > made transparent. The bridge rail in Carrera for example. Note that > half-transparent textures (if you use .tga files) will not render in > the right order. > - ACT Labs wheel control sets were added (thanks to Dedlast) > - Logitech wheel axes support. > The Logitech wheel used slider axes and these could not be > specified. Support was added. Use 'slider0' or 'slider1' as the axis > name in the control set files (no Logitech ctrlset exists yet; please > send one to me if you get it working) > - More axes are supported; the Logitech wheel uses a slider axis, and > this was not recognized. Now if someone can do a control set that I > can include... (use slider0 or slider1 instead of the regular axes, > X/Y/Z/RX/RY/RZ). > - Full compilable source available for Win32, SGI and Linux. > I might have forgotten things (I'm almost 100% sure), or getting it > to work is (very) unclear. Mail me with what you had to do to get it > to compile, or what your problems are in getting it up, so I can make > it easier for future releases. Warning: it might be a bit of a mess > right now; didn't put enough time I guess into testing and searching > for omissions. > - Force feedback improved. > I now have friction working to get rid of the loose feel at low > speeds. Look at the msff_ff.ini file in data/ctrlsets for some new > parameters. > > TrackEd changes: > - AI waypoint editing now possible (consisting of direction markers > and left/right road markers; possible speeds are learned by the robot > drivers). > - (Track) Camera handling is now a bit where it should be; not too > much funny alpha stuff. It's ok to do your own track camera's now, in > fact, I'd love to see what you make of it, since I don't really put > time into those much. Surfers' Paradise needs them, badly. > > BONUS BONUS BONUS: > - The track 'Surfer's Paradise' has been uploaded. This is originally > an SCGT conversion of the Microsoft Cart game. It's not really > complete; adding track cams is left as an excersise for the user (and > those skyboxes!). If you have created a nice track cam set, please > mail the 'special.ini' file to me so I can include it in a forthcoming > release. Thanks to Gary Johnson for letting me premier this track. > > Enjoy! > Ruud van Gaal > > Ruud van Gaal, GPL Rank +53.25 > Pencil art : http://www.marketgraph.nl/gallery/ > Free car sim : http://www.marketgraph.nl/gallery/racer/

I'll do some Beta testing for the Linux version if you want. I just installed Linux and am now only using Windoze for GPL and MS Office. My newsgroups actually work now! Linux hasn't crashed yet! I thought my DSL connection was crappy but it was just Windows. I'm reading up on networking stuff now so maybe I could help with the multiplayer. Kirk House Ruud van Gaal wrote:

Quoted text (96 lines)

> Hi all, > > I've just uploaded a new version of Racer (it's getting a bit complex > at the moment, 3 platforms to support, hm). > > Get it at: http://www.marketgraph.nl/gallery/racer/ > > Note that work is slowing down a bit; I've been enjoying the sun and > pumping through real-life work. But that's life. ;-) > > Racer v0.4.5 BETA > ----------------- > > Changes from v0.4.1: > - Rendering optimized a little. > It now sorts by material, made some 10% difference on my O2. > - Texture loading optimized. > Multiple occurences are only loaded once. This was previously only > used in TrackEd. Should load (much) faster now and use much less > memory. If you previously tried to convert tracks like Monza, try > again; this version should be much better. > - Differential code has been placed in. It's quite beta, and will > exhibit quite a few problems while driving (or should I say trying to > drive). > > New features: > - Linux version added! Not a fully compilable version probably (mail > me your problems; it SHOULD be doable), but perhaps interesting for > future serving of races. Strangely enough, the Linux version was up & > displaying in 3 days. No sound yet though. Any takers? > - Collision detection (car vs. track). A bit crude still; no angular > impulses yet (and no angular rotation collision prediction). Thanks to > Dave Eberly for his OBB-Triangle detection code. > - Power oversteer. > Note the Ferrari is quite heavy and requires some pushing to make it > spin out from a start. The addition was possible thanks to a friction > curve which specifies maximum friction based on the slip velocity. But > oh gosh, ofcourse there's too much oversteer! > - Lit tracks; some lighting can be added to the tracks (it's on by > default). > Although subtle, it does add a bit of flavour. To get lighting to > work on old exported tracks, use TrackEd, select your track, and click > 'Add Lighting'. This will process all track .DOFs and add normals so > that lighting works correctly. > - Transparent track objects (fences and such). > Although the rendering order is not be correct in all cases, alpha > rendering is supported. For .BMP texturemaps, any purple pixels are > made transparent. The bridge rail in Carrera for example. Note that > half-transparent textures (if you use .tga files) will not render in > the right order. > - ACT Labs wheel control sets were added (thanks to Dedlast) > - Logitech wheel axes support. > The Logitech wheel used slider axes and these could not be > specified. Support was added. Use 'slider0' or 'slider1' as the axis > name in the control set files (no Logitech ctrlset exists yet; please > send one to me if you get it working) > - More axes are supported; the Logitech wheel uses a slider axis, and > this was not recognized. Now if someone can do a control set that I > can include... (use slider0 or slider1 instead of the regular axes, > X/Y/Z/RX/RY/RZ). > - Full compilable source available for Win32, SGI and Linux. > I might have forgotten things (I'm almost 100% sure), or getting it > to work is (very) unclear. Mail me with what you had to do to get it > to compile, or what your problems are in getting it up, so I can make > it easier for future releases. Warning: it might be a bit of a mess > right now; didn't put enough time I guess into testing and searching > for omissions. > - Force feedback improved. > I now have friction working to get rid of the loose feel at low > speeds. Look at the msff_ff.ini file in data/ctrlsets for some new > parameters. > > TrackEd changes: > - AI waypoint editing now possible (consisting of direction markers > and left/right road markers; possible speeds are learned by the robot > drivers). > - (Track) Camera handling is now a bit where it should be; not too > much funny alpha stuff. It's ok to do your own track camera's now, in > fact, I'd love to see what you make of it, since I don't really put > time into those much. Surfers' Paradise needs them, badly. > > BONUS BONUS BONUS: > - The track 'Surfer's Paradise' has been uploaded. This is originally > an SCGT conversion of the Microsoft Cart game. It's not really > complete; adding track cams is left as an excersise for the user (and > those skyboxes!). If you have created a nice track cam set, please > mail the 'special.ini' file to me so I can include it in a forthcoming > release. Thanks to Gary Johnson for letting me premier this track. > > Enjoy! > Ruud van Gaal > > Ruud van Gaal, GPL Rank +53.25 > Pencil art : http://www.marketgraph.nl/gallery/ > Free car sim : http://www.marketgraph.nl/gallery/racer/ >

Thrustmaster Nascar Pro Digital Wheel I can't get it to work!! :(

On Mon, 14 May 2001 16:39:49 -0700, Kirk House <[email hidden]> wrote: >I'll do some Beta testing for the Linux version if you want. I'll have to cook together some Linux exes very soon and put those up for download. I don't know however how compatible it will be across versions, so some compilation may be necessary; I'll make that easier in the future too. > I just >installed Linux and am now only using Windoze for GPL and MS Office. My >newsgroups actually work now! Linux hasn't crashed yet! I thought my DSL >connection was crappy but it was just Windows. I'm reading up on >networking stuff now so maybe I could help with the multiplayer. Ideas are always welcome. There's no multiplayer on the list as yet, but as I now have 3 platforms and they're all in the same network, it should be at least possible to test some, ONCE some code gets in. But remember, network play may be the toughest thing to get working yet. I'll probably do some AI first. Hm, might have to make it networked AI, hm. Thanks, Ruud van Gaal, GPL Rank +53.25 Pencil art : http://www.marketgraph.nl/gallery/ Free car sim : http://www.marketgraph.nl/gallery/racer/

On 15 May 2001 04:00:40 GMT, [email hidden] (TRUSRS85) wrote: >Thrustmaster Nascar Pro Digital Wheel > >I can't get it to work!! :( I don't have this wheel, so it's hard; if anyone can do a control set for this one, please send it to me so I can include it for future releases. Note there's a small tutorial on control set files on the site (in the section 'tutorials'; you might figure it out by yourself. Good luck! Ruud van Gaal, GPL Rank +53.25 Pencil art : http://www.marketgraph.nl/gallery/ Free car sim : http://www.marketgraph.nl/gallery/racer/

I've never get your baby to work above 1 (one!) fps on my computer (celeron 366@415 + gforceIImx, win 98). in qlog.txt, i read that it can t import DOF files, and can t load FLG.BMP from the image directory. I've only done a basic install (the package with "carrera" and the ferrari), tweaking or not the resolution with config.exe always give me the same painfully result. someone can help me ? "Ruud van Gaal" <[email hidden]> a �crit dans le message news: [email hidden]...

Quoted text (95 lines)

> Hi all, > > I've just uploaded a new version of Racer (it's getting a bit complex > at the moment, 3 platforms to support, hm). > > Get it at: http://www.marketgraph.nl/gallery/racer/ > > Note that work is slowing down a bit; I've been enjoying the sun and > pumping through real-life work. But that's life. ;-) > > Racer v0.4.5 BETA > ----------------- > > Changes from v0.4.1: > - Rendering optimized a little. > It now sorts by material, made some 10% difference on my O2. > - Texture loading optimized. > Multiple occurences are only loaded once. This was previously only > used in TrackEd. Should load (much) faster now and use much less > memory. If you previously tried to convert tracks like Monza, try > again; this version should be much better. > - Differential code has been placed in. It's quite beta, and will > exhibit quite a few problems while driving (or should I say trying to > drive). > > New features: > - Linux version added! Not a fully compilable version probably (mail > me your problems; it SHOULD be doable), but perhaps interesting for > future serving of races. Strangely enough, the Linux version was up & > displaying in 3 days. No sound yet though. Any takers? > - Collision detection (car vs. track). A bit crude still; no angular > impulses yet (and no angular rotation collision prediction). Thanks to > Dave Eberly for his OBB-Triangle detection code. > - Power oversteer. > Note the Ferrari is quite heavy and requires some pushing to make it > spin out from a start. The addition was possible thanks to a friction > curve which specifies maximum friction based on the slip velocity. But > oh gosh, ofcourse there's too much oversteer! > - Lit tracks; some lighting can be added to the tracks (it's on by > default). > Although subtle, it does add a bit of flavour. To get lighting to > work on old exported tracks, use TrackEd, select your track, and click > 'Add Lighting'. This will process all track .DOFs and add normals so > that lighting works correctly. > - Transparent track objects (fences and such). > Although the rendering order is not be correct in all cases, alpha > rendering is supported. For .BMP texturemaps, any purple pixels are > made transparent. The bridge rail in Carrera for example. Note that > half-transparent textures (if you use .tga files) will not render in > the right order. > - ACT Labs wheel control sets were added (thanks to Dedlast) > - Logitech wheel axes support. > The Logitech wheel used slider axes and these could not be > specified. Support was added. Use 'slider0' or 'slider1' as the axis > name in the control set files (no Logitech ctrlset exists yet; please > send one to me if you get it working) > - More axes are supported; the Logitech wheel uses a slider axis, and > this was not recognized. Now if someone can do a control set that I > can include... (use slider0 or slider1 instead of the regular axes, > X/Y/Z/RX/RY/RZ). > - Full compilable source available for Win32, SGI and Linux. > I might have forgotten things (I'm almost 100% sure), or getting it > to work is (very) unclear. Mail me with what you had to do to get it > to compile, or what your problems are in getting it up, so I can make > it easier for future releases. Warning: it might be a bit of a mess > right now; didn't put enough time I guess into testing and searching > for omissions. > - Force feedback improved. > I now have friction working to get rid of the loose feel at low > speeds. Look at the msff_ff.ini file in data/ctrlsets for some new > parameters. > > TrackEd changes: > - AI waypoint editing now possible (consisting of direction markers > and left/right road markers; possible speeds are learned by the robot > drivers). > - (Track) Camera handling is now a bit where it should be; not too > much funny alpha stuff. It's ok to do your own track camera's now, in > fact, I'd love to see what you make of it, since I don't really put > time into those much. Surfers' Paradise needs them, badly. > > BONUS BONUS BONUS: > - The track 'Surfer's Paradise' has been uploaded. This is originally > an SCGT conversion of the Microsoft Cart game. It's not really > complete; adding track cams is left as an excersise for the user (and > those skyboxes!). If you have created a nice track cam set, please > mail the 'special.ini' file to me so I can include it in a forthcoming > release. Thanks to Gary Johnson for letting me premier this track. > > Enjoy! > Ruud van Gaal > > Ruud van Gaal, GPL Rank +53.25 > Pencil art : http://www.marketgraph.nl/gallery/ > Free car sim : http://www.marketgraph.nl/gallery/racer/

On Tue, 15 May 2001 16:17:35 +0200, "koko" <[email hidden]> wrote:

Quoted text (7 lines)

>I've never get your baby to work above 1 (one!) fps on my computer (celeron >366@415 + gforceIImx, win 98). >in qlog.txt, i read that it can t import DOF files, and can t load FLG.BMP >from the image directory. I've only done a basic install (the package with >"carrera" and the ferrari), tweaking or not the resolution with config.exe >always give me the same painfully result. >someone can help me ?

Yes, two .dof files that it can't import ("Can't import DOF ''") and FLG.BMP. Have that here as well, but the DOFs that it can't find are the skyboxes; if it's annoying you can remove the geometry entries in data/tracks/carrera/geometry.ini (where probably a line "file=..." was changed in "_file=..."). The FLG.BMP files are not missing actually, but they are animations in SCGT, so you have FLG00.BMP-FLG09.BMP. Not too worry, these are replaced by a red texture, so they are easily identifiable onscreen. As for the 1 fps, hm. It does much more on my GForceIIMX, but on Win2K. One possible thing to check might be to hook up a terminal on COM1 and see some more debug output at the start (where it selects a visual/pixelformat). One thing that might be the cause is that window-rendering isn't h/w accelerated. Try running fullscreen. If that doesn't work, try tweaking the #bits. Or set your desktop at 800x600x32 for example. It seems the s/w rendering is used. Hope this helps, Ruud van Gaal, GPL Rank +53.25 Pencil art : http://www.marketgraph.nl/gallery/ Free car sim : http://www.marketgraph.nl/gallery/racer/

This looks awesome! Really great work. I played the Ferrari Arcade simulator yesterday for the first time, and Racer looks a lot like it. Is it supposed to be raining at the Carrera track? There is very little traction. -- Pat Dotson

Thanks for letting us try out your work-in-progress... It's looks good, coming along nicely. When I try it on my Athlon 1.2GHz 266FSB with RadeonDDR it's really fast and looks good, but the physics are having a hard time. The car slides like it's on ice and many times started bouncing (or teetering, see-sawing, vacillating?... hmmm you know what I mean) uncontrollably until it digs itself into the ground and eventually disappears and the program crashes. I know it's beta, I just though I'd let you know... Alain. "Ruud van Gaal" <[email hidden]> wrote in message news:[email hidden]...

Quoted text (10 lines)

> Hi all, > > I've just uploaded a new version of Racer (it's getting a bit complex > at the moment, 3 platforms to support, hm). > > Get it at: http://www.marketgraph.nl/gallery/racer/ > > Note that work is slowing down a bit; I've been enjoying the sun and > pumping through real-life work. But that's life. ;-) >[...]

On Wed, 16 May 2001 12:36:31 -0500, Pat Dotson <[email hidden]> wrote: >This looks awesome! Really great work. I played >the Ferrari Arcade simulator yesterday for the first >time, and Racer looks a lot like it. Thanks. :) It should definitely get environment mapping to get a shine on the car though before I really think it looks good. >Is it supposed to be raining at the Carrera track? There >is very little traction. Ah, that's the excuse I was looking for! Rain! ;-) Actually, the traction is not that bad when just going straight, but the lateral forces get killed pretty quickly. I noticed you can even turn into an eternal spin. Something to do with the brakes, and has the highest priority, except for the source code adjustments (for people trying to compile it) I'm going to do this evening. But you're right, when turning, things are getting way off realistic scales. My brain is trying to crack it this evening again. :) Ruud van Gaal, GPL Rank +53.25 Pencil art : http://www.marketgraph.nl/gallery/ Free car sim : http://www.marketgraph.nl/gallery/racer/

On Wed, 16 May 2001 21:49:07 -0400, "Alain" <[email hidden]> wrote: >Thanks for letting us try out your work-in-progress... >It's looks good, coming along nicely. >When I try it on my Athlon 1.2GHz 266FSB with RadeonDDR it's really fast and >looks good, but the physics are having a hard time. The car slides like >it's on ice and many times started bouncing (or teetering, see-sawing, >vacillating?... The ice sliding is the biggest problem at the moment. Hope to get that out somehow. Difficult sometimes to debug, although I did do a 'Save race in progress' feature yesterday so I can more easily get at situations and look more closely at what the car's doing. > hmmm you know what I mean) uncontrollably until it digs >itself into the ground and eventually disappears and the program crashes. Yes, esp. when hitting the kerbs it can do this. Has to do with the collision detection; this isn't quite right and can get you stuck, not only in the fences, but also in the ground itself. I might have to manually decrease the box size (the black box drawn around the car), so that it doesn't get stuck into the ground that easily. > I >know it's beta, I just though I'd let you know... Thanks anyway; I'll keep working on it. Once that damn tire sliding is out of the way... @#$#$! ;-) Ruud van Gaal, GPL Rank +53.25 Pencil art : http://www.marketgraph.nl/gallery/ Free car sim : http://www.marketgraph.nl/gallery/racer/

Sorry I didn't mean to discourage you... I guess you knew about these issues, I just though maybe it was something that was only happening on my machine because of the clock speed or something. Well, I have to say that I'm envious of what you are working on... that's a great project. I think simulating real physics (well, at least Newtonian physics) on a PC in real time is one the most impressive things for me. I really believe it is the key to an immersive experience on the PC. Good luck, don't give up and have fun! Alain. "Ruud van Gaal" <[email hidden]> wrote in message news:[email hidden]... > [...] > Thanks anyway; I'll keep working on it. Once that damn tire sliding is > out of the way... @#$#$! ;-) > [...]

On Thu, 17 May 2001 13:45:13 -0400, "Alain" <[email hidden]> wrote: >Sorry I didn't mean to discourage you... You didn't discourage me, not to be afraid. :) > I guess you knew about these >issues, I just though maybe it was something that was only happening on my >machine because of the clock speed or something. No, it should run fine on machines >=1.5GHz. ;-) (unlike GPL :( ). Actually, it runs on Onyx supercomputers as well (SGI), and those can have 256 CPUs. I don't know whether it WILL run on those, but actually I don't see why not, but in any case I have a screenshot of an Onyx which have at least 2 CPUs I think. The clock speed should really affect little; I can tell because it's about 5x slower on my SGI O2 than on the PC here. >Well, I have to say that I'm envious of what you are working on... that's a >great project. I think simulating real physics (well, at least Newtonian >physics) on a PC in real time is one the most impressive things for me. I >really believe it is the key to an immersive experience on the PC. I must say that Windows2000 makes you think your computer is slow, while if you program these physics things they actually turn out to be the fastest damn things you ever saw since the 'look at what we can do in 1 frame' attitude on the C64. ;-) >Good luck, don't give up and have fun! Thanks! Ruud van Gaal, GPL Rank +53.25 Pencil art : http://www.marketgraph.nl/gallery/ Free car sim : http://www.marketgraph.nl/gallery/racer/

That's so true... I remember when I programmed in assembly on the C64, Amiga and PC to try and do primitive 3D vector graphics (no textures) and I had to optimize to death to get a decent frame rate. I went on to have a full-time job and life (so little time for play :-( ) and I assumed that todays programmers must be brillant to squeeze such incredible visuals out of todays PCs. Well recently, while I was on holiday, I tried to write a quick C++ D3DX program which had full 3D objects and real-time gravity (objects fell to the ground, but simple stuff)... I finished writing it and thought "I'm going to get seconds per frame instead of FPS with this code..." well the thing ran so fast it was ridiculous! All this unoptimized C++ code with floating point data was flying at an incredible rate... today's CPUs are real beasts! Now's the time to have a full-fledged real car simulator! (sorry this was a little off topic) Alain. "Ruud van Gaal" <[email hidden]> wrote in message news:[email hidden]... > I must say that Windows2000 makes you think your computer is slow, > while if you program these physics things they actually turn out to be > the fastest damn things you ever saw since the 'look at what we can do > in 1 frame' attitude on the C64. ;-)

On Thu, 17 May 2001 16:38:11 -0400, "Alain" <[email hidden]> wrote: >That's so true... I remember when I programmed in assembly on the C64, Amiga >and PC to try and do primitive 3D vector graphics (no textures) and I had to >optimize to death to get a decent frame rate. I went on to have a full-time >job and life (so little time for play :-( ) and I assumed that todays >programmers must be brillant to squeeze such incredible visuals out of >todays PCs. There's still the rule that brilliant programming can lead to even more brilliant visuals though; there's a lot of knowledge behind it all, but for starters, you indeed don't need to do much anymore. That's cool. ... >Now's the time to have a full-fledged real car simulator! Define 'full-fledged' ;-) It can always go higher, better, more. Until we get per-polygon dents (damage), the work is never finished. I'm often looking at races and looking at why the screen is so much more lifelike than on a computer. Perhaps I'll do a human-motion type delay in the track cameras. Also, the track cams should be more intelligent in the future, in that they notice more than 1 car and zoom in/out accordingly. BTW The spinning bug was reasonably squashed out last night. I could sleep a happy man as I found a major bug where longitudinal forces were applied IN THE WRONG DIRECTION where it comes to rotation of the car. Gosh, that it ever worked. ;-) The lateral forces were bigger than the longitudinal forces though, so it turned the car the right way, but braking and throttling in a turn was actually fighting the car rotation the wrong way! Now, though, there are some trickery things left; I will probably chop up a new version tonight, but I stumble upon some new things now. I will get some options in to turn on/off some 'realism' features. If I tuck those in, I get more slippery responses once again. But this may be due to the open differential code; open diffs are bad (for the car handling) I think when the car starts to slide. So it might be on the spot. I don't know, I don't know how an open diff feels (in real life) wrt a locked or limited slip diff. But it sure drives a lot better! :)) Keep an eye open tonight for the new release (v0.4.5), if I get round to it. Will be at most this weekend. As long as the weather stays this shitty. ;-) Ruud van Gaal, GPL Rank +53.25 Pencil art : http://www.marketgraph.nl/gallery/ Free car sim : http://www.marketgraph.nl/gallery/racer/

Wow, great, we'll be sure to check out the latest... Thanks! "Ruud van Gaal" <[email hidden]> wrote in message news:[email hidden]...

Quoted text (23 lines)

> [...] > BTW The spinning bug was reasonably squashed out last night. I could > sleep a happy man as I found a major bug where longitudinal forces > were applied IN THE WRONG DIRECTION where it comes to rotation of the > car. Gosh, that it ever worked. ;-) > The lateral forces were bigger than the longitudinal forces though, so > it turned the car the right way, but braking and throttling in a turn > was actually fighting the car rotation the wrong way! > > Now, though, there are some trickery things left; I will probably chop > up a new version tonight, but I stumble upon some new things now. I > will get some options in to turn on/off some 'realism' features. If I > tuck those in, I get more slippery responses once again. But this may > be due to the open differential code; open diffs are bad (for the car > handling) I think when the car starts to slide. So it might be on the > spot. I don't know, I don't know how an open diff feels (in real life) > wrt a locked or limited slip diff. > But it sure drives a lot better! :)) > > Keep an eye open tonight for the new release (v0.4.5), if I get round > to it. Will be at most this weekend. As long as the weather stays this > shitty. ;-) > [...]

On Fri, 18 May 2001 15:50:44 -0400, "Alain" <[email hidden]> wrote: >Wow, great, we'll be sure to check out the latest... >Thanks! It's there. :) Ruud van Gaal, GPL Rank +53.25 Pencil art : http://www.marketgraph.nl/gallery/ Free car sim : http://www.marketgraph.nl/gallery/racer/