Sorry your browser is not supported!

You are using an outdated browser that does not support modern web technologies, in order to use this site please update to a new browser.

Browsers supported include Chrome, FireFox, Safari, Opera, Internet Explorer 10+ or Microsoft Edge.

Work in Progress / Fiero Racing

Author
Message
Kjelle
21
Years of Service
User Offline
Joined: 29th Sep 2002
Location:
Posted: 11th Nov 2002 10:35
I have worked with car physics, peeking a bit at the masked coder. Not as advanced but still fun to code.
This code is written in DBV113.

My wish is to use QuakeRally BSP's but I cant get some fundamental things to work in DBPro, must wait for next patch.

ftp://anonymous:anonymous@hostname.nu/ftp/MTEC_Productions/SnowFreaks_other/Fiero.zip
The Masked Coder
21
Years of Service
User Offline
Joined: 6th Oct 2002
Location:
Posted: 12th Nov 2002 09:47
Cool, downloading.

Anything, with car and physics in the title grabs my attention, lol.



The Masked Coder
21
Years of Service
User Offline
Joined: 6th Oct 2002
Location:
Posted: 12th Nov 2002 09:50
Ahh, seems to be a broken url, or else i can't ftp into the site to get the file.

Any chance you can upload to a standard web page somewhere ?

Kjelle
21
Years of Service
User Offline
Joined: 29th Sep 2002
Location:
Posted: 12th Nov 2002 10:54
Yes youre right, this is the right path:

ftp://anonymous:anonymous@hostname.nu/MTEC_Productions/SnowFreaks_other/Fiero.zip

Now it should work. I have used Vapour's heightmap to create the world, but it's a bit to 'bumpy' thats why I wanted to use Quakerelly BSP's instead. I think that would be fun to try .

The instructions are in swedish, but I think you get the idea. Speaking of car physics, which game do you think have the most accurate car physics today ?

I think that these games have very good physics:
TOCA-2, Colin 2, INSANE. The INSANE engine is particularly interesting with its damagemodel which is completely sick ! Now there are some tools for the creation of almost any vehicle, I've even seen GNU's in the game

I have looked a bit into creating own vehicles in insane and it looks real cool.

Maybe something to create in DarkBasic ? A general vehicle editor, in which you can put any physic parameters in collision gravity, limbs which can replace and deform easily.

Or maybe that ruins all the fun

Well dont expect much from the fiero game, the code is only made for fun.
The Masked Coder
21
Years of Service
User Offline
Joined: 6th Oct 2002
Location:
Posted: 12th Nov 2002 15:41
Yup, got it downloading now ok, 10mb

On the subject of car physics, to me, the codemasters game INSANE has the best car physics i've seen, and i always like to make the point, that even the mud flaps on the vehicles seem to have theyre own physics, lol.

If you study what they have done, it's truly incredible.

God knows how many degrees of freedom they have. Fully independant suspension, excellent 2d physics, don't know if theyve used pajecka or not, looks damn accurate to me though, Damage model as you say, mesh deformation on the vehicles and drivers, Excellent flight model and rolling and crashing, very realistic, all in all incredible physics, it's aproaching perfection.

Well, my aim is to eventually code, something similar, physics engine wise, but there is an awful long way to go before i reach that standard. As you probably know, car physics coding at first glance, seems quite a simple and basic problem to solve, but the lack of any decent ametuer car games will show you, just how bloody difficult it is to acheive and the lack of any decent resources on the net to help ameteur coders doesnt help much either. Youve got to get your hands dirt with math formula if you wan't to create anything half decent.

If youve checked my web page then youll know the links to marco monsters site and racer, those 2 web sites offer the best car physics resources i've yet found. Well worth checking regularly.

The racer open source car physics engine/project, uses the open dynamics library, by the way, which is a c++ open physics library, as the basis for it's engine. The source does included the games industry standard pajecka algorithm to model tyre forces, and is well worth investigating, if only just to familiarise yourself with the algorithm.

Look forward to trying your demo, and hey, fun is the name of the game

Kjelle
21
Years of Service
User Offline
Joined: 29th Sep 2002
Location:
Posted: 12th Nov 2002 16:20
I have been at Racer's downloaded and tested the engine. Actually it is a company called invictus who made the Insane game, http://www.invictus.hu They also have a new release 'street legal' which is interesting but very bugfilled. You can reconstruct and modify your vehicle in that game, very interesting !
Btw, do your webpage work now, and what is the adress, I have tried your advertised IP number but it does not work anymore it seems ?!
hexGEAR
21
Years of Service
User Offline
Joined: 3rd Nov 2002
Location: Naytonia
Posted: 12th Nov 2002 17:54
Hey masked coder, i kinda need your help, that is if your not that busy and have the time. Later on i'm going to be making a game that has a rocket pack as one of the main items, i was just playing your 3d shoot em up demo and was wondering if i could get the source code for it as well as some explanation on how it works (maybe rem's in the code)?

to live is to suffer, to survive, well, that's to find meaning in the suffering - DMX
The Masked Coder
21
Years of Service
User Offline
Joined: 6th Oct 2002
Location:
Posted: 12th Nov 2002 18:58
Hi Kjelle .

Yeh, your right, invictus did create insane, codemasters published it.

Tried your game, and thought it was a fair attempt at car physics. Wasn't too keen on the engine noise, but the landscape looked cool and nice car model too.

Yeh, the landscape is real bumpy, and a suspension system would make for smoother gameplay.

Made some awsome jumps too, lol, loads of fun.

Hope you continue with car physics, as it's something not many people succsefully pull off. You could build on this code and over time create something quite spectacular.

Yeh, i havn't bothered altering those ip addys advertising my site here, as i only really post on the rgt forums now, my latest demo being there for download.

But heres my website addy anyway, the new one.

http://www.vapournet.com/maskedcoder/index.html

or direct ip

http://66.227.66.133/maskedcoder/index.html




Hello hexgear.

I guess your after my thrust code, for ship flight, instead of a normal aeroplane flight control method.

Well, the source i'm affraid isnt available for download, sorry, but the technique i used, was to calculate a thrust vector, pointing down out of the bottom of the ship, and continually applied motion to the ship, based on a direction of motion vector, with gravity pulling the ship downwards at all times.

These direction vectors, are unit vectors, which are an x,y,z vector with a value of 1 or less. They have no size or position but direction only.

I used a standard surface normal unit vector algorithm to calculate these.

Sorry i cant help further.

Masked

Shadow Robert
21
Years of Service
User Offline
Joined: 22nd Sep 2002
Location: Hertfordshire, England
Posted: 12th Nov 2002 23:19
couldn't tell much due to the terrain, perhaps just giving it a flat surface with matrix and i dunno perhaps a ditch of something somewhere.
Oh and take out that annnoying sound is very offputting

for car physics i'd recommend NOT following the Marco or alike code, as they don't utilise actual Vector code which gives bugs and inherintly makes it pretty useless to make realistic physics.

you should really have 3 Vectors and from them calculate position and rotation based on the cars values...
Gravity, Actual Heading, Set Heading
all of these should be through the Cars CG, so when the wieght shifts the CG does as well and the Vectors move with that.
Add to this Collision (unless detected polygon collision/evasive calculation collision) which should have 2 boxes for the actual car (also 1 per door IF you add or need them) one to cover the body & one to cover the cabin.
Also will need one per wheel which will allow you to calculate Suspension & Body rotation.

Always working from the center of gravity and the calculation of where that is at a given point, the body wieght of the car and CG should be the outwieghting principals (plus traction which isn't easi to calc) will allow you to calculate the Real Vector which the car heads on ... it is also best to think about interpoling this value to smoothness between calculation and loops

I'll see about getting up the whole math behind the calculations which is no easi thing to understand without a good knowlage of Algebraic Math and Physics laws.
Have to convert from my current C++, because i've recently been developing a realistic physics for the Zoids Battle Engine ... i can get the Biped and Quadrapeds to move, but the Slither's are proving troublesome

Anata aru kowagaru no watashi!
The Masked Coder
21
Years of Service
User Offline
Joined: 6th Oct 2002
Location:
Posted: 13th Nov 2002 09:48
LOL.

Of course raven vagetta knows it all.

He realy hasn't the foggiest idea about car physics, and frankly i havn't seen any car physics code from him, although he often claims to have some miraculous code, hidden somewhere, or secretly in development.

Marco Monsters car physics code , of course uses vector maths, and utilises heading, yaw speed, rotation rate, wheel and body slip angles, acceleration, revs ect, the whole gammut. The one thing that isn't used at present, as far as i know from marco, is the pacjka tire slip formulas, which are what most of the Comercial games use. I would reccomend studying his algorithm as an introduction to the Proper maths involved in car physics.
It's not perfect, of course, and you can't expect to create comercial quality physics, without understanding the basics of particle physics, dynamics ect.


You won't find a better tutorial on the net for beginners and it is laid out in a friendly manner and in laymans terms.

To simply dismiss this code, as crap, is ludicrous, and Raven is basically basing his assumptions on my interpretation of marco's code, which he obviously doesnt think is up to much.

I lay a challenege to vegeta, now. Write a better car physics code, in dark basic and post the source here for all to use. If your such an expert, this should be an easy task for you.

But we all know, nothing will transpire. And further damage to his reputaion will ensue.

It's worth mentioning too, that Raven Vageta was banned from the RGT forums, for various reasons, and i see his attitude here still hasnt changed. Still a big head know it all.

I particularly don't like the way, he along with many others, order you, to make changes in your code, to please them. A polite request or sugestion would be met with far more respect and would actually be helpful, instead of insulting the programmer. One of the reasons i don't post here much. RGT in my opinion is a much more user friendly db forum.

But to give him credit, i'm sure he has helped a lot of people, just a shame i hate his attitude.

For anyone wishing to have a look at marco mosnters car physics tutorial.

Here is the link

http://home.wxs.nl/%7Emonstrous/

Kjelle
21
Years of Service
User Offline
Joined: 29th Sep 2002
Location:
Posted: 13th Nov 2002 10:10
I am glad for any input and response, and I dont know about RV's earlier attitude, so I dont want to judge anyone altough you maybe are right

Can you give me the address to the RGT-Forum?
I will definitely check out Marcos, I'm open to all suggestions right now as I am new to the subject.
The Masked Coder
21
Years of Service
User Offline
Joined: 6th Oct 2002
Location:
Posted: 13th Nov 2002 16:36
http://www.realgametools.net/forums/

And is run by Guy Savoie.



Shadow Robert
21
Years of Service
User Offline
Joined: 22nd Sep 2002
Location: Hertfordshire, England
Posted: 13th Nov 2002 21:17
The Vectors in Marcos demo' arn't real vectors and are calculated seperately, its why you can't stop properly with fix code... it is also why you've had so much trouble with adding gravity because rather than using the variables already there you're trying to just patch it in - and also noticed that you're still using Matrix "Get Height" rather than polygon collision.

The vectors for the Real and Set Speed and Direction should be setup for an axis point (CG) which the WHOLE car moves by...
All the programming I have done for this is within someone elses demo as it stands which runs in DarkBasic and i've used David Brabhams ByWire Physics as a Base for all of my calculations.
The whole mess with Marcos code is quite unrealistic, especially as he substitutes alot of code for unknowns or make do values, not to mention the actual math runs in the wrong order

I've not had a chance to download your latest demo Masked, however from what i've heard the inherint problems that had cursed the first few versions are still there ... and if you havn't figured by now a better way than seperating the XZ axis calculation a 2D scale and actually utilising all 3Axis and plotting them based on a single vector than there really isn't alot you can be snapping at me for here.

Its kinda the problem i had (and i believe more the reason why i was banned from RGT) with Simple's tutorials.
People learn what they're taught from day one, so if you start off on the wrong foot to begin with then you're doomed from the start!

It is far better for someone to fumble through something on thier own than to be taught something that will cause them not to achieve what they want to in the end...
Rather than panning what i had to say like that i think you should take some note, just because you disagree is no reason to reply in such a mannor!

You have your way... I have my way - i think sheding someone into a single track of thought on this kinda matter is quite wrong.
And at the end of the day it is better to have 2 seperate ways to think and work from as a base than one.

Anata aru kowagaru no watashi!
The Masked Coder
21
Years of Service
User Offline
Joined: 6th Oct 2002
Location:
Posted: 14th Nov 2002 16:07
Yeh, this is a typical response from you vageta.

Lots of waffle with no substance.

No source code, from you either, no supprise there then, and in fact you admit to using a professionals physics engine, if it's from the david brabham of elite fame i'm thinking of. I did a quick search on the net about this bywire physics engine, but yielded very few relevant links, so i havn't been able to investigate it yet. So infact, you havn't coded car physics by yourself, but have used someone elses algorithm to base yours on. Which i have done also, which just illustrates the need for car physics tutorials on the net and how difficult coding and modelling the physics really is.

Again you embarrass yourslelf, with liitle to no knowledge of what you are talking about.

The use of the term real vectors, is hillarious and i think you are obviously confused somewhere along the line, maybe with polar vectors, interia tensors or cartesian coordinates. ? You talking so much rubbish it's difficult to know where to start.

The vectors in marcos code are calculated seperately. ?
I think what you mean is, that car heading, is calculated using a 2d vector along the x,z plane. Whats wrong with that, it's very common practice to handle heading in this way, and makes perfect sense. A very common technique used in many games and sims for a very long time. ?

Umm, cant stop properly with fixed code.
Well, yes, the physics do break down at very low speeds and stopping, does throw up errors, and this is inherant in the physics code, where numerical instabilities arise at low speeds. The term fixed code, has me a bit mystified and the claim that that is the reason why stopping is in error in the code, again has me a bit baffled.

And why do i have so much trouble calculating gravity. ?
I have no problem with adding gravity to my car physics code, simply calculated with a surface normal, pointing out of the front of the car, and the vertical componants of that normal are scaled by gravity, this causes and downward pull on the car, depending on how much the car is pitched.
There is no problem with gravity in my code. ?, another flase claim by you.

Matrix get height inbstead of polygon collision.
I suppose you mean, why am i using a matrix and not a 3ds or x file landscape. So, what, whats wrong with that. ?
An easily generated landscape with the matrix command, serves the purpose very well, and leaves out a lot of the overheads with an x,3ds object landscape.
By the way, i havn't yest seen decent matrix aligning on a 3ds or x object landscape. Just a point to make.
A matrix does have disadvantages, ie, speed being the main one, but just because i use one does not mean the demo/code is inferior. So i assume, you think that anyone who uses a matrix as a landscape in dark basic, is an inferiror coder to yourself ?

The maths runs in the wrong order. ? What the hell are you talking about. it's marco's algorithm, he can use whatever order he chooses. And in fact it is the standard convention for car physics, so again your wrong in that statement.
Incidentally, i use the same convention.

Unless you can produce a demo or source code to back up your claims, I WILL A CALL YOU A LIAR, FRAUD AND A CHEAT.
I've seen you BULLSHIT for a long time now vegeta and this really just illustrates the point.

I am more than happy with my car physics engine, in it's current state, given the algorithm used. It is in continual development and i never claim that it's perfect. Something you seem to assume it should be. Maybe a better programmer than myself could create even more realistic car physics using the same algorithm.

You seem, to dismiss marcos tutorial and demo's as being poor, and claim that a car physics engine should be created from scratch even though you admit to copying someone elses code.

This is a very difficult task for most people, and marcos site has been a tremendous help to a lot of people in getting them started in coding car physics. It is a basis to start from and build upon.

If you know of another better car physics tutorial on the net, then post the url here for all to see.

I highly reccomend marcos site for his car physics tutorial and code. It isn't perfect, but it's the best youll find on the net, unless you know different.


For one last time vegeta, i'll say, put up or shut.

Show us what you claim to have created.


Note to moderators.

Might be worth locking this thread, if it degenerates to much into a flame war.

Kohai of UWDesign
21
Years of Service
User Offline
Joined: 12th Sep 2002
Location: France
Posted: 14th Nov 2002 21:11
Masked Coder : Thk you ...

[Kohai]
[url]www.underwaredesign.com[url] [url]kohaistyle.com[url]
pIII 733/256 Ram/Geforce 256
rapscaLLion
21
Years of Service
User Offline
Joined: 29th Aug 2002
Location: Canada
Posted: 14th Nov 2002 22:33
...
(no comment)

Alex Wanuch
aka rapscaLLion
Kousen Dev Progress >> Currently Working On Editors
Shadow Robert
21
Years of Service
User Offline
Joined: 22nd Sep 2002
Location: Hertfordshire, England
Posted: 14th Nov 2002 23:21
Isn't much of a flame war to be started here
A Vector is a Bi,Tri,Quad Calculation of a Set point in space at any given time based upon set or variable values.

For a 2D Vector this is a single Figure which represents speed and direction which you can use to plot against 2 planes of veiw/axis

Calculating these planes seperatly is technically not a Vector as the actual values for these numbers at no point pertain to each other as a single value.

The Marco code and all spin off i've seen keep the bug of stopping simply because the low speed can't be calculated properly at low speed wihout greater and greater float acuracy, whereas a vector doesn't require this acuracy because the speed is a constant within it for ALL points, not each point

David Brabam's ByWire isn't a simple algorithm and isn't outlined like a tutorial or engine. It is more an explaination of the simplicities of Vectorised Math when dealing with Free Flight modes which he used for Elite and Virus.
The fact that you're unaware that there should be an order to calculating the aspects for the Physics is kinda of point enough here would you not say?

When you get into Vectorised Algorithm, Logrithm and Algebraic Mathematics, then some good A Level math does help out

Rather than looking at numerous engines that are trying to fake these actions through coding perhaps you should check more within some physics and mathematical books.
The Hungerford & Mercer as well as Hewitt ranges are particularly good and easi to understand

Rather than just being abusive about what has been said, please look deeper into all of this. I don't think you should spout of your mouth stopping people from reading what I have to say ... just because i choose not to show any code what so ever to everybody doesn't mean it isn't there with other people.
And it is upto them not me if it is to be shown

Anata aru kowagaru no watashi!
hexGEAR
21
Years of Service
User Offline
Joined: 3rd Nov 2002
Location: Naytonia
Posted: 14th Nov 2002 23:22
vegeta is gonna be vegeta until he dies (go watch dragonballZ), either except it or don't. His simillar behaviour here as was observed by you guys from his former forum should have been a sign to you that he was not gonna change no matter what. The way i see it, if he bothers you, ignore him, simple. Yeah he may seem pushy at times but the advice he gives helps in one way or the other. When he told me, and i quote "your game is painfully slow", and gave me his advice ("orders" if you would like to call it that), i've taken a better route to making my game. dunno about you guys but to me, he's alright.

to live is to suffer, to survive, well, that's to find meaning in the suffering - DMX
The Masked Coder
21
Years of Service
User Offline
Joined: 6th Oct 2002
Location:
Posted: 15th Nov 2002 01:27
No point continuing this thread.

I rest my case and will not reply further.

masked.

Mr Tank
21
Years of Service
User Offline
Joined: 25th Nov 2002
Location: United Kingdom
Posted: 25th Nov 2002 14:49
Meeow!

Perhaps you should `take this outside`?!

I think i can see what veggie is going on about- ie the `monster` website does not seem to put stuff into the frame of the car, and only really (from what i saw) does stuff for when the car is on the level, but I still found it to be useful and informative. I didn`t check out his code- just the written stuff.

You can check out my super car physics thing when

1) RGT is working again
2) I`ve finished it

Whichever comes last. At the moment i`ve got a sort of poly collision thing happening, and rolling with steering is working fine, but the whole stopping /starting thing is causing the car to jitter around in a stupid fashion. I`m probably going to go with `massless` wheels, and see what happens.

richardneail
21
Years of Service
User Offline
Joined: 26th Aug 2002
Location: United Kingdom
Posted: 26th Nov 2002 12:19
Just one question for masked, what overheads with a x, 3ds file for a landscape ?????

As far as I can see the only overhead is actually making it, but then again to get a nice looking matrix track it would take you quite a while too.

Rich

You might never win but if you give in will you ever know ??????
Mr Tank
21
Years of Service
User Offline
Joined: 25th Nov 2002
Location: United Kingdom
Posted: 26th Nov 2002 16:44
Rich,

How did you manage to get your car to go on an object? Did you get the information out of the .x file somehow, or have you got it separate, so you only check for the part of the track you are on, or what??

I`m tempted to get darkmatter and use memblocks to create my tracks. I would just use db triangles but these are slow, and lighting doesn`t work properly on them. Good idea?

Alternatively i could model track building blocks in milkshape or whatever, and stick tracks together like a scaelextric.

richardneail
21
Years of Service
User Offline
Joined: 26th Aug 2002
Location: United Kingdom
Posted: 26th Nov 2002 17:45
He He,

Everyone asks me this and for now everyone gets the same answer, I am sorry I cant tell you for now.

Hope you all understand y, I have developed Dark GT over about 4 years now so it's been pretty long haul.

Rich

You might never win but if you give in will you ever know ??????
Mr Tank
21
Years of Service
User Offline
Joined: 25th Nov 2002
Location: United Kingdom
Posted: 26th Nov 2002 17:49
fair enough!

Login to post a reply

Server time is: 2024-04-19 01:48:59
Your offset time is: 2024-04-19 01:48:59