When developing a racing game or simulation, great effort has to be put into racetrack design.
Some real-life racing circuits have been licensed for use in racing games and are converted
accurately into race track files.

All racetrack files are readable only for the game they were
supposed to be used for. On the other hand, the players enjoy some tracks more than others -
why not let them choose a racetrack from their own library of racetracks included in the
racing games they have purchased? Why not race on some older racetracks? If you can continue
to use the tracks bought with a racing game, every bought racing game is an investment because
you can continue to use its content.

To me, racing tracks are a bit like places I've been. I'm almost longing to revisit the coast track
from the original The Need For Speed, yet my current PC can't run it (properly). I'd also like to
race different cars on this road or with a different physics engine.

So, what we need is a common file format for race tracks. Of corse it has to be open source, so that
everybody can use it. And we need tools to convert racing track from existing racing games into this
open format, so we can use them in future games, which support this open format.

There are many challenges ahead. For example, simulation vs. arcade tracks. Quality of the conversion.
Some games do not allow they player to leave the road, others do. Checkpoints. Traffic. Jumps. Powerups.
The new file format should allow for a head that contains lots of meta data: Which game the track is from,
which year it was created, its license, its focus (sim or arcade), point-to-point or round track, rallye or road,
grip levels, etc

To mention licenses: Of course it's illegal to spread a file based on a commercial racing game (that is not yet
abandonware). But it is lega to spread a tool which creates the open file format from the original game files.
Therefore, there might be a challenge who can create the best "export filter" for certains racing games. This
could be fueled by bounties. No one said it was easy

Well, what do you think about this idea?

Do you really think a game developer would take a 5, 10 or more years old game out of the dusty cupboard just to add support for a open source file system that probably will only result in thousands of crappy tracks in racing games, or even worse - no differences, all games would be the same?
That doesn't make sense, besides the fact that LFS doesn't support modding atm, and it's not sure if it will ever support modding - maybe with S3.

Sidenote:
Any data from a game that is older than 5 years is obsolete anyway - low poly, crappy textures etc. So you can't convert this track data into a newer game anyway.

The conversion would have to be done by fans. There are fans of old racing games around, trust me Low poly count doesn't really matter when it comes to nostalgia.

Would all racing games be the same? Is a racing game really just a bunch of cars and tracks? I guess not - see lfs

Quote from fastpeon :

But it is lega to spread a tool which creates the open file format from the original game files.
Therefore, there might be a challenge who can create the best "export filter" for certains racing games.

Actually this is a huge gaping grey area. More like a black hole.

By creating a tool that takes licensed content out of a game, and makes it publically available, would make whoever made that tool open to attack from lawyers.

Quote from fastpeon :

This could be fueled by bounties. No one said it was easy

Yes, it could be fueled by bounties, but in most situations, large projects cannot be fueled by bounties because the bounty vs time to complete ratio is too low. At the end of the day, you have to understand people who take open-source bounties generally develop software full time; and they have to beable to live. If that bounty is £1000, and it takes the guy 5 days to create the software, thats £200 a day. Which is below-par with what he/she could get as an employed developer (circumstance depending, of course).

In theory, I like the idea. In practicality there will be so many different nodes of data that will be so radically different between games, you'd have a nightmare on your hands. You'd have to come up with a standard, that includes scaling, objectives, geometry, lighting, etc. Quite a lot of games have specific hacks built into the maps, etc.

Orion's points are also still valid.

Quote from fastpeon :

The conversion would have to be done by fans. There are fans of old racing games around, trust me Low poly count doesn't really matter when it comes to nostalgia.

Would all racing games be the same? Is a racing game really just a bunch of cars and tracks? I guess not - see lfs

Then it makes much more sense to re-model all tracks from scratch, and use a game that supports modding. I don't know about rFactor, but maybe there are some tracks from old games already.

The racing games/sims wouldn't be "the same" of course, but I think every set of tracks has it's own atmosphere, and this can differ a lot. For example, a rFactor track in LFS wouldn't be LFS anymore, because it's not Eric style

Quote from fastpeon :

Well, what do you think about this idea?

i think you should understand more about what you have ideas upon.

Quote from the_angry_angel :

By creating a tool that takes licensed content out of a game, and makes it publically available, would make whoever made that tool open to attack from lawyers.

It's not illegal to publish a conversion tool. And it's legal to use the tool, as long as you bought the original content, and the conversion remains for your own use. But publishing the conversion may get you into trouble.

(The exception: the US have also outlawed conversion tools. Google for "DMCA" and "DVD Jon" to get the details.)

A 3d model is more than just a 3d model when you mix it with a game engine. Let's make a little example - LFS uses a path of nodes to determine not only what to draw onscreen, but what position to calculate you in. I have no idea what defines this path, whether it's an editor or a 3d surface with a particular lane and vertices ordered in a certain way.

If it's the latter you're in trouble, because that will only apply whilst a) The object stays in it's original format, b) the object was designed for LFS.

The problem here is that there is a whole lot more than just paths and nodes in LFS that contribute to a given games 3D format. Some games might use multitexturing to define the wet racing line and some games might use a second 3d surface at half alpha for the same thing.

There's just too many variables and too many specific engine customisations to each model to make a standard format unless the format is some kind of very generic script that allows a lot of freedom to each engine as to how to interpret it, but that's only going to make modelling tracks a horrible inconsistent nightmare and would results in a format so limited that artists wont want to use it.

I wonder how tracks for new games are currently made. New tool for each new game?

It depends on the developer and the requirements of the game. Some modellers use 3ds Max to create race tracks using splines, but then the model gets passed to somebody else with a custom tool to "rig" the track for use. It's all down to how the team works and what the requirements of the enginer are.

When I was working on a racing game I asked the artists to adhere to certain guidelines, I then rigged them for use by defining certain 'limbs' as kerb, tarmac, sandtrap etc.

LFS seems to use some kind of spline system, but whether the tracks are created in a modeller or a custom tool I couldn't tell you. I know the devs have some customised tools for this project, such things are not uncommon to be honest. That's not to say they had to write several complicated applications before starting LFS - my own custom tools tend to be very technical with peculiar front ends ... they're only for me after all, so I just write a bit of code to do a job and think of some way to instance it. Often I dont have an interface at all and I just call up the bits of code I want and recompile to do a certain thing.

Actually a spinoff from this would be a developer pool of tracks. Essentially we have professionaly done tracks that work in certain games. We, the public, would have to buy the tracks and the game seperately but we'd know that the more racing games we own that use this database the more we can use the tracks we love. The developers no longer need to worry about the quality of the tracks or diversity as the racers will decide what they want to drive and obviously the developer for the game could provide some 'free' or included tracks with thier game. Virtually impossible to do as getting the money men to agree on what is worth what and who should get the money from the tracks and games and licensing ect sadly. It would be great to know that the next version of GTR, LFS and rFactor all used the same track data. Of course there would then be a repository for fan made tracks that could be properly looked at and given not just community approval but the support of the track developers. I wonder if we can convince an F1 team to start dumping all thier track data from last year?

we can dream