One of the biggest advantages of the open source world is the ability to freely fork whatever you want, with no limitations on what you are going to make out of it.
Collaborating with other developers is not always working out, as sometimes we see selfish behaviors coming from the founders of projects, or because some developers want the project to become something different from what it is now, or because some projects have a lot of potential in more than one sectors and people would like to explore them.
What we are going to see on this article, is two examples of racing game forks. Fasten your seatbelt and take your position at the starting line cause here comes…
Speed Dreams is a TORCS fork that aims to be a fully featured racing game. I consider the case of Speed Dreams to be a very successful fork because first of all TORCS is a very mature and stable project that shouts “fork me”. Why fork TORCS instead of helping in the developing of it? Well…TORCS aims to be easy to modify, compatible, performant and stable. To achieve these goals you have to be conservative in what you add to the game, and always try to have the least possible maintaining volume.
Speed Dreams is actually fulfilling a very natural possibility for TORCS which is a fully featured racing game. That means adding new cars and new tracks, working on the improvement of physics and simulation, and also better graphics and visuals. The work of Speed Dreams developers has been great on this context so far. Currently the game is in RC state with a little more than 200 bugs to go before version 2.0. You can try the stable 1.4 version and have all the fun in the world racing . It may not be the best racing game you ever played, but it surely is the best for the linux platform.
The case of Stunt Rally is completely different. It is based on the Vdrift project which is in heavy development for 7 years now. Vdrift may be the most promising open source racing game, aiming to be high quality in all respects, but it still isn’t very playable. It suffers from a lot of bugs that show themselves once every few seconds of gameplay in the form of invisible bumps and other crazy stuff. Vdrift developers still work on all sectors sometimes changing main components completely. Is it wise to base your project on Vdrift? Let’s see…
First of all there are big performance issues. Stunt Rally offers high quality optical effects and visuals but with great cost. The minimum requirement regarding the graphics card is a GeForce 9600 GT or a Radeon HD 3870. I tried to play the game using a GeForce 9600 GT and it is only playable using the “medium” settings. Using the same hardware I can play Oil Rush at “High” settings… Besides the performance issues, there are also bad gameplay elements present. All of the available cars feel more or less the same. I don’t know if that is somekind of simplified physics or just my idea but that is what I felt while playing Stunt Rally. Default steering settings are not too good either. Thankfully you can optimize the steering through the options but it still isn’t very good using the keyboard. I suppose that a steering wheel is the right gadget to use when playing Stunt Rally. Last but not least, the physics sometimes become as “strange” as in Vdrift.
Stunt Rally looks amazing and it sounds stable (version 1.5), but things are not quite that perfect. I don’t want to degrade the hard work of the Stunt Rally developers though. Stunt Rally offers many great maps to splash into, enough cars to drive, and a very nice optical experience if you have the corresponding hardware. I find the idea very good, and I am sure they will make this game better and better as the time passes. No one can complain about an open source project not being very good. Stunt Rally is a free game after all and it has the potential to be really great in the future! It’s just that my own experience with it wasn’t very enjoyable. What was yours?