VDrift is a cross-platform, open source driving simulation made with drift racing in mind. The driving physics engine was recently re-written from scratch but was inspired and owes much to the Vamos physics engine. It is released under the GNU General Public License (GPL) v2. It is currently available for Linux, FreeBSD, Mac OS X and Windows (Cygwin).
This game is in the early stages of development but is already very playable. Currently the game features:
Here are some screen shots to be considered for a new VDrift splash screen.
Please rate the images in this album and say which you'd like most to be the new splash screen.
You may also submit your own screenshots to this album for consideration.
This upgrade improved my framerate very slightly. If anyone notices any negative changes from the previous version used (2.68), please post in the forums about it.
Contributed by: joevenzon on Wednesday, November 19 2008 @ 08:44 PM PST · Views: 413
As of R2225, here's a list of some of the recent fixes in SVN since the refactor was moved to the trunk:
* console logging is now saved to ~/.vdrift/log.txt
* view distance re-implemented
* initial work on automated car performance testing
* racing line re-implemented
* support for new unified coordinate system
* driver meshes re-implemented
* fixed aspect ratio problems on widescreen resolutions
* replays re-implemented and working better than ever
* in-game menu re-implemented
* car paint selector re-implemented
* screenshot control re-implemented
* mouse release on abort (thanks alex25)
* texture size option re-implemented
* many bug fixes
Contributed by: joevenzon on Tuesday, October 21 2008 @ 09:00 AM PDT · Views: 610
The refactor is complete enough that I have moved it back to the trunk. Hopefully people can give it a try and let me know how it works out. Only the Practice Game mode is working, and the graphics are very simple because the fancy shaders haven't been turned back on yet. Also, you may notice a lot of little things not quite working correctly: positioning of the spinning car, inability to change car paint, no joystick calibration, lots of spurious error messages, and so on.
You may want to move your old ~/.vdrift folders out of the way if you have trouble running it, although it *should* work with them.
The issues list (see the Issues link at the top of the page) shows what isn't working yet. All features that worked in the last release but don't now are marked with Critical or VeryHigh priority. If you find something isn't working like it did in the trunk isn't shown on that list, please post in the forum about about it and I'll add it.
Contributed by: joevenzon on Tuesday, September 09 2008 @ 10:12 PM PDT · Views: 823
It's been a little while since I've posted an update on the refactor status, so here goes. Progress has been steady and I've mainly been working on implementing all of the car physics. Tonight I just finished up camber and toe angles. The refactor car physics are almost totally complete now, although the cars may not drive exactly the same as the trunk mainly due to the lack of bumpy off-track surfaces and joystick input filtering. I also made progress with the GUI and you can now select any car or track you want. The F1-02 seems to perform better with the integration scheme in the refactor as opposed to the trunk, so that's a good indicator -- it's definitely the most demanding on the physics engine.
Next, I need to finish off the few remaining unimplemented GUI widgets and then move on to completing some simple features like tire squeal sounds, tire smoke, control filtering, off-track bumpy surfaces, TCS, and so on. Then, it's on to the more complex features like better graphics, AI, auto-X mode, replays, and eventually multiplayer.
Contributed by: joevenzon on Saturday, July 26 2008 @ 03:45 PM PDT · Views: 927
It's been exactly a month since I last posted about the refactor!
I'm pleased to announce that as of rev 2078 I've gotten basic driving working.
Many features are still missing or incomplete, but you can actually get around the track. Collisions aren't supported yet, so keep it on the track or you'll fall off of the world, the track smoothing isn't working yet so you get bumps when you drive from one polygon to the next, the partially implemented autoclutch isn't that great, the suspension travel is completely linear, there are no anti-roll bars, there's no downforce, there's no ABS/TCS, the steering controls don't have any filtering, performance is poor, and so on.
Anyway, give it a shot if you're interested. I've created a wiki page for the refactor that has information about how to check it out: http://wiki.vdrift.net/2008_Refactor_Project
Contributed by: joevenzon on Thursday, June 26 2008 @ 07:25 AM PDT · Views: 1,059
Here's a progress update on the code refactor. Most of the core subsystems (input, models, textures, math, collision, etc) are in place; the only exception I can think of is sound. The GUI is about three quarters of the way to being fully functional. I've got all of the low-level physics classes in place, but still need to write most of the car classes. Rev 2042 loads track graphic models and generate collision structures for them, creates an XS graphic model, and simulates the car body with gravity and a fake, simplified suspension. Of course, there's no engine yet, so you can't move the car at all.
The next steps are to get the car physics to load in the .car file and attach wheels and real suspension elements to the body.
This turned into almost a complete re-write. The new code is MUCH leaner. Avoiding coupling has really payed off so far; one result is that changing the contents of a header file and recompiling only results in two or three dependencies needing to be recompiled and then it's done, instead of nearly the entire project needing recompilation like with the old code. I've also been valgrinding and unit testing along the way and am pretty confident that the new code is going to be more stable and less buggy.
I'd guess I'm about half-way to having a working game back up and running, and maybe a third of the way to having all of VDrift features re-incorporated.
I'm not an expert on the subject, but from what I can see the NSV/Symplectic Euler/Euler-Cromer algorithm and the Velocity Verlet algorithms both look decent. Runge Kutta 4 is too complicated, too slow, and surprisingly isn't even as accurate as NSV/Symplectic Euler/Euler-Cromer.
Contributed by: joevenzon on Friday, April 25 2008 @ 07:49 PM PDT · Views: 1,042
Hacking on VDrift has been getting harder and harder. Each new feature that is added requires more work and makes it that much more difficult to add the next feature. So, I think it's time for some major code refactoring... and maybe even a complete rewrite of the top level code. What this means is that I'll probably end up writing down bugs or new feature suggestions in the issues system as future "to do" items and probably won't be able to address them right away. But I thought I'd post this to let everyone in on my thinking and why you might not see new features getting added as quickly as you'd expect.
I'll keep the current VDrift code in the trunk and will make a new branch for the refactoring.
EDIT: the refactoring branch is:
branches/refactor2008
Contributed by: thelusiv on Sunday, March 09 2008 @ 11:38 PM PDT · Views: 1,389
It seems worth posting that VDrift has two web sites that have recently been added or updated. We used to have a Subversion repository browser but it's been down for some time; it is now active again at http://svn.vdrift.net/. Also, you can visit http://code.google.com/p/vdrift/issues/list for the VDrift issue tracker, which is the new site to manage VDrift bugs, features, and other issues. It is hosted on Google Code.
