by James "Nutty" Hallows

Now, we unplug our perfect joystick and plug in our standard joystick. Our frame time is extended to 34 milliseconds as a result of the added millisecond. Taking the inverse again to get back to the new frame rate, we find that it's dropped to 29.4 FPS. Well, a USB stick is no better than our 'perfect' joystick. I'll leave it to you whether that 0.6 FPS drop is worth spending a few hundred dollars on. 8).

Another point is about USB and speed. It does take less time for the game to return the data, but overall USB also has the capability to slow the system down equivalently. USB generates polls whenever it detects data change, the system collects the data and stores it in case an application asks for it.

The process of collecting the data eats CPU cycles, and it does it continuously. Even when the stick isn't installed in the Control Panel. USB data actually passes through a minimum of 5 system drivers doing this ongoing poll, and the poll from the App still passes through at least 2 drivers collecting the data. Using a USB stick, you might not see the 0.6 FPS drop keyboard control to stick control, but there's every chance you'd see a constant FPS less than 29.4 to begin with.

The second misconception is that USB is going to be inherently more stable. Many people confuse the medium with the message, or at least the protocol with the data. USB is a serial communications link, nothing more. Hang a crap pot on the far end of it and it won't work a whole lot better than an analog port with a crap pot on it. Hang a good optical sensor on an analog stick and it won't work a whole lot worse than a USB stick.

On the other hand, there are some real advantages. It gets rid of a lot of little problems with analog transmission. The 'crossover' glitch that happens when the axis values are identical goes away. It allows much more data to be moved without having to resort to non-standard methods. Not noticeable with a standard stick, but the 'jitter' that the standard TQS causes in the normal F22 wouldn't occur in USB because the communications would never need to be established in the first place.

But, actually, I'm in the rare position of being able to compare apples to apples here. The new chipset Bob has developed is virtually identical to what a USB F22 would be if one existed. The serial link is slower but, other than that, there's no real difference in what would be happening. A USB F22 might give you (maybe) 1 or 2 FPS at 30 FPS over what we're seeing with the new chips. Other than that, you'd be hard-pressed to notice the difference.

What we see with the new chips and the improvements made over the standard F22 are a result of the change in comms methods (the ability to do the mapping in the PC instead of in the stick is a result of that, too), and the application of CTFJ techniques to the driver. If you're seeing there what I'm seeing here, I think you'll agree that the new chips would be hard to improve on insofar as stability, noise, etc., and it's worth noting that it uses those nasty old pots and that nasty old gameport.

Click to continue

FoxTwoPro

So what stage is the project at now?

Well, Bob's got new chips almost ready for testing. There's still a lot to be done, but certainly, the end is in sight. It looks as though the chips will work well under Windows 2000, although obviously we have to wait a while before we can test them out exhaustively under it. I've got a lot of work to do on Foxy to get it ready and so my work is cut out - I think you can safely say that Bob and I have been busy on this for some time now, and we'll continue to work hard to bring this to fruition.

When will these chips be ready? That's hard to say. This is Bob's project, and my input is really from the Windows software side to support his work. A lot will depend on whether people are willing to purchase the chips and support Bob's efforts. Bob's currently talking about setting up a batch run for say 25 chip sets. Bob is very much in the position I was in when I started selling Fox Two Pro. It's a huge gamble not knowing whether it'll take off, and required a considerable initial outlay, and a lot of time creating the product.

With the way that Fox Two Pro has sold and continues to sell well, I'm just trying my best to convince Bob that this will be way more popular than that, and I forsee that there would be no problems shifting a lot more of these chips than I've sold of my software. So, if you're interested, drop by the Stickworks, Bob's site and check out the latest news on the project, and send him an e-mail ([email protected]).

If anyone from Guillemot is reading this, have a think about the implications of Bob's new chipset. With a minimal outlay, (ie. swap the chips and add your own labelling to the front of the stick) you could relaunch the Thrustmaster top end flight controllers as the ultimate digital high performance controllers, fully supported under Windows.

With the correct marketing, there's no reason why we shouldn't see these controllers sitting alongside every other brand of controller in the shops. With the removal of the archaic DOS software and setup that plagued these controllers, and making it easier for people to get their hands on them, you'd have a winner on your hands.

And if any of you are unsure as to whether you'd want to upgrade your controllers, well let me tell you this: You won't find me going back to the chips you're all flying with!

Join a discussion forum on this article by clicking HERE.