Since most of the mobile gadgets weren't designed with the car in mind (iPods, mobile phones, etc.) they aren't car friendly inherently.

When designing an in-car interface, there are roughly two ways to go about it: stick with what people know, possibly losing functionality in the process, or give the user something new to learn, possibly losing users in the process.

The iDrive is supposed to be great - if you spend the half hour to study it and the two weeks to get used to it.

Everyone has gotten along for the last 50 years with a simple paradigm: a volume knob, a tuner knob, and five or six preset buttons. Even AM radios worked this way, and the current top of the line sat radio tuner works similarly.

But when we add 3d navigation + localized search for points of interest, all tied into google and yahoo, automatically tracking where your car is and taking into account your current outlook calendar and when your next appointment is coming up, calculating the traffic conditions in real time and telling guiding you to a Thai restaurant, what is the user interface designer supposed to look like? Should it work with the UP/DOWN knob and the six buttons you already have on your stereo? Or a touch screen? Or eight buttons on the left and right sides of the screen? Or steering wheel controls? Or some über-knob between the driver and passenger seats that shapeshifts depending on the function?

In my own in-car UI designs, i've tried to model the "lazy user" - someone who gets out of his car, hops into his friends car and expects to be able to figure it out WHILE DRIVING.

Although this is certainly not the ideal, I'm not just trying to cover our potential legal liability (I'm doing that too) but I'm trying to be realistic about people's use of the vehicle. If you decide that the majority of people DON'T ask for help, don't read the manual, and fumblingly try to apply their existing, fragmentary knowledge of human-computer user interface to the new gadget and get impatient if it doesn't instantly understand them.