Today’s operating systems, motherboards and devices comes with Plug-and-Play support, to let the device work without any manual configuration. Having said that, the system is not perfect.
The problem lies with device drivers. The device drivers which ship with each operating system release is only what is available at the time, so it cannot support new hardware which has been released after that time. This means when you plug in the device, the system is not going to find a suitable driver or it’s going to end up becoming a generic device. I’m sure you too will be quite pissed when you just installed a $300 gaming keyboard and mouse, and it defaults as a generic USB keyboard and mouse.
Up to and until now, operating systems and hardware vendors have tackled this problem by the use of internet updating mechanisms which seek out new drivers when you plug in a device. This of course relies on vendors actually actively updating the drivers in this drivers pool which so far they’ve failed, but more importantly, it requires an active internet connection. The paradox of installing a new network adapter which requires a network connection to download a new driver is a good example where this fails.
Microsoft has just patented an idea that solves both of these problems with one stone. Patent application 20080071935, “Self-Installing Computer Peripherals” for those of you playing at home.
The idea involves adding a tiny bit of non-volatile flash memory right into the device. It would only need to be large enough to store with it the device driver and any additional software required to support the device. In addition, a USB hub needs to be built into the device so both the functionality and flash memory parts of the device can be accessed by the system.
Microsoft, Plug-and-Play, Patent