Re:The real Ajay Bhatt, co-inventor of the USB port
by weaksauce on Sunday October 11, 2009 @01:47AM
(#874383)
It was cheaper because it was not proprietary like the firewire port. I think that was the only issue really. Firewire has a smaller protocol overhead so at a given bit rate the throughput was faster. Did the original firewire spec allow for multiple devices on the same bus or was that included in a later revision?
Re:The real Ajay Bhatt, co-inventor of the USB port
by cnvogel on Sunday October 11, 2009 @12:56PM
(#874898)
The physical data interface of USB is simpler (only two signalling wires, D+, D- instead of the two differential pairs for firewire). Supply of power to the devices is simpler (+5V instead of >20V with firewire, which needs an additional DC/DC converter on the device). For most primitive low-speed devices (mice, keyboards, toy rocket launchers) the data transfer implementation is simpler (see e.g. the bit-banging software-only V-USB implementation for microcontrollers).
So on the low-performance side, the advantage is clear. And end-users like the simplicity of having the same physical port for all their computer peripherals (which can use low- (<1Mbit), full (12MBit) or high (480MBit) speed, without the user knowing or noticing).
While I personally think that Firewire is the more elegant, better performing and more flexible communication technology, I can see why USB surpassed it in the marketplace.
USB 1.0 seems like one of the many interfaces that will eventually end up obsolete just like ISA PCI AGP PCMIA etc...
I really don't see anything technically superior about USB (feel free to correct me). The main reason it won out over Firewire was of course Intel and Microsoft backing, and that it was in the sweet spot for the price/feature trade off.