USB 2.0 has more bandwidth than Firewire 400.
What proof do you have that it's a hack ?
Read the spec and you'll have all the proof you need. It's a hack. And no, it doesn't actually have more bandwidth.
1. First, USB 2.0 in real-world tests gets about 280 Mbps to FireWire's 400. (Source: Wikipedia) The reason for this is that USB silicon lacks true DMA, lacks any on-chip engine for isochronous data handling, etc. As a result, it is nowhere near as reliable and cannot come close to the bandwidth of even FireWire 400 in the real world.
The highly simplified explanation is this: with FireWire, the FireWire controller itself does about 90% of the work of moving audio data into buffers in your computer's RAM. With USB, your CPU does about 90% of the work of moving the audio data, usually at interrupt time.
2. USB is half duplex. That means it can send data or receive data, but it cannot do both at once. FireWire (at least FireWire 800) is full duplex. That means it can send and receive data at the same time. Therefore the full bandwidth USB 2.0 is only 480 Mbps (or realistically, much less), while the full bandwidth of FireWire 800 is 1600 Mbps total, almost 4x the theoretical bandwidth of USB 2.0, and almost 6x the real-world bandwidth.
I've been reading people have been recording 16 channels with USB 2.0 and one guy gets 13 with Vista.
I can handle an 8 channel interface with FireWire on a 400 MHz PowerBook G3 (approx. equivalent to a P3/400) in Mac OS X 10.4. Any machine capable of running XP, much less Vista, should handle 64+ channels easily. Handling 13 channels is a complete joke.
Firewire is on its way out, as many manufacturers have switched from FW to USB 2.0. Too many interfaces to support. Heck, even video camcorders are now USB 2.0 for HD transfers.
Camcorders only use USB for reading files off disk/flash, and that's exclusively the domain of low end consumer toys. Real camcorders all use FireWire with the exception of a handful of pro models that use their own wire protocol. (Even most pro video formats use FireWire as a transport.)
The reason for this is that in the real video world, outside of special setups for ENG, you want to be able to transfer video losslessly between devices, not just from a device to a computer. USB is fundamentally incapable of supporting that because it is a master-slave bus, while FireWire can trivially support that because it is a peer-to-peer bus. I can state definitively that USB as it is currently architected will never replace FireWire in video applications except among less discerning consumer customers, and that if FireWire goes away for pro video gear, whatever replaces it will be a
lot closer to FireWire than to USB in its architecture.
And just wait for USB 3.0, which might be out next year and 100 times faster than USB 2.0
Just wait for FireWire 3200. It will smoke USB 3.0 unless they redesign USB 3.0 to use smarter silicon that takes the data transfer load off the CPU. Heck, I'd be amazed if USB 3.0 even beats the existing FireWire 800 except in burst performance given how much of a pig of a bus it is. It's really that horrible.
On the contrary, IMHO, USB is dead. It just doesn't know it yet. For keyboards and mice, Bluetooth is already cutting into its sales. For disks, eSATA works much better and will likely decimate USB in the near future, leaving only cheap memory sticks and low-end consumer junk gear as its last bastion of its use. To put it bluntly, USB is the short bus in the computer world. It manages to merely be acceptable at everything---jack of all trades, master of none.
USB 3.0 is no exception. Instead of sensibly adapting the electrical signaling to faster speeds as every other bus did, they chose to replace it entirely with an optical signaling mechanism in the same cable. If you've ever worked with optical cables, you already know why this is going to go horribly wrong. Start with the "you have to clean your USB cables or they stop working" problem, work past the whole "cables break if you wrap them too tightly" problem, and you'll eventually end up running smack into the "every hardware vendor has to redesign the hardware to support a radically different piece of PHY silicon with extra outputs to additional pins on a USB connector that now costs ten times as much per unit" problem.
IMHO, USB 3.0 is, or at least should be, completely DOA. It has absolutely no benefit whatsoever for consumers over even existing alternatives like eSATA (which not only will
always be faster than USB because of the inherent simplicity of the bridge silicon, but also is almost certainly much cheaper to manufacture than USB 3.0 will be). USB 3.0 is basically the "20th Anniversary Mac" of busses---overdesigned, underperforming, and way too expensive---built primarily "because they could" rather than because it was supposed to actually fill any particular niche. It was a last gasp attempt at making USB viable again for storage devices, and it is utterly doomed to failure because it is in every way the wrong design.
Oh, and when USB thoroughly loses the storage war to eSATA, all those consumer camcorders will likely move to eSATA, too. If all you're doing with the bus is providing a disk, you might as well use a protocol that was designed for disks instead of a protocol that was designed for keyboard and mice.
