The linux tulip NIC drivers had some fast switching support, at least as of a couple versions ago. I never looked into it too deep; because 128port+ Xylan omniswitch boxes are $100 or less on ebay; and even with my budget constraints that just blows anything pc based right out of consideration.

The problem with arXiv.org is that as it gets more popular, the signal-to-noise goes way down (a la slashdot), as anyone with a pet theory can throw it onto the preprint server. It can take a lot of energy to sift through papers to decide which ones contain crackpot ideas, and which ones are simply ahead of the curve.

Not really. At least the high energy sections are pretty crank free. The general physics section gets a bit of crank stuff, but not much. They have a filter, so you have to have an institutional email address (or a sponsor who has one) in order to put a paper there.

Though their motivations may not be so philanthropic, at least consumer electronics corporations are on the side of the people like you and I. After all, they know that consumers will not purchase crippled, copy-protected products. Hopefully, this will result in a somewhat more balanced result when laws are passed. Call me cynical, but I feel that the Hollywood lobby's advantage is quite large and the laws will likely get passed.