The memory is going to stay on the processor cards. It would be somewhere between slow and ridiculously slow (by modern standards) to do anything else. The slot interface is PCIe x8. An I/O interconnect. Not memory, certainly not SMP. More like a tightly coupled cluster.

Cable companies pay television stations for content, not for delivery. It is not yet quite practical to let users pay for content independently because cable channels occupy a fixed portion of the bandwidth of the entire downstream network. When the fixed allocation of bandwidth required for a real time channel becomes irrelevant, cable channels as we know them will go away, and users will pay for content directly. The cost for delivering the content is another story.

That is what network peering arrangements are about - nothing to do with content, but rather with traffic. Big difference. It makes sense from a traffic engineering point of view for a provider like Netflix to cover part of the cost of delivering the traffic to the end user. The cost of producing the content, on the other hand, is none of the ISPs business. That is between Netflix and the end user. The ISP does not charge for content, the ISP does not pay for content. The ISP deals in traffic.

The summary is ridiculously misleading. The Senate didn't vote before midnight either. They have yet to hold a vote, although supposedly one is scheduled for about 4:00 a.m. EST. Sometime tomorrow morning the House will reconvene to consider whatever the Senate passes.

Ditto. That is the PR of the patent industry, which lives and breathes by placing impediments to the progress of science and the useful arts, but there isn't the slightest evidence that the social benefits out weigh the social costs, to say nothing of the of the rent seeking, soul sucking patent attorneys who profit from this perverse deprivation of moral and intellectual rights.

There is no shortage of literature in defense of that position. The patent system, as presently constituted, is a first class train wreck. Virtually every informed observer who is not on the payroll of the patent bar, and who lacks a vested interest in some trivial non-invention, understands that fact.

The OS has nothing to do with it. Firewire ports are DMA, as are Thunderbolt ports if I remember correctly, which means access to the port means direct access to the RAM. That means you can not only read the data, but you can also potentially manipulate it (killing processes, injecting code into already running processes, etc.).

Serious computers, running serious operating systems, use an IOMMU to restrict the access of externally accessible DMA devices to main memory, in much the same way a regular MMU is used to restrict access by user processes, by making other memory basically invisible.
http://en.wikipedia.org/wiki/IOMMU

More detail: It should not be possible to get a patent on anything. Granting monopolies on ideas is economically counterproductive, morally suspect, and intellectually perverse.

AT&T acquired a near monopoly long before the government briefly nationalized it during World War I. It was the governments concern that AT&T be required to interconnect with other carriers, and they did just that.
http://en.wikipedia.org/wiki/Kingsbury_Commitment

In addition, if you want something to blame for the establishment of the AT&T monopoly I suggest you squarely attribute it to U.S. patent law, which is highly susceptible to the creation of such monopolies, cartels, and conglomerates. There is nothing "free market" about a patent. The patent is one of the greatest offenses against free market principles ever devised.

My first instinct is to think so what? Shouldn't non-profit foundations have ambitious fund raising targets that they fall short of most of the time? Is FreeBSD in danger of ceasing to be a viable operating system because the target wasn't met?

Long distance and national phone calls are charged at a higher rate as it is the simplest way of getting businesses and wealthy people to subsidize the maintenance of the local telephone network.

That is why they invented progressive income taxes. No need to stunt commerce and communication too.

SQL Server? That is great, if you are a Microsoft only shop, and your needs don't get too out of the ordinary. Large OLTP databases tend to run on Oracle or DB2, on an operating system actually designed for the purpose. Neither is cheap of course, by any means, but together they dwarf SQL Server in market share (measured in dollars, not installations).

A bunch of rich guys injected liquidity into the system because there was no central bank to do so.

Quite true. What you do not mention, however, is that non-fractional-reserve currencies do not have liquidity crises. If fractional reserve banking (meaning legalized embezzlement of deposits) was prohibited, as it had been in common law for most of the past two millennia, the Panic of 1907 would have been over before it started.

An economy can handle massive swings in equity indices without serious problems. But as soon as people start demanding the return of a deposit all hell breaks loose. In the current system a demand deposit isn't a deposit at all, it is a loan to the bank that a depositor (lender actually) can call at any time.

No economy deals well with callable loans on a large scale, and yet our current banking system is structured on them. We just call them demand deposits instead, even though they have nothing legally in common with deposits except the name.

If too many customers start calling their loans to the bank, the whole house of cards starts falling apart. A house of cards that can grow so big that no combination of honest monied interests can bail them out. Hence the advent of fiat currency, where the central bank can make money out of thin air (roughly speaking) so that it can keep those highly leveraged hedge funds we call banks afloat.

like gold - you severely limit the flexibility of the currency to cope with a crisis

There is nothing wrong with gold. The problem is that legalized embezzlement of deposits (aka fractional reserve banking) is incompatible with hard currency. With FRB, just like any other embezzlement scheme, anytime anything goes seriously wrong, either depositors are out their money or their holdings are devalued.

It is economically impossible for holders of a debt backed currency to be made whole in real terms. All FRB institutions are ticking time bombs that can only be kept alive by debasing the currency - meaning negative real interest rates on all "deposits".

In a report I wrote last year, I estimated the markup for Internet services was 6,452 per cent

It is actually better than that, much better. Somewhere along the line, a customer sent that data to an Internet Service Provider for free. By a simple exercise of elementary arithmetic, we can see that the markup charged on data transmitted and received is actually infinite. (Shhh! don't let anyone know).

There is no American version of Brussels.

Washington D.C. is the American version of Brussels.

But what purpose does opening up their communication hold?

The issue at hand is not whether scientists should be required to publish their emails on the Internet, but whether they should have legal privilege against subpoenas. A better question is what is it about science that requires granting legal privileges that approximately no one else has? Courts exist to decide legitimate cases and controversies. How are they supposed to do that if the they cannot gain access to the relevant information? Flip a coin?