Okay fair enough, but then still, the argument falls on what it means to be a "site" and whether the new form is similar enough to the old form to be considered fundamentally the same thing.

But the point is, challenging TPB in the state it's in now would challenge the entire premise of DHT - and I don't think the case can be made that DHT is skewed towards any particular use, let alone illegal uses. They can't actually have any knowledge of what is being shared because the hash doesn't reveal what it is without first having the file (or otherwise, having an additional database that links the hash to a description). This would seem to make a stronger case for this incarnation over the previous being lawfully neutral.

The question then becomes: does the md5 hash of a file, being linked to a swarm of peers with the files themselves, become symbolic of the property that is being pirated ("stolen") in a convincing enough manner to implicate the hashtable host? It seems to be a stretch.

That's kind of the whole point. This isn't a show-stopper, it's a show-starter. It's going to force similar innovation from Google, which means that everyone wins.

It's not often I'll cheer MS on, but this is awesome. I can't wait to see how Google responds.

"It's understandable that researchers are wary of having their work referred to as mind reading. They call it neural decoding." Mind reading is a severe overstatement of the reality of these technology, considering that the "reading" here is stimulus driven, and requires dedicated stimulus-training in order to be "read". These algorithms cannot "read" anything that they have not explicitly trained in the user, so it's much more like recognizing the conditioned brain-responses with respect to the task and/or algorithm.

Hey, this very logic once saved the human race, afterall.

Okay good point. I was gathering my statistics from Wikipedia on nuke counts, which are apparently decades old?

I understand the difference is that Twitter has forced limits on length, but the argument would be, what, that there's a memetic competition for writing format, and that Twitter is going to win the fight? The popularity of Twitter is way overestimated, wasn't there a study showing how most of the people who join Twitter only ever use it once and then forget about it?

Err. For the mathematically anal, it would ACTUALLY be: ((Dual GTX + Dual Radeon) / 2), but you get the point in the OP.

I like the methods used in the tests here. It gives a clear picture of what you would expect in ideal conditions. That is, the combination of two different graphics cards (GTX + Radeon) should be the average of proprietary dual-configurations (Mathematically this would be: Dual GTX + Dual Radeon / 2). Here, it's pretty clear there is a large benefit (as compared to single configurations), but unfortunately it looks as though the faster card is still being somewhat bottlenecked by the slower one, where the performance is much closer to the slowest card than it is to the mean as you might expect.