As I said, I don't think secret MITM attacks are viable for many connections and large amounts of traffic. And AFAIK some browsers these days (Chrome for sure) have learned from the recent Root CA disasters in that they explicitly expect some "well know" site certificates (and google.com would be one of those), so they wouldn't trust any other certificate that's been been issued to google.com and signed by a trusted root CA.

IMHO this PRISM thing does what we always more or less suspected the NSA is doing -- it intercepts large amounts of public internet traffic, and automatically or semi-automacally tries to extract intelligence from it. A lot of SMTP traffic on the public internet is still unencrypted (even Gmail still accepts incoming SMTP deliveries without requiring STARTTLS or anything like that). So there's no really big surprise, but at least now we know for sure.

The question I've been asking myself is this: Snowden claims that he could "get your emails, passwords, phone records, credit cards", i.e. that NSA analysts can tap into pretty much any communication and read pretty much any mail that anyone sends. If that's true, why didn't he prove that to the Guardian people before he blew his cover? If I imagine I was someone who could read anyone's mail at will, and I wanted to expose myself to an outside reporter, the least I would do is show that reporter some emails that he wrote to his wife last week, or something. Just to prove my allegations.

If it's really true that the NSA essentially has random access to any Gmail/Apple/MS mail, this means that they either have a direct back door into Google's datacenters, or they're intercepting the communication at some backbone routers (which would probably imply a considerable additional load on the routers' CPUs, and they would have to have the private key to Google's server certificate (or the signing certificate, if they were to perform MITM attacks) -- I think it's well-nigh impossible to do those things in secret for a large number of connections and for an extended period of time). Frankly, I have a hard time believing that this is really what's going on.

So they minimized the number of statistical radiation-induced deaths down to some small value by permanently evacuating an area of hundreds of square miles around the reactor? Still doesn't sound like a trivial thing to me.

They wouldn't be in the privileged position that they're in now though. It would be much harder for them to lock customers and developers into their products. They would probably survive, but they would shrink substantially. MS's strategy always revolved around controlling the platform, not just writing software.

In my parents' household, they have all but done so.

That wasn't "a" flesh and bone based aircraft.. It would've taken at least two of those to take out both engines.

Maybe at his mom, when he was little.

BT Retail has started testing Carrier Grade NAT (CGNAT) with its customer.

Has the customer been informed already? How does he or she take it?

This "incident" happened in a state government (in the state of Mecklenburg-Vorpommern), not in the federal government.

Not more valuable, less common. The media primarily reports rare and unusual events. Somalis and Iraqis getting killed in terrorist strikes happens all the time. The same is true for people dying from cancer in the west (and those people aren't less "valuable" either). People getting killed in bomb attacks on western streets is rare. Hence the media attention.

They provided it with several hundred million new potential users, by building and selling the most successful (by number of users) platform in history[1] that uses OpenGL as a primary API, thus creating a huge incentive for developers to write OpenGL-based software. I think this does deserve credit.

[1] at least it was at the time; has since been overtaken by Android (which probably wouldn't have enjoyed the same success against WinMobile it it hadn't had iOS to copy from)

If Microsoft thought like that, DirectX never would have happened in the first place.

The design proposed here uses inertial confinement and pulsed (rather than continuous) operation.

It doesn't even have to be exothermic (net energy gain) on the spacecraft, without considering any ground processing. In other words, it's perfectly fine if, for each kWh of electric energy you supply into the engine, you only get e.g. 0.4 kWh of kinetic energy of exhaust gases (plasma) coming out of the engine's nozzle. What's much more important is that the engine puts that 0.4 kWh into a very tiny amount of plasma, so that the plasma's velocity is very high (for a given amount of kinetic energy, the velocity is proportional to the reciprocal of the square root of the mass). That velocity is the "specific impulse" of the engine, and it determines how much fuel mass you need to achieve a given delta-v of the vehicle.

Not really. The project may consist of several parts or subsystems, and you may want to fork in order to replace or refactor one or two of the subsystems. In that case, all changes you make to one of the other subsystems should still be easy to integrate back into the original project.