Nobody argues there weren't things that really needed fixing. But there was also a lot of nice-to-haves that didn't have to be fixed at triple rates on rush schedules by anyone qualified to use a keyboard. It sailed by so smoothly I can't help to feel it was overhyped and overfixed. As in a paramedic can give you a band-aid, but it really wasn't necessary to go that far.

You must be providing Facebook with more data than I am, I just keep an account because some insist on using it as their RSVP system. The people it recommends are typically totally random people who have one friend in common with me.

Here in Norway we're up to 20% last year, it increases by about 3% per year (11->14->17->20) and all major rollouts (ex-DSL, ex-cable, ex-power companies) are doing fiber for new apartment buildings or housing areas. We're expecting major investments in fiber over the next years as the copper network has officially been declared a phase-out technology to be shut down in central areas by end of 2017 (first tiny test county has already shut down, it's now all fiber + mobile), it'll still exist as a legacy option in rural areas but many of the 45% currently using xDSL will move to fiber. By the end of the decade I'd not be surprised if a majority is on fiber. But then I do think we're #2 in the world on fiber penetration.

Ahahahahaha no.

And that could have been their leverage. Imagine an x86 tablet that works okay but as a bonus can run that odd must-have application you really missed from desktop Windows, like a cut down Surface Pro. Instead they totally messed up the "Surface" brand with RT and Pro that are really two completely different worlds. Yes I can understand the talking heads wanting a product that can compete in the space Surface/WinRT is, but they should have realized this was going to suck and suck bad. And it's dragging down the Windows brand with them.

But we were talking about brute force.

Or aliens could invade and use your datacenter for parking! Your only safe bet is the cloud!

Seriously, that sort of crap is vastly less likely than your business going under in the normal course of events. Encryption is useful if it keeps data out of the hands of you're competitors for long enough that they gain no advantage from it - that's all it has to do (well, and the algorithms chosen need to pass any auditing requirements).

What you're missing is there's an exponential vs exponential thing here. Compute growing exponentially means the number of bits you need in a key grows linearly. And in a couple hundred years we can't finally get to 256 bits being brute-forcible, because we hit hard limits first.
 

I have worked on a similar concept for police back in the 90's and can prove it.

If you're trusting Apple with you're stored data you're trusting Microsoft too.

256bits is really, staggeringly huge.

There's a minimum amount of power required to flip one bit in quantum mechanics. We are many orders of magnitude less efficient right now, but that power requirement is a hard limit set by the universe. Just to count to 2^256 would require more power than the output of the Sun for the lifetime of the universe.

No one recognizable as human will ever brute-force a strong 256-bit key.

"Fascist" has never meant that, except in the fantasy of socialists who can't accept that the National Socialist party really was socialist (which, BTW, was very progressive on stuff like minimum wage and universal health care and pensions and so on, at least for non-Jews - didn't stop them from being totalitarian militarists).

The "ism" for government-by-corporation is "mercantilism". Remember, for a couple of centuries it was normal and expected for the government to send the army/navy to protect the interest of the nations corporations (each of which would have a government-granted monopoly in its area of business). We may still have mercantilistic tendencies in the US government, but its not accepted (by most) as normal or desirable, unlike a few hundred years ago when that's what government was for.

Well yes the iOS front-end perhaps, but unless they also rewrote the whole decoding backend that they have to link to that part will still be under the LGPL.

When he spoke about spying on Americans, he was a whistle blower. Had he been smart, he would have stopped right there.

Sadly, that idiot carried it into treason and has not only harmed America's interest, but his own: his life.

Snowden will never ever have a normal life. More importantly, no nation will trust a man that is such a traitor. Sure, they will USE him for a time, but he will not be allowed into any place in which he could damage that nation. And within 20 years, he will want to come back to the west, and will be willing to do his time.

"unreasonable" carries a heavy load in interpreting the 4th

Only by those who would usurp the constitution's actual intent for random ideas of their own. The legal system is rife with such people; that's a good deal of the cause of many of our problems. I'm not arguing that there aren't bad judges, lawyers and legislators out there -- clearly, there are, and quite a few of them at that. I'm simply pointing out the obvious: The 4th cannot be read by a sane and honest person to turn on any individual or congressional idea of "reasonable"; that's already defined as the highest law in the land. Anything else is disingenuous (or clueless) at best, and may be an indicator of a much, much worse problem.

Last decade called and want their post back, this decade fiber is your last mile. The rest is just for in-house distribution or on the go.

Because computers typically is much faster than me at doing something, they can read Wikipedia faster than I could read an A4 page, my math speed is measured in seconds per floating point operation not the other way around and it could query huge database much faster than I could find the index cards at the library, much less find anything. What they're short on is the ability to comprehend and learn, not process. If they're so slow it's because they're waiting for humans to give them feedback or tweak their programming because they lack the inherent ability of self-learning, if they were able to review their own performance and adapt/improve/get creative they should be spinning evolution cycles faster than mayflies.

Let me express this another way, assume that you had an AI that already speaks English fluently. How long would you guesstimate that it'd take before it could speak every language fluently? My guess is that if you could do one, it could do the rest in hours - rush through all the dictionaries, the grammar, the expressions and idioms and tomorrow it'd speak Japanese, Swahili and Inuit fluently too. Meanwhile it'd take many, many years for a person to speak a hundred languages, if one could do it at all. What you're talking about is the limit as to how fast humans can teach a computer, it speaks nothing of how fast a computer could teach itself. After all us humans are used to figuring out things on our own and to surpass our teachers, a true AI would need to have that capacity as well.