Spares should be precisely what there's a lot of. To deal with actual, meaningful contingencies (trees taking out power lines, trucks driving into power lines, drunk Air Force commanders ordering live-fire practice on power lines, etc) there should be zero points of failure. Anywhere.

If a meteorite of the kind that lit up Russia early in the year, or the kind that lit up California the year before, hit a substation, no amount of armour will prevent serious damage. The CA one, discussed here as I recall, was the size of a minibus. The fragments that reached the surface - and reports say there were many - were certainly far more dangerous to a transformer than a few grams of lead.

You have to assume such a strike is inevitable. Prevention is impossible. Shielding would be stupid. That leaves option number 3 - make it not matter. It's cheap, easy, effective against any type of outage and provided you have decent routing protocols operating over a bidirectional mesh topology, resilience increases anywhere from superlinearly to exponentially.

Then what? Then you don't care if it's a meteorite, an airliner falling out the sky, an army tank driver on speedballs or Bob Bobkins, the brother and first cousin of Joe Bobkins, out hunting things that'll stay still long enough for him to point his rocket launcher. It. Just. Won't. Matter. Worth. A. Damn. The flicker of your LED house lights will barely register with even super-sensitivities. The routing protocols would have established new pathways to all destinations in microseconds, with the decisions being implemented a millisecond or two later. Nobody would notice and nobody would care.

There's an expense to redundancy, just as there is an expense to not having bridges fall in rivers. But it's a very small expense. The outages from the ice storms and rain storms? Those are big expenses. Big RECURRING expenses. With redundancy alone, due to the statistical nature of line loss, you could get extremely close to zero outage for anyone. Ever. Redundancy (down to as small a scale as practical), smarter placement of utilities (ie: not on thin poles in ice storm prone areas) and better material choices (aluminium cables?!) combined could guarantee the system would survive uninterrupted anything short of a nuclear bomb.

(You could design a complete infrastructure on a national scale that actually could withstand a full-blown nuclear war, but a lack of users would make it pointless. Unless we have developed AI by then. In which case, they and The Machines they'd need to maintain the system could endure pretty much indefinitely.)

If so, it might backfire. The NSA weren't able to prevent the attack, and if law enforcement are baffled then clearly the NSA have nothing that can identify the attackers. One genuine attack and one possible attack, nothing the existing system could do before, during or after. Fifty claims about things the NSA freely admit were fiction - well, those remain fiction.

Fifty claims that can be legitimately called false positives and one, maybe two false negatives. If you were running a company and one of your employees screwed up major decisions 51-52 times in succession, you'd probably fire them. From a canon on the top floor.

In this case, I'd argue the intelligence services and crime units have proven themselves unfit for purpose, and that the power company is too negligent on providing robust, fault-tolerant services and should have their business license withdrawn.

Scented or unscented? Basic or self-relighting? With or without wax skulls carved on the outside?

These are important decisions!

We've known the US has crappy infrastructure since, well, as close to forever as matters in America.

Attacks on a power station or substation would be immaterial if the grid was a grid, redundancy was built into the system, and getting things done was a higher priority than ego strokes and profit margins. (Yeah, heresy, I know.)

The moment you deliberately create single points of failure is the moment you hand out invites to nutcases, lunatics, wannabe cowboys and the rest of the US security infrastructure*. The moment you make such violence nothing more than a public nuisance, something not even worth a writeup in the local paper, is the moment it stops being interesting for the fringe groups to do.

*Yes, the local crackhead with the M16 and armoured personnel carrier is the "militia" the Constitution speaketh of. They are part of the national defence system. Due to two major wars inflicting a massive drain on reserves and an exceptional loss of forces due to PTSD and injuries, said crackheads form an increasingly large part of the regular forces, police and intelligence services. Frankly, I'd be far more concerned about a coup from within than a bunch of moonshine-laden rednecks who have watched too many Dukes of Hazard episodes.

Of course, given the NSA can dictate terms to the President, Congress and Federal judges, the coup might have already happened. Would you notice if it had? Would you care?

Optical SETI is an active field of research. It doesn't get talked about as much as radio SETI, in part because it is only recently that optical interferometry arrays became possible, in part because optical telescope time is expensive and in part because the atmosphere limits the quality of data for optics. There are (very recent) developments in autocorrection that reduce atmospherics, but the reality is that until someone parks an optical SKA telescope in space, the quality of telescope data won't be sufficient.

I'm busy with Chuckie Egg in one window and Elite in another. I really need to write an OXP for Oolite that will let me play both in just one. Can't wait to use military beam lasers on that giant yellow rubber duck.

Since Twitter, Facebook and the NSA don't store anything with a known value to anyone, all three should be wiped immediately.

I would agree with this (it is very largely true) but material I put online in the 1988-1992 timeframe is definitely incomplete. Massive holes in what has been retained. Yes, I will be remembered, for a whole, despite the best efforts of those who know me, but those gaps mean the Internet has forgotten.

Has it forgotten whole people? Entirely possible. If someone less wildly and eccentrically prolific in posting over those same years encountered the same gaps, they may have vanished entirely.

Could that happen today? Less likely. Archives are better and more numerous. Storage is cheaper, as is bandwidth. But one only has to look at the typical tweets and fanfiction.net contributions to realize that although more is transmitted, less is said. The total information per unit time added to the Internet may turn out to be constant.

Back in the stickshift era, Netflix directors drove the interstates/motorways at the proper speed. In first gear. Then parked by slamming directly into reverse. After all, you're supposed to burn things out, right?

Anyone checked the suicide rate of ex-Netflix employees? My guess is that it's above national average. Considerably above. Once an employee has been burned through and is no longer A lister because they're mentally shot, why would anyone else hire them? Their experience means they'll need to be paid more than the graduates who are more functional and more able. Not a good bargain. And the experience is worth nothing because web programmers are a dime a dozen and recommendation algorithms are common.

Work experience should always add value, but in this modern culture, who wants to help another company? The best short-term returns are from squeezing minds till they're dry, then throwing away the rinds. Why would companies worry about the long-term? Not on the balance sheet. As for helping others... that's..... Socialist thinking! Even when there's no competition. All for one and one for me.

Nobody knows what would piss off the wrong people to that extent. The CIA apparently had "rogue" missions being launched by "enthused" controllers. We don't know if that's true, but since I am defining the scope of ignorance, anything we can rationally say we are ignorant of is in scope. In this case, we can rationally say that the best information we have makes it possible that upsetting relatively low-level employees of any security agency may be sufficient to warrant (in their eyes) a visit.

I dispute the conclusion that you should avoid saying X, Y or Z, because avoiding the elephant in the room (or the penguin on the television) may be precisely what upsets them. It would be considered suspicious by some and if those some are amongst the controllers, not mentioning things could also get you a visit.

Hell, we know Rumsfeld held prayer sessions as head of the DoD and is alleged to have held strange superstitions about cats. If someone that bizarre could become head of a department, someone just as bizarre could be involved in CIA operations. There is a finite, non-zero probability that being an Odinist or a crazy cat lady could also attract CIA attention.

The fact is, if you are breathing (or not breathing but still functioning), you will upset someone. There is nothing you can do to avoid it, there will almost always be nothing you can do to defend against it (sorry, that's just how it is), so the old advice still holds true. Eat, drink and be merry, for tomorrow you may die.

Very true. Instill an element of fear in someone who you know will talk about it, creating an element of fear over the wider community. PsyOps. Which we know governments practice.

The Russians know no more than the rest of us - Snowden has made it clear he gave all documents to others, and this is extremely believable. It makes it pointless to limit damage - or even establishing what damage there is to be limited - by capturing or killing him.

Standards have never been as high as I'd like - typos and grammatical errors abound in articles - but noospeek is definitely a new low. I would suggest sacking the current editors and replacing them with Grumpy Cat and Happy Cat.

It is not only valid but required, so although your thoughts are appreciated, they are nonetheless wrong.

Actually, archaeology shows that only some societies are greedy. It happens that those are the societies that dominate, but that is a consequence of short term gains being militarily better than long term gains, in early history. You were very vulnerable back then and even small losses had large impacts.

Ultimately, though, it means that humans are not compelled to be a bunch of arrogant twits. At the very worst, some societies may have a genetic propensity for it, but that dictates nothing. Even if it did, sending the right-wingers to Mars (or, better, Venus) and using gene therapy or eugenics to reduce (not eliminate, that would be bad) violent tendencies should be sufficient.

The Pledge is an affront to all that school stands for. Unthinking obedience simply isn't compatible with intellectual growth or rational questioning. Obedience to a nation is also incompatible with the international semi-borderless worlds of science and art. Neither paints nor positrons have any respect for local laws or political boundaries. Boundaries exist to maximize the benefits within and minimize contagion from flawed systems, the notion of "loyalty" to any standard is relatively modern as society goes and has been a failure from start to, well, it hasn't finished yet but it's time for philosophers to stop poking at their navels and start thinking about metanations and paranations, how to draw on what has always worked (cooperation across strengths) to derive a notion that is functional, rational, sane and likely to (as an early Megadeth noted) work this time.