There's a little unspoken benefit about what a true, affordable, universal-coverage broadband system could provide for: drones. Envision drones that can provide high quality real-time streaming (commands to the drone, imagery back) without requiring line of sight or effective cellular service.

Individuals and companies could get the sort of drone communication that today only exists for militaries. Buoyant drones (hydrogen, helium) could stay aloft for long periods and go anywhere. Conceivably a hydrogen-powered drone could stay aloft until its electronics failed, via condensing atmospheric moisture via a hygroscopic material and electrolysing it to replace the slow rate of leakage (using solar power). So picture a world where, say, anyone could buy a mass-produced mini spy drone and send anywhere, even a war zone with no infrastructure, and have it fly at a height where it would be almost impossible to spot. It would in most cases cost significantly more to take down than it costs to build (barring "drone killer" drones, but then you get to needing to maintain a large distributed inventory of them and a sensitive nationwide detection system that works at all altitudes, and you're just inviting people to come up with countermeasures). It would make it increasingly difficult to lie about human rights abuses, war crimes, armed incursions, etc.

I once looked into what it would take to make such a drone previously but quickly realized that the bandwidth costs alone via today's satellite internet services would get pretty astronomical quite fast, turning a "cheap drone" into a prohibitively expensive one. But this could change the picture. If satellite internet is cheap and widespread, not only will your bandwidth be cheap, but it also means that your connectivity hardware will also be widespread and cheap.

On the home front, one of the big concerns by regulatory bodies for all of these drone-based services companies are eager to launch is of course loss of connectivity - which is one reason why, for example, the FAA has been resisting them in the US. But if satellite service to a drone is much less likely to suffer from the reception irregularty that plagues cell phone towers. And you always have cell phone connectivity as a backup. You're greatly improving the overall reliability of your drone communications, which should make it easier to start getting commercial drone services approved by regulators.

Even in modern countries there are holes. I live in Iceland and we have one of the best rates of broadband connectivity and fiber deployment in the world. But my land is in a sparsely populated valley so it hasn't paid off to run a line out there, most people just use their cell phones for a net connection. If satellite could beat that (and wouldn't be too blocked by mountains), even in highly connected countries there's a real potential market here.

Heck, there's a lot of people who would get it if the price and stats were right even if they had ground-based broadband. Everyone here has bandwidth caps on international net traffic, only domestic is unlimited. So people who want to do a lot of downloads of foreign content might well choose that instead of or inaddition to regular broadband.

If technology had jumped directly from the 1979 Impala to the 2014 Honda Accord, yes, that would have been really f*ing amazing and a huge technological innovation, and the company that did it worthy of every bit of praise thrown at it for doing so.

Yeah, I had written a section about this but must have messed up my tags and Slashdot ate it.. Delta clipper highest achieved altitude: 1 kilometer. Falcon 9 first stage alone highest achieved altitude: 130km. Delta clipper furthest flown from the landing pad before landing: 300 meters. Falcon 9 first stage alone, furthest flown from the landing pad before landing: 300km. Delta clipper mass ratio, 2,5. Falcon 9 first stage alone, mass ratio 20 (and the boosters on the Falcon Heavy have a mass ratio of 30). And on and on and on. Not to mention that they're built utterly differently.

Aww, Slashdot ate my whole paragraph comparing Falcon 9 to the Delta Clipper, that was the best one.... must have messed up my tagging :(

Internet satellite thingy - almost identical to Teledesic

Teledesic: Launched on Pegasus rockets which cost your firstborn child. SpaceX: Launched on Falcon rockets which are cheaper than the Russians and Chinese even without reuse. Teledescic: 90s computer and communications tech (this was the era where playing the original Doom game took a high end computer and nerds envied those with ISDN connections). SpaceX: 10 iterations of Moore's Law later. Teledescic: Communcation sats have to be large objects with heavy hydrazine thrusters for stationkeeping. SpaceX: Much smaller satellites available (all the way down to cubesats), with a wide variety of ion thrusters for stationkeeping available.

Yeah, totally the same situation.

Hyperloop - first theorised by Robert Goddard nearly a century ago and a staple of SF for decades

Goddard and sci-fi: vaccuum tube. Hyperloop: tube full of thin air. Goddard and sci-fi: maglev. Hyperloop: ground-effect aerofoils. Compressor on each craft. Goddard and sci-fi: massive trains holding huge numbers of passengers. Hyperloop: small computer-timed trains to spread out the load on the track and thus reduce construction costs. Goddard and sci-fi: Trains implausibly deep underground. Hyperloop: built like a monorail. Goddard and sci-fi: tubes take the shortest route to their destination. Hyperloop: Trains go primarily over already-built and permitted infrastructure to reduce right of way and environmental costs / challenges.

Yeah, totally the same situation.

Falcon 9 - It can land vertically, like errr, the lunar module or the Delta Clipper

Tesla - Okay, they're quite nice but electric cars aren't exactly a new idea

Aww, you didn't give me an example to compare it to! Let's just go with the EV-1, since that was probably the most modern commercially-produced EV before Tesla EV-1, range 60 miles (older version) to 100 miles (newer version). Tesla Roadster, range 230 miles, and Model S, up to 300. EV-1, 0-60=8 seconds. Tesla Roadster and Model S Performance, 4 seconds. EV-1 production: about 1100. Tesla: produces that many cars in *1 1/2 weeks*. EV-1: Loved by owners but panned by critics. Tesla Model S: not only loved by owners but has been getting some of the highest ratings for any kind of car period.

Your "analogies" are akin to saying "So what if he won the Indy 500 - I raced my go-cart down the street the other day and beat a soap-box racer!"

"Cruising speed" is otherwise known as "terminal velocity" and is hundreds of meters per second. And I'll reiterate: the *point* is to go slow. Drag is a *good thing* on the way back down.

Other corrections: It's false that there's no part of the rocket that reliably faces a given angle - it doesn't tumble, it maintains an orientation generally between 0 and 15 degrees relative to the direction of travel. And the concept that bloody air is going to kill a pneumatic piston in a matter of minutes is the height of absurdity. .

Apart from actually launching a rocket to space and then having it descend and attempt to land, what's your proposed method to determine how much the fins have to move during a real-world descent and thus how much hydraulic fluid they'll consume? (beyond the simulations, which SpaceX uses extensively; they're invaluable but don't correspond 100% to real-world flight scenarios)

Yeah, it's kind of funny, but with this "failure" SpaceX will still end up recovering more of their first stage than is recovered for 99% of first stages launched in the history of man. Not like any of it will likely be of any use except for analysis and then scrap, but still...

I believe what they're saying is instead of any hydraulic system at all, which would be a simplification, no? And I have no idea what you mean by "efficiency leeching ramscoop", the whole point is to slow down.

I'm sure that SpaceX had reasons for not going with such a design. One that comes to mind for example would be during hover/low speeds - no ram air. But you don't need to be mean to the GP for suggesting that.

Yep. From 1300 meters per second to maybe 3 meters per second, with a CEP that would make military ballistics engineers salivate. So, so close.

It's funny, but there's really three analogies I use to explain the "whats" and "whys" of the hyperloop concept, and one of them is a roller coaster (the other two being the "super-high altitude airplane" analogy and the "building a pipeline" analogy).

Compare a roller coaster ride with going on a train. Are roller coasters built suchly that you have to wait half an hour or more between rides because they haul many hundreds of people at once? Do you have to spend 5 minutes boarding and later 5 minutes disembarking because of the scale? Does a pilot have to take the controls to maintain spacing and occasionally handle the risks of merging traffic and the like? And the tracks massively heavy and expensive to support these giant roller coaster cars?

No, of course not. Roller coasters are well optimized. Roller coaster cars are small, maybe two dozen or so riders at once. Because of this, they load and unload quickly. They're predominantly computer controlled with only a bit of human "central control" to send craft on their way and the like. They're all "expressways", no intersections, so all the computer has to do is make sure that it's not too close to the cars ahead of or behind it. Because the cars are small, the track can be made light, which makes it a lot cheaper.

Hyperloop implements the roller coaster paradigm to a tee.

That said, the current stage they're at, I wouldn't put people on it. They need to make sure that things are going to go as expected. Most of what they're doing is mature tech, but a few of the things, like the air-bearing skis, are going to need a lot of testing to prove their reliability. Right now they need a proof of concept and to iron out the basics. The next step up, where they have to prove the predicted reliability, repeatability, throughput, economics, maintenance etc, that would be more of the stage where an amusement part ride would be a possibility. Though I'd personally prefer that their next testing stage be built as something that, if it goes well, one could just expand into an actual hyperloop route. Maybe several dozen kilometers here - that should be enough room to accelerate up to top speed, coast a bit and deal with some curves and the like, then decelerate back down. And if it works out well, I have trouble picturing that some Vegas casino magnates wouldn't pay to link it up between them and LA. 6-ish billion dollars to enable millions of people in the LA area to pop over to Vegas in half an hour for $20 and unload a couple hundred dollars in the casinos? The amount of additional traffic they'd get would pay that off in a heartbeat.

Although... hmm, you know, they designed Hyperloop to limit passenger vertical acceleration to 1G and lateral acceleration to 0,5Gs, for reasons of passenger comfort - but not reasons of structural integrity or acceleration capability. So you know, even on actual routes, they actually could potentially let people purchase tickets to a... ahem... less G-force limited experience. ;) It'd require more car spacing, so the tickets would cost more, but when your base price is only $20... Plus, you'd get there a little faster. ;)

but why would it be lighter than a conventional train?

The main reason is that hyperloop isn't designed to achieve throughput by bundling everyone together into (proportionally) rarely launched trains, but by frequently launching smaller trains fully under computer control - spacing on the order of a few minutes instead of a half hour or so. There's only 28 passengers per pod. That launch rate is easier than many other computer controlled transportation systems, mind you, because there's no intersections - the only thing you could possibly hit is the car in front of you or the car behind you, and only by doing something really ridiculous (it's also more than enough time to stop if something goes wrong, as per the calculations, and the numbers look quite realistic).

but at the least it's going to have to support the tube plus the train cars itself

The tube isn't actually as heavy as you might think. Unfortunately I dont have my numbers on me right now, but it works out to not materially change the picture.

Is it simply that new materials and not having to share tracks with existing trains allow for different, lighter construction?

The real enabling technology for this is high launch rate, and the enabling technology for that is computer control with a simplified control problem (one way, don't hit the car ahead of or behind you). Yes, they'll use modern materials to try to keep things light, but that's not the key factor; the key factor is spreading out the load. It also adds a great deal of convenience for passengers - near constant departures and quick to get in and out of. The proposal really has more in common with a roller coaster ride than with a train: frequent small cars, quick loading and unloading, computer control maintaining spacing, etc.

But I'm concerned that it might be too optimistic

In the beginning I did too. But I've read the proposal and cross-referenced the numbers, and I'm sold. The cost figure is totally different from rail because it's really nothing like rail. For example, the track construction is far more like pipeline construction (really, it *is* a pipeline construction), so you need to compare to pipeline construction costs, not track construction costs. And it actually favors comparably with most types of pipeline, like oil pipeline, which are bogged down in environmental regulations and almost always lots of lawsuits, plus face high construction costs from having to generally go through wilderness areas, and a ton of other things. In most aspects oil pipelines have a far tougher time of it; the only thing that Hyperloop has harder is establishing and maintaining tolerances (but they have some very good proposed solutions for achieving the them quickly and affordably).

That it runs TF2 well isn't saying much. That wasn't very intense when it came out and it is very old. TF2 runs great on integrated Intel cards. Try a game that is a heavier hitter, and uses more modern API calls. Then you'll see issues.

Se what you are really saying is "A problem can be fixed by throwing enough hardware at it." Your GPU and CPU are unimaginably powerful compared to what was available in 2007. So of course it runs well, it could be running at 25% efficiency and still run well because your monitor's scan rate is the limiting factor.

However that's not so easy to do with new games that push the envelope. You can't just throw tons of hardware at them because they are already pushing the high end hardware that is out. So efficiency matters. If the driver is slow, you are going to have poor performance.

Further there is the issue of crashing. AMD drivers seem to have a tendency to 'asplode when you start throwing some of the new features at them. These features are there for a reason, they allow greater detail, more efficient rendering, new visuals, etc. If you can't support them, then that's an issue.

If you want a real test, fire up Metro Last Light Redux, see how that works.

"Oh that's an old version, we aren't going to patch the bug." Really? That's an acceptable response that something that's 3 years old is too old to patch? But somehow, taking 100 days to patch a product that's 5 years old (in 7's case) is too long? Much easier to deal with patch issues if you just declare you only support the latest greatest and require everyone to upgrade all the time, no matter the issues.

MS's response is particularly understandable given the complexity of doing regression testing on the wide variety of hardware, software, and patch sets the patch might need to be applied against. If they released it and it caused issues, well then people would cry even more about how shitty they were for not testing it.

I think you are right about the mud slinging/political office: What with Chrome books Google now wishes to directly attack MS. They want to make Windows look bad, and thus make their own product look good by comparison. This isn't motivated by being a good citizen, it is motivated by something else.

For that matter one can get all conspiracy theorist and say maybe they chose their reporting date knowing MS's patch cycle to try and create just such a situation.