Catch up on stories from the past week (and beyond) at the Slashdot story archive

 



Forgot your password?
typodupeerror
×

Comment Re:And it's already closed (Score 1) 81

It's also a pretty tough negotiating strategy. If the governor A doesn't match the bid of governor B, Musk actually fires a lot of governor A's constituents, and the whole thing is egg on his face. Since governor A doesn't want that, he might decide to offer Musk some terms that are actually bad for the state, but will cause less personal blowback for the governor than the mass firing would have. But then there's governor B has the same incentives, and also doesn't want headlines about mass firings in his own state. He might actually decide to accept an even worse deal for the state, so as to avoid the bad headlines and instead look like a hero. So this sets up a race to the bottom which could easily save Musk more money than he spent on the cancelled construction project.

Comment Re:And it's already closed (Score 1) 81

It is possible that this is just an interruption of work, while Musk brings in new contractors who can actually keep deadlines. According to the article, the fired construction crew missed all their construction milestones. That could be the reason for the layoffs, not a cancellation of the plans.

Comment Re: Such a Waste (Score 1) 156

This is a good point. The book certainly didn't feel that way. The problem is that for the Hobbit movies, Jackson started with the original material and then decided to overdo everything about 5x beyond how Tolkien wrote it. So they can't just ride in barrels down a river - an incredibly perilous thing to begin with.

Here's how I picture Jackson deciding to "improve on" the original. They can't just ride in barrels, they have to ride in inexplicably stable barrels that don't take in water, down a river with some crazy fucking rapids, yeah!, while ... let see, why not make them be shot at by elves who otherwise never miss, but this time each shot will miss by like a milimeter, and then some other fuckers are gonna come attack them and come within a milimeter of killing them like a million times, and then good-guy elves will start shooting at those other fuckers, doing acrobatics and other bitchin elf shit, meanwhile the dwarves will be all "hoooo noooo! OOOooooo!" and we'll just keep doing that for like 20 minutes, and then more shooting and swiping and rapids, and by the end, each dwarf will almost-die like 200 times, because, you know. Tension.

You know that if Tolkien sees this he'd be like "Duh, I totally should have written it that way to begin with, it's waay more radical and gnarly!"

Comment A good step forward... (Score 3, Interesting) 82

I feel an explaination may help. This project is based on a formal specification (in Isabelle/HOL) about what should be true about the microkernel. This specification rules out things like buffer overflows, null pointer dereferences and other properties by recasting these ideas in terms of higher order logic and uses automatic theorem proving tools to verify the proofs and that implementations match the specification.

There's even a binary verified version for ARM, so you don't even have to trust that your compiler works (but, there is progress in verified compilers, so hopefully an x86_64 version is on the way). The value in this is in using the tool chain and creating new, formally specified abstractions and implementing them in an verified manner to implement more secure, robust programs on top of this kernel. Of course, the microkernel makes assumptions about the hardware, boot loader, but formal verification is used more often in hardware, and you have to trust something at some point.

This opens a whole set of possibilities to the community as a whole. As a random example, you could formalize the Arduino language (or a kernel for that language) and create a verified version of that system that runs on this microkernel. This would be a big effort, but you could do it.

Overall, this is a positive step in lower the costs of verified co-designed systems and I hope it attracts more interest in software formal verification.

Comment For the price, there's better options... (Score 2) 502

There are plenty of external boxes that allow for more options for recording and output at that price range. There's are good 2x2 boxes out there for less even.

If you are working in audio, you are using different kit. If you are an audiophile, you are probably just using the digital output into an amp anyway.

Comment OMG, not my tooth brushing!!! (Score 3, Funny) 150

This is so scary! If somebody learns every detail of the motions I make when I brush my teeth, they will basically have all the info they need to turn me into a zombie servant of the NSA-corprotocracy! And now they also want to know the humidity in my house!? Goddamn it, didn't our founding fathers say that the moisture content of our residence shall not collected? I'm so outraged! Now excuse me while I upload all my photos, featuring everyone I've ever associated with, to Facebook.

Comment Losing rights to Linux kernel? (Score 1) 349

IANAL, but I wonder if a case couldn't be made that by limiting the distribution of parts of their Linux kernel source code, they've violated clause 6 of the GPL2 in every product of theirs that uses a Linux kernel with that code, and therefore every other Linux kernel developer is in a position to sue them for copyright violation, or at least file DMCA notices with anybody distributing their stuff. I am assuming they ship products using that kernel code.

Comment Should be denser! (Score 1) 120

From the picture it looks like it takes just as much space as a regular parking garage, but I think the real potential in a system like this is in maximizing the density of parked cars. I'm picturing something like an Amazon warehouse, but with cars on each shelf. In places where space is at a premium, this sort of ultra-dense shelving system seems like the right way to store a lot of cars. What would also be awesome would be a smartphone app that gives the garage a heads up 5 minutes before you arrive to pick up your car, so that "Ray" can stick it into a pickup spot. For example, if it's in a city and on a subway line, you could choose "I'm on the northbound C train" and the dispatching system is wired into the subway system, figures out where that train is and can estimate accurately how long it will be before you arrive. Then you get a return message about which spot to go to, get in and drive off. Yes it's a bit more technology than self-parking, but the technology is mostly fixed costs, and in many dense cities, those costs are probably much lower than the equivalent number of traditional parking spaces. Also, these costs are likely to fall over time, unlike the cost of space in Gangnam or Manhattan. If it's coordinated right, it's also more convenient.

Comment Good idea, bad marketing (Score 0) 131

Think of it this way: this will be a trivially cheap device to install in a car, and it will be pretty much invisible in how it functions, until someone tries to steal your car. It will probably be bundled with other functions that count your blinks and warn you when you're too drowsy to drive safely. This is the kind of device that will pay for itself many times over in insurance savings. Also, if it records your car data in some hard-coded way, that data could be very useful in fighting wrongful traffic tickets. To market it as a spy-on-your-kids tool is not a good move. It sounds sinister and gross. Basically, it should be described as a password device for your car, which you can enter just by looking like yourself, or else typing something in on the owner's phone. If your computer requires a password to operate, why shouldn't you car, especially if entering it doesn't require any actions?

Slashdot Top Deals

It is easier to write an incorrect program than understand a correct one.

Working...