Pop quiz, hotshot: what can an adversary who can do some IP spoofing but is not otherwise in a privileged position do? A week ago you'd say something like 'ddos or something' but nothing scarier because TCP streams aren't vulnerable to muckracking unless you can listen in. Well, thanks to an improvement in security, Improving TCP's Robustness to Blind In-Window Attacks, the answer is now "learn the sequence number, window size, RST the connection and inject arbitrary DATA packets into the stream". This is being somewhat soft-sold as a javascript injection attack but of course it's actually quite a bit more severe. Fortunately, the fix is pretty straightforward as it relies on the nuances of a specific implementation in the Linux kernel. The lesson is clear: don't even try. Wait no, the lesson is "global state is the enemy". Many thanks to cp for bringing the full scope of this to my attention.
Government agency gets hacked, yawn right? Well, this time it's the NSA (a.k.a. Equation Group (pdf)). Snowden -- who isn't dead -- provides some context. More coverage here and here with a (dead) link to the files and some indications that at least one of the exploits (against Cisco ASA) is functional.
nom, a rust parsing framework. NB, this is very reminiscent of well-built regexes: r'(?P<areacode>\d{3})[ -]+(?P<exchange>\d{3})[ -]+(?P<extension>\d{4})'
Cubesats are neat, basically a small, modular-ish common form factor and bus brings the bar down for building your own space probe (PC architecture anyone?) Generally though, they don't have propulsion right now, they just orbit until they re-enter and are destroyed. Many folks are working to change that, most interestingly Deep Space Industries which swear they'll be launching a commercial asteroid mining mission this decade using their Comet-1 engine which is literally a steam engine with a charming Isp of 175s. The idea of using water as your root propellant isn't crazy of course: it's inert, safe to handle and launch, safe to have hanging around on the ISS (unlike hydrazine which hates you) and extremely common in the solar system. It's such a good idea that there are a few other folks working on similar systems: HYDROS, which electrolyzes the water into H2 and O2 and burns that for a respectable 300s Isp, and CAT (which might be the same as the one from this kickstarter!), which just turns it into a plamsa and claims a totally absurd 1623s Isp. That's ion engine territory.
Of course, the BLOCKCHAIN being a libertarian paradise means there's no recourse here, the DAO screwed up and its investors screwed up by trusting it and the thief was just plain smarter than you are so them's the breaks, right?
Don't be silly, of course they decided to go back in time and hard fork the network. Problem solved, right?
Nope. Now you just have 2 Ethereums, Ethereum and Ethereum Classic! (Same great taste, less hypocrisy!) What happens next? ETH wins, ETC wins or everybody loses.
"How Vector Space Mathematics Reveals the Hidden Sexism in Language". arXiv paper. Choice quote: "One perspective on bias in word embeddings is that it merely reflects bias in society, and therefore one should attempt to debias society rather than word embeddings,” say Bolukbasi and co. “However, by reducing the bias in today’s computer systems (or at least not amplifying the bias), which is increasingly reliant on word embeddings, in a small way debiased word embeddings can hopefully contribute to reducing gender bias in society."
How do you get energy on Mars? Everything you bring to the surface costs (optimistically) $thousands per pound so you're not going to be lugging a tank of unleaded up there. Nuclear would be ideal but it turns out people don't like radioactive chunks falling on their heads if something goes wrong. Solar is tempting but since Martian insolation is only 42% of Earth's you already have to bring twice what you'd need on earth. If only you could manufacture the cells in-situ.
Enter Perovskite cells:
They don't need to be manufactured in a billion-dollar chip fab, in fact you can make them yourself in an undergrad chem lab.
They're nothing more than a series of spin-coated layers of magic liquid on top of a generic glass substrate.
You're already doing Methalox ISRU so you need a large bootstrap power supply to cryogenically cool all that rocket fuel, once the tank is full you've got excess power on hand, why not use it to make some glass? Along with a small spincoater and some jugs of magic Perovskite juice you could be turning (lots of) energy into solar panels.
It'll still probably be up to humans to do lots of the final assembly, though I'm eager to see what automation looks like when human labor is outrageously expensive instead of minimum-wage.
The primary job of a rocket is to make fast things come out one end. To do that they have a lot of other ancillary problems like pressurizing tanks, changing direction in space and making sure the fuel is at the bottom of the tank. Instead of a dozen different systems to do this, why not just bolt an inline-6 internal combustion engine to the thing?. Lots more information in this forum thread.
Some stock market return visualizations: A triangle and Meteors and Snakes. Note that all of them are implicitly for a single lump sum investment not the inherent dollar-cost averaging that us working schmucks do.
I can't quite believe that a six-engined, double-fuselaged megajet will actually exist, much less actually launch a payload. If they find a way to duct tape a Dream Chaser to it as has been proposed, expect to see a Gulfstream version of it in a decade.
The moon is closer but much more annoying to get to than Mars because you can't aerobrake. If, however, you can manage to put a big metal highway on the surface, you can magnetobrake. Yes I just made that word up, deal with it.