I'll just mention for those unaware that the author of the post is one of the founders of PLOS, which is an umbrella organization under which a number of journals (including PLOS ONE) are published.
The point about having an outlet to destress is a really good one; its something I struggle with (feeling guilty about "wasting time" even though I know some amount of relaxation is necessary to stay productive).
Tangentially related, the author's product looks awesome too. I'm probably going to need to make some websites for my parents businesses soon and I'd like something that I can maintain happily with emacs but that they can post to and modify from their browser; this thing looks like it might just be the ticket.
I think Julia has the potential to be a contender in this space as well. The advantage over Rust or Go is that you get a dynamic language (with a REPL and everything), and still can have C-like performance. You also get metaprogramming and optional typing (so its there if you want safety but you can ignore it if you just need to crank something out). To me it really feels like a blend of the best of Python/Ruby and statically typed languages.
Sure, but it provides similar concurrency semantics (putting things on and taking things off of channels), which is what I think David is really trying to emphasize.
Meshnets don't scale well anyway so perhaps strength in numbers applies here: if there are so many small meshnets that getting a mole in each one is totally impractical, spying on meshers en masse becomes totally impractical.
In my experience D3 is very much like that. Before you wrap your head around the way its data binding model works it seems totally arcane (esp. if you are like me coming from using other visualization tools, like matplotlib, where the hello world example is just "plot(x,y)"). Keep banging your head against the wall, it will eventually crumble and all will become clear.
I (and I'm guessing some of the people involved with Julia) would argue that regardless of how good the numerical programming tools are, you'll never get the average working scientist to use Haskell. That sounds like an interesting project though, I've become curious lately about the prospects for integrating high performance numerical computing with functional programming.
Spot on. We are laughably far away from being able to do the kinds of things with living systems that the author suggests. An analogous claim would be that because the Wright Brothers just invented the airplane, we will soon have cheap faster than light interstellar travel. I don't mean to be overly negative, this stuff is exciting! But it is important to have perspective and keep in mind how far we have to go.
I'm acutally playing with Clojure and Julia right now and I'm running into the same problem. Never used Java and a lot of the Clojure material assumes you're coming with at least a little bit of Java background. That said Clojure is a much more mature language, Julia is still very new, its APIs are changing rapidly, etc.
Agreed about too many interesting languages though :)
I had the same experience when I stopped using Opera and didn't have the RSS reader anymore. I do miss reading webcomics regularly, but the productivity gain is nice.
Nitpick, but its worth pointing out that Hacker School as a program and an institution is a very different beast than App Academy, Dev Bootcamp, etc. and is somewhat mischaracterized in the article.
I think what the GP is arguing is that if you want to be really fast you should forget using Java arrays and just wrap BLAS, LAPACK, etc., which are written in "close to the metal" languages, optimized within an inch of their life, have been around for decades, and are used by others with similar goals (numpy/scipy, etc.). As the GP says, Java libraries that already do this are probably available, so this may be a pretty trivial task.
Nah, onion routing just bounces HTTP requests/responses around before sending them on to their final destination (which is still a just a server somewhere).
What the GP is talking about sounds something more like freenet or cryptosphere. When you query for some data in these systems you have no idea where it is stored and the people who are storing it / sending it back to you have no idea what they are storing.
There are also a number of extenuating differences between his Baltimore example and PRISM, the most critical being that the recording of data in Baltimore was limited both in time and scope by the case it was being collected for. Presumably once the investigators either caught the culprit or gave up on doing so, the payphone taps were removed. They also were only allowed to tap phones they thought might be relevant to that particular investigation. The war on terrorism used to justify mass wiretapping, on the other hand, will literally go on forever and relevant data could be anywhere. Terrorism is not an enemy that can plausibly be captured or defeated (like a drug dealer or the Nazis), nor is it localized to a few city blocks or ISPs or websites, therefore there is no limit on how long this data will need to be collected for or what sort of data might be required.
I'm all for brief and limited surveillance authorized by warrants tied to a specific case (e.g. his Baltimore example), what I find unacceptable about PRISM is exactly the thing Simon glosses over as irrelevant, its scale: virtually unlimited in both duration and scope.
Hey I'm a bit late on this but I figured I'd chime in. I'm part of the current batch and I left school two weeks early to attend, so its possible even if your schedules don't line up. That said, I was only a part time student last quarter and it was still pretty tough doing HS by day and finishing up my finals by night, but I made it happen, feel free to email me if you want to talk about it (email in profile).
To me this seems like the root of the issue. Politicians give the voters what the want; If most are willing to trade civil liberties for security (or at least the illusion of security) than that's the way it will be :(
Awesome! A really cool version of this would let the user write their own algorithm and then have users compete against each other for whose algorithm can evolve the best car.
Saying that Myriad was the first to discover BRCA1 is somewhat disingenuous; other labs did it independently at roughly the same time. Moreover, deciding when a gene was "discovered" is a minefield. What constitues discovery? Identification of a mutant? Cloning the gene? Sequencing it? There are many steps to this process.
just use a long, randomly generated passphrase and write it down on paper. you'd be surprised how quickly muscle memory sets in after you type it a few times, and if you forget you always have the paper backup (keep it in you wallet or desk).
Thank you for bringing some sanity to the discussion. "our country has weathered worse" is true and an excellent way of putting all this shit in perspective.
No. That chain ends at the second step "digital genome sequence". In principle, we could synthesize a bunch of DNA strands corresponding to the genomic sequences and ligate them all together (although it would be absurdly expensive and laborious), but this would not constitute a chromosome. Eukaryotic chromosomes are organized in the nucleus by a huge variety of scaffolding and modifying factors (histones, etc.) into structures called chromatin. Recapturing this from sequence data is impossible, even in principle; the information is just not there. Indeed, understanding nuclear organization at a bunch of different leves is one of the big challenges of modern genetic research.
I think it would really, really help to have a few example images of what the damn thing actually looks like on a webpage. People are inherently visual creatures and imho this would go a long way.