[citation needed]

Seriously, show me a double-blind test with a decent sample size where it's found to be statistically significant that people can distinguish a non-retina ipad from a retina one. Not just calculations on the theoretical resolution of the human eye.

I will note as evidence supporting my theory that "retina==overkill" that monitors recommended for professional photo/video editing, including Apple's own "Thunderbolt MC914LL/B", have less than half the DPI of a retina screen. If it's not popular among people who have virtually unlimited budgets for buying displays, then that's because it's not necessary.

I would add Fortran (90/95/2003/2008) to that one-item list. Not to spark a flame war here, but Fortran can often be 10x faster for number crunching, given the same amount of programmer-hours to code it. Not saying you can't make a C program equally fast, but the default way people structure Fortran code for number crunching results in faster code than the default way people stucture C code for number crunching.

Take IPA for instance: you have to put in some work in order to write C code that will actually benefit from a compiler that has IPA support. Limit your use of pointers, only strict aliasing, etc. For Fortran, you have to put in some work in order to write code that won't benefit from a compiler with IPA.

I care about Moore's law winding down. For my applications (CFD) it means that not only do I have to start paying attention to the fact that I'm running close-to-metal, e.g. I have to minimize the amount of cache misses, but also that if I want to have a scalable application, I can't do it without MPI. And MPI is tricky.

Well, the US military has one called Active Denial System, but that's not the one I was thinking about. I can't seem to find any good links at the moment, sorry :( But it looks like the ADS is not very popular anymore with the military (never been used in the field), so I'm guessing they scrapped the civilan versions. Too much negative PR potential for a riot control device.

You could have a look at the ATLAS-I machine built at Sandia in the 70's, they used it for simulated EMP testing. 200 gigawatts @ 10 megavolts, focused at ~50 m distance.

For reference, the vehicles that have been built for riot control using RF radiation (which obviously put out much less power than an EMP, otherwise they wouldn't be usable in cities) have charge-up times of more than a day, and they're probably supplied by 480V three-phase at several hundred amps.

The "Project Thumper", as someone below mentioned, charges the capacitor bank at 110V 20 A for a few minutes, so that's roughly 4*10*60*24= 60 000 times less energy than something which has much less energy than an EMP. Myth busted.

http://xkcd.com/908/

The other reply pretty much nailed it. BTW, you can get converters that run HDMI over cat6 in order to have non-pathetic range, but the good ones are expensive.

Out of curiosity, does anyone know how much microwave radiation you have to submit a modern car to before the ECU craps out? I know S-band radar, which is basically what microwave ovens use, can disable an ECU. There are law enforcement agencies using such devices. But I haven't seen any numbers on the mW/cm^2 needed. I do know that brief pulses of 10 mW/cm^2 is the human safety limit for ovens, so it's probably in that ballpark.

So I'm thinking, with a $50 oven magnetron at 1100 W and a parabolic reflector, it shouldn't be hard nor expensive to drive around in an old pre-ECU car randomly disabling other vehicles. You might even be able to do it at a decent range!

I should also say: I've gotten some very cool photos in bars by using the camera built-in flash and a full glass of beer as a diffuser. At 1/8@f/2.8, ISO 400 (with a good IS system/small sensor) you get quite a bit of ambient light to balance the shot nicely.

This is by the way something professional photographers tend to overlook: that larger sensor produces less noise at a given ISO/shutter/aperture, sure, but with a smaller sensor and an equivalent IS system you can often go as much as a stop or two lower in shutter speed before it gets tricky to hold the camera steady.

The kind of photography discussed in the article uses active illumination. That changes things quite a bit; I have no problems taking pictures in a wedding with what i call the "IKEA Fong sphere" on top of my camera, I can even do it at f/2.8 @ ISO 200.

See this for an example photo and how to build one yourself.

Analysts have pointed out that the way Amazon is investing, it is clear that it intends to use it's current advantage in online retail to build a distribution network that is sufficiently advanced that any competitors will have to spend huge amounts of time/money to catch up. It's a long tail gambit.

Imagine in twenty years being able to look at your infrastructure and be confident that no-one on the planet has anything on the same scale (except perhaps the US military). Then you can start churning a profit like nobody's business (pun intended).

The code snippet you showed may be "clean", but I have NO IDEA what it does, and I program daily in 5+ different languages. That tells me that Ruby is a poor choice, since the syntax doesn't resemble common languages like C, python, perl, fortran or bash.

The important question is: does it do analog input? That's the biggest limitation on the RPi IMHO. Adding a cheap AC current sensor for home automation monitoring would be a lot simpler if the RPi had analog inputs.

Informative and funny! Not bad for an AC, not bad at all.

Allow me to translate that into non-grant-proposal-writing-scientist-language:
"It was fun to study this, please give us more money so we can continue having fun."