That paper is from 1996. The updated epilogue contains this quote:

The only conjecture I heard which didn't sound totally "out there" was that the animals sensed the changes in the electric field caused by the piezoelectric effect due to stress changes in the rocks before an earthquake.

Something along those lines at least.

The time you spend on the road depends linearly on the speed. However the kinetic energy of the car, the main factor w.r.t. the breaking distance, depends on the speed squared. Thus while you're somewhat less likely of being involved in an accident when you drive fast due to less exposure time, the results are probably significantly worse if an accident happens.

If you're doing 50 mph you will manage to come to a complete stop in about 180 feet[1]. At 60 mph initial speed, you'll be doing 35 mph after 180 feet.

Personally I'd prefer the risk of a daily slap on the wrist over a knock-out punch once a month.

[1]: http://www.brake.org.uk/facts/speedscience.htm

Only for memory. There's a lot of other resources out there where the GC complicates things a lot. Files and mutexes to name some common ones. At least .Net has "using" which is a poor mans RAII, but it's still easy to mess up and have a lingering resource handle.

I admit I've only programmed in C++ for 6 years, so I'm a bit new. But I've not once come across any code which sounds like the mess you describe. On the other hand, the code which does use it is significantly more clear because of it.

offsetTangentPos = LocalToTangentMatrix * GlobalToLocalMatrix * (globalPos + offset);

is a lot clearer and closer to the math to me than

offsetTangentPos = Matrix.multiply(LocalToTangentMatrix, GlobalToLocalMatrix).transform(globalPos.add(offset));

I'm no C/C++ expert, but I think I'd list RAII, operator overloading (especially for math-oriented code) and templates.

Sure operator overloading and templates can be abused, but when used with moderation I find they make it much easier to write clear, concise code.

Not curious at all. And, as mentioned in previous posts, the shutdown has been planned for several years already. If the Higgs' energy was higher than the roughly 125 GeV it seems to have, LHC would have found it a lot sooner.

The reason for not shutting down earlier was that they wanted to be sure that if it wasn't found by the time they shut down, then the accumulated data would be sufficient to rule out the Higgs. They calculated that in order to do that, they had to run until the end of 2012.

In fact, the current run at LHC was extended after the Higgs discovery was made, for the sole reason of gathering more Higgs data.

The kernel changes sound good, similar to how Vista introduced a lot of good kernel changes. However none seem like it will significantly increase my productivity. And without installing 3rd party apps, the whole clusterfuck that is forced Metro significantly reduces my productivity. I know. I tried.

So, again, there's not really a whole lot of reasons to upgrade a desktop computer if you're a power user. This is in stark contrast to Windows 7 which I found to significantly increase my productivity compared to Windows XP.

For me the bottom line is this: what does Windows 8 do that Windows 7 can't do, or can't do well? Besides playing some hypothetical DirectX 11.1+ only games.

Windows 7 was a major step forward compared to Windows XP and Vista. Windows 2000/XP was a major step forward compared to Windows 9x.

So in which positive ways are Windows 8 a step forward? Ok so they've increased performance a bit, but I got a beefy PC with SSD, so I doubt Windows 8 is considerably faster. And that's about it.

And yes I did use Windows 8. And I don't dislike it just because it's cool. I dislike it because it's forcing me to work in a way which just isn't productive for me. If there was an option to disable the metro stuff and having the Win7 start menu, I would probably find it quite nice. Instead they decided not to give me any choice...

Which is nice if you use a handful of applications. However some of us regularly use a lot more than that. Off the top of my head I can count 35 different applications I launch on a daily basis, most of them many times a day. I'm excluding all games and such.

Pinning 35 apps to the taskbar would take so much space there's hardly any left for the windows that are active, and so the windows would quickly become grouped, reducing efficiency significantly.

I just cannot phantom why Microsoft didn't give us the option. Metro might be nice on the desktop for some people, and perhaps on a tablet. But why force us? I would have upgraded if I could have kept my start menu, so a lost sale here. Great move MS.

Not sure what rig you had, but my when I ran my i7 860 + AMD 4870 full tilt (prime95 in background and playing Battlefield Bad Company 2) it drew about 360-370w, measured with a kill-a-watt-like meter at the socket.

True, however if you're planning on using the FX in a rendering farm or similar, then the Intel CPU is definitely more bang for the buck. Not just the CPU itself but you'll be able to use less machines in the farm to get the same performance, which would be a huge saving. Using the LuxRender data, 10 i7 2600k boxes would give you the same rendering power as 17 FX-8150 boxes! I've assumed FPU performance scales similar to the PassMark score there, which I think is reasonable.

So that's 7 more motherboards, RAM sticks and PSUs compared to using Intel. And then there's the power drain...

So yeah, for anything FPU intensive the FX series could end up being an expensive investment. Which is sad, I'd really like to buy AMD gear.

He's probably referring to the Buldozers which do have terrible FP performance, primarily because two "cores" share a single FPU.

For example, while the FX-8120 gets about 77% of the score the i7 2600k gets in PassMark, in a very floating-point heavy, multi-threaded rendering application the 8120 gets about 51% of the 2600k.

While I'd love to buy more AMD, I've had quite a few bad experiences with the motherboards for AMD, while I haven't had any issues with the Intel based motherboards I've bought over the last 5 years or so. As such I'm keeping to Intel for now.

Generally kept to Asus, and not the cheapest models. Is this just my bad luck or?

Guess it depends on your measurement. Blender uses (hosts) Python 3 for scripting.