1) While I don't know specifics, I'm sure streaming, probably oct-trees, and possibly compression/decompression of nodes in the oct-tree.

2) You need to add or modify the scene in an editor and create animations if you want those. No dynamic lighting or shadows at this time, pending hardware advances (it can be done, just not very well in realtime).

3) Laser scanning can quickly build rooms identical to real world rooms without having to go in and model everything. The flaw with it is anything behind any blocking objects won't be seen, so I suspect you end up with a partial model of a room. It is great if you want to observe the world from a single point, but most of the time you want to move around (this is my main skepticism so far). It is possible maybe with software to take several laser readings and then assemble the complete room. As I said, though, this is where I'm skeptical.

At the end of the video they say they can do animations and a demo of it will be coming soon. With them, that means around 2016+, but other companies and researchers have demonstrated voxel rendering and RAGE proves animation can be done with it.

Early graphics hardware was designed to texture triangles and didn't contain a bunch of highly parallel general purpose units like today's graphics hardware. Since raycasting (and raytracing) and voxels can be done in parallel, voxels are making more sense again. You can make entirely unique terrain without overlapping and blending a bunch of textures using disk streaming methods and some structure (Carmack used a sparse voxel octtree in RAGE [here's a BSD licensed example of a sparse volume octtree).

But as I said in other posts, don't expect fancy lighting and shadows until next gen cards are available (in fact, they maybe are by now, but they're way out of my price range).

But John Carmack is the entire opposite, and his previous game, Rage, uses voxels for the entire terrain, so I think the technology is feasible. Really the main problem is lighting as raycasting and raytracing both need very expensive algorithms for soft shadows and soft lighting and the hardware to do it justice is not quite here yet. The methods I've seen for realtime result in either blocky shadows or grainy lighting.

First off, many genres of games don't need dynamic global illumination, so that is less of an issue that you make it. In fact, the only games I've played lately that have a global illumination model were Guild Wars 2, the Sims 4 (and yeah, 3 is generally better, but it isn't terrible) and Tomb Raider (2013 version). I have dozens of strategy, adventure, RTS, and racing games that have no global illumination at all. I don't play a lot of shooters outside of the few I know don't make me sick, but yeah that market uses a lot of global illumination.

Second ray casting and ray tracing both can support forms of dynamic lighting, assuming that's what they're using, and I seriously doubt they are doing something else like marching cubes to extract polygons, The problem id had with it in RAGE (which was also voxel based) was the GPU just couldn't keep up for lights and shadows, so they didn't put it in (and are working with hardware manufacturers to make next-gen cards more voxel light/shadow friendly). Techniques like real time radiosity tend to leave blocky shadows due to needing unacceptably large patch sizes. Real time photon mapping (a faster technique than radiosity) tends to create "grainy" images. AFAIK, there really isn't an acceptable solution yet.

In college I wrote a radiosity renderer that was certainly not realtime on a 100Mhz processor with some dedicated graphics hardware (an SGI Onyx, I believe), but it rendered scenes with eye catching detail, including dust particles in the air, which is really difficult to simulate with polygon renderers, or at least was last time I tried (but rendering took something like 72 hours... sigh).

Where polygon rendering tends to look fake to me in that architecture video is with curved metal objects. Something doesn't quite look right, especially with the sinks, but maybe it's because reflections are missing. Some polygon based graphics actually run something akin to mini raytracers in a hardware shader to get reflective curved surfaces (in fact, I've written one for shadows and I could do curved surface reflections with it).

Sadly, leaking radioactive plutonium-239 would probably be better than leaking radioactive strontium or cesium due to the inverse relationship between half life and danger to tissue. Plutonium has a half life of around 24100 years, so you could probably have a brick of it in your bedroom and it would never be a health threat to you. Yeah, it isn't the 1.26 billion years of potassium or 14 billion years for thorium (which is in granite), but it still is a very long time.

Actually, had Fukushima had US standards, the backup generators would have been placed above flood levels and the disaster likely averted. Japan's ignorance of this known and acknowledged design flaw was largely their own fault, IMO. Chernobyl, as you said, was an intentional test that wasn't communicated properly. The other major non-test nuclear disaster, Three Mile Island, was caused by an equipment failure followed by misdiagnosis by engineers (a light indicated a valve was closed when it was in fact stuck open).. Again, though, shoddy safety standards were at fault.

As for the oceans around Hawaii, they probably were damaged more by the 106 above ground nuclear bomb tests the US did at the Pacific Proving Grounds - I doubt Fukushima and Chernobyl will ever do as much damage as those did, even if all of the reactors there had resulted in full meltdowns..

Cold fusion is a myth as far as we know, and the other powers you mentioned are actually forms of nuclear.

Solar is fusion energy from the sun
Geothermal is fission energy from the earth's core

I find it kind of funny that clean energy is usually nuclear (with gravitational such as hydro being the exception), despite anti-nuclear people favoring them.

There is actually a reasonable solution to the waste, but anti-nuclear people say it is a proliferation risk and cannot ever be built, unfortunately. It seems on-site reprocessing can never overcome being a security issue and will always be a proliferation issue despite several countries already pushing forward with Gen IV technology that can run on it (because breeder reactors can make it into fuel). Most designs being implemented (like Russia's BN-350/600/800/1200) are once through without reprocessing, but largely because the US design had reprocessing, it was killed in 1994 and never resurfaced. Russia is actually using the BN-800 to reduce their weapons grade plutonium supply

In other words, Russia is using this reactor for the exact opposite reason of proliferation, consuming weapons grade plutonium for energy.

Microsoft has every other consumer OS hits going back to Windows 97 - ME flop, XP hit (2000 was generally considered a server OS, the follow up to NT), Vista flop, 7 hit, 8 flop, 9... hit?

I expect they'll fix the desktop experience in 9.

Oh wait, you said disappointment - yeah, usually that happens too :D

For me it usually is WHY THE HELL DO YOU NOT HAVE A MODERN FILE SYSTEM!? NTFS is way long in the tooth and barely supports metadata, much less user metadata. I like to tag things so I can find them later.

To be fair, all communism so far has been dictatorships from the beginning. Since communism is an economic system, there is no reason they have to be, however. A communist-republic is perfectly feasible. Just saying they don't turn into dictatorships, they already are.

  Fascism I wouldn't say, as it ties too much into expansionism and racial conflict rather than social conflict, though it seems Putin is having a go at part of Fascist doctrine (the belief that strong countries have a right to claim territory from weak ones).

Heh, well bull shit (and cow shit) are useful field fertilizers. My grandpa had his herd in the fallow field every third year (on a corn->soybeans or wheat->fallow rotation or maybe it was the other way around).

Misogyny isn't the right word, but objectification probably is. The entertainment industry is really shallow that way, and I'm saying that as someone that has worked professionally in both music and game programming/game music (I've also done game art and 2d and 3d modeling, but I probably couldn't cut it there in modern games - it didn't take as much talent in the 1990s). I lobbied to have female characters in one of the games I worked on and was mocked and told by the owner that there was no market for that (and all I really asked for was ONE female character NAME in that game, which was a driving game - incidentally, someone modded them in later). I also have a friend in TV and that industry isn't much better, and he is probably the best example of that (for years behind his back my friends joked of him as our male slut friend, but he finally settled down with a long term girlfriend and even moved in with her and sold his bachelor pad, which he refused to do with the other two fairly long relationships he had).

Not sure what specific issue you have, but usually when I look at comparable sites like Tom's Hardware, they agree with each other. And it isn't like AMD/ATI was much better - they were completely in Microsoft's pocket for years going by driver support, as in they didn't support OpenGL extensions for years forcing the OpenGL group to make more and more frequent releases. I worked on a cross platform graphics engine until about OpenGL 3 and ATI's support was bleak at best.

As for Intel vs AMD, Intel has had the better processors for years at the high end. If you want to go mid-tier or especially low end, yeah AMD is a good choice. When I priced out mid-tier AMD vs Intel I still went with Intel based on performance comparisons given by multiple sites including AnandTech and Tom's Hardware. In no case did AMD have a better product in the price I was looking at ($150-170), but they did have a Intel outperforming product in the price range about $20-40 cheaper. Their SoCs are also extremely good if you are really cheap. I have a friend that built a nice A10 box a few years ago.

Except these drives use the SSD as cache, or at least mine does. In other words, you don't actually install anything on the 8GB, the drive decides what should be there by demand. In my experience, it does speed up most disk operations,but compared to a dedicated solid state drive it is still much slower. Personally, I can live with the slower speed with 2TB solid state drives (non-hybrid) ranging from $2000-7000 right now, at least for any with a reputable brand name. I've seen 1TB drives for about $500 as well, but my entire build was about $760 and even the 1TB drive would have pushed my build well over $1100.

Back in college (LOOOONG ago) I often used pico because of vi's arcane command set and emacs equally bad commmand set and abhorrent load time (it took 2+ minutes to start on our minicomputer, which consisted of 6 6502 processors - it was basically 6 Apple ][s and we had faster PCs and macs in other labs, but they didn't have Pascal, which was still the learning language of choice until they year after I had it). Oddly enough, I learned vi later because of some of the things it could do that pico couldn't (at least not at the time), especially searching for and removing linefeeds and other hidden characters from files. Hidden characters caused problems when I started writing my code on my home computer and transferring it to the UNIX server via network drive for testing/debugging.