Hmmm.... is there some financial equivelant to Hawking Radiation? And can we calculate the length of time it would take such a singularity to evaporate?
Wait, a piece of software moving towards a slower, more enterprise-friendly release system, in direct contradiction of recent trends (see: Firefox 10)?
One cent is effectively nothing on most stock transactions. Remember, that's regardless of shares transferred or their total value - selling 10,000 shares of Google's $500 stock (total value: $5M) would still get only a one-cent tax. Even small stuff, it's not a barrier - I could sell my ten shares of the now-bankrupt Circuit City for $0.04 (current price), and have made a one-cent profit after that tax (I bought it at $0.0021)
Hell, lower it to a tenth of a cent if you still think it's a problem. For million-transactions-a-day HFT, that's still a huge cost. But for common trades that last more than a minute, it's negligible.
They're not a threat to the US. South Korea and Japan (NK still hasn't gotten over that) are still rather threatened - a 10 kiloton warhead hitting a major city would still kill tens of thousands, minimum. Might hit the millions, if it's the right city (Tokyo is within range, and has 13 million people). That's hardly a bloody nose - more like a knife to the leg.
True, if NK did that, they would get bombed back to the Precambrian era, but they'd have still done a rather nasty amount of damage.
It's been shown repeatedly that conversing on a phone requires substantially more attention than talking to someone face-to-face.
Plus, if the person you're talking to is physically in the car, they know instantly "crap something's going on I need to shut up to let the driver focus". Hell, they might spot a potential problem before the driver does and alert them ("watch out, that moron's trying to cut you off").
Better idea - add a low fixed tax on any stock trade, regardless of value or size. Just a $0.01 per transaction would make HFT extremely costly- if you're executing millions of trades per second, that quickly brings you up to billions of dollars per trading day. HFT can't be that profitable. Yet it wouldn't really affect people making actual investments - ones where you actually investigate the company to make sure it knows how to earn a profit.
Uh, GP never said that. You're taking a reasonable position ("distracted driving is dangerous") and taking it to absurd extremes. Nobody has ever proposed banning talking or radios, to my knowledge, and pretending that's even relevant to the discussion at hand is bad debate form.
Nanotechnology is no more a risk than any other technology. Seriously, look at some of the things that are classified as nanotech. If you make a powder fine enough it can be classified as nanotech.
There's two reasons people are so irrationally afraid. First, nanotech is, by definition, invisible to the human eye. Since they can't see how it works, they're afraid.
Second comes all the media, both news and entertainment, that uses nanotech the same way the 60s used radiation or the same way the 1800s used electricity. Quoth TVTropes: "Nanotechnology has become an all-purpose magic substitute for soft science fiction and sci-fi-flavored fantasy. Nano is the latest sci-fi name buzzword".
Anyone with a modicum of education in the matter can tell you that nanotechnology, as it now stands, is completely unable to destroy the world. That famous "grey goo" scenario? Yeah, that's not only extremely unlikely to ever happen (comparable to a virus taking over every machine on the Internet and turning it against us), but completely impossible with current engineering. We don't have nano-scale robots. That's probably further away than fusion power, honestly. The most advanced nanotech we have now is the processor in your computer - the actual transistors and wires and such are made at 100nm, possibly as small as 32nm. That qualifies it as nanotechnology. And it's as likely to destroy the world as the chair you're sitting in.
Fact: The Predator drone was originally created to defend against the flying squirrel menace. However, due to normal government incompetence, it was instead used in the War on Terror.
Soldiers at the front continued to be baffled by why the drone's decoy launcher is filled with acorns instead of flares, something they, too, attribute to typical government incompetence.
Even engine developers don't want to do that unless necessary for some really, really cool feature (realtime ray-tracing, maybe). It's just far too much work.
NO games programmer wants to get involved in bare hardware coding. That would require so much redundant code to be written, and testing would be an absolute nightmare. Even the vaunted Intel Larrabee design was going to have drivers and code so that it would appear to games as a regular OpenGL/DirectX card. You could write your own code, sure, but it would default to acting just like any other card (as far as the software can tell).
Uh, quite a few Windows games use OpenGL, even the forthcoming Rage (so OpenGL on Windows is hardly "dead"). A lot of them (especially older ones) even offer both - I can set Half-Life or Unreal Tournament to use OpenGL, Direct3D, or even software rendering.
Most popular engines support both. UE3, used by about half the games on the market, uses OpenGL on the PS3, Wii, Mac, iPhone and Linux, and D3D on the XBox and Windows.
See, you're thinking too much about Windows VS Linux VS Mac, when the developers are thinking PS3 VS XBox VS Wii VS Windows. Coincidentally, half of those ONLY support OpenGL, while only one is pure Direct3D. Since every developer big enough that isn't owned by a console maker targets as many platforms as possible, most games end up having multiple renderers.
Problem is that free-to-play only works for certain kinds of games. It took a few years of work to convert TF2 from the traditional release to the current F2P form, and the game changed so much that they lost quite a few of their early players (myself included).
Basically, it ONLY works in multiplayer (many people either can't, or won't, play online), and it ONLY works in high-variety games. You couldn't make a F2P version of Portal, for instance - there's not enough variety of useful items to sell, and you can't go far at all selling cosmetic-only items.
You can't have the entire game industry based around F2P. At least, not one that was previously established using the "traditional" model. You'll lose a significant chunk of your market just by making a game that [em]could[/em] be F2P (from the Portal 2 stats, about 40% of players never even start online mode), and you'll lose another major chunk if you do make it F2P (many long-time gamers refuse to play "free" games as they've been conditioned to expect it to be low-quality, others will refuse because quite often bought items DO trash the game balance). Even in China, F2P is a trend, not a rule. There are many, many Chinese games made according to the traditional formula.
Except that most of the latency is inherent to DRAM - in the past decade, average latencies have dropped from 10-20ns to 6-7ns, while total bandwidth has skyrocketed from 1600 MB/s per channel to 10666MB/s per channel and capacities have surged from average systems having 128MB total to average systems reaching 2GB or 4GB.
Basically, DRAM latency isn't a problem that can be fixed by moving it on-die. Nobody's found a solution yet, other than piling on more and more cache and hoping your branch predictor works well. Eliminating the standard interface in order to plop it on-die won't do anything to change that.
Uh, you seem to be confused on your units. RAM latencies are measured in clock cycles - a CAS of 11 means 11 clock cycles for a content address strobe. The exact time will vary depending on the actual clock speed. It comes out to be 6.875 nanoseconds for 1600 mHz. Meanwhile, that CAS of 9 on the 1333 MT/s comes out to 6.75 nanoseconds - pretty much the same.
This is why "faster" RAM has higher latency numbers, and slower RAM has lower latency numbers. DDR3-2400 normally has latencies in the double-digits, despite often costing more than some computers and requiring liquid cooling. And that's also why you can go down all the way to DDR-400 and see latencies of 2-3-2 - the units have "changed" from 2400 to 400.
A latency of 11 clocks is a bit low for DDR3-1600 (average seems to be 9-9-9, but 7-7-7 isn't unheard of), but not terrifyingly slow.
Cache almost always uses SRAM, which uses a whopping six transistors per bit instead of one (although it's MUCH faster, latency-wise). Thus you get much higher memory densities from DRAM, with the corresponding price advantage (which is further inflated by economies-of-scale, as DRAM is a much higher-volume product than SRAM). So that's not really what he's saying.
It's still not really a good idea, though. RAM is generally one of the cheapest parts in a computer, and is often the only thing upgraded before fully replacing the computer. Merging the CPU and RAM would just make people upgrade the entire CPU more often (since CPU sockets change so often, even for AMD).
Funny. I recall reading a report a few months ago about shockingly poor build quality on the new wave of MacBooks. I seem to recall the soldering being described as "what an amateur would do in shop class" and "surprised that the thing even works without frying itself or the user".
Apple outsources to the same company pretty much everyone else does. Their designs may be better (debatable), but your build quality is going to be identical to any other laptop.
The thing is, there's not much important left to write about. All the things people generally go onto Wikipedia for are well-covered - there's data on every country, language, mountain range, planet, president, prime minister, prince and poet. There's a ton of placeholder articles, yeah, but does anybody really want to write an article on a Venezuelan political party from the '40s, a minor asteroid of no special significance, a particular bird species (already well-documented in the family page), or an early-90s Congressman from Ohio? Those are all real examples, by the way.
There's still current events, yeah, but history isn't being made that quickly. And the rise of topic-specific wikis is draining Wikipedia of otherwise useless articles (and their authors). Why have a page for every Pokemon on Wikipedia when you can have a page for every Pokemon on a Pokemon-specific Wikipedia? That's actually a good thing - it lets Wikipedia be an introductory course to pretty much everything, and more specific wikis can be more thorough and detailed. But this does mean that all the obsessive fan-nerds will be moving their Star Wars expanded-universe character bios from Wikipedia to Wookiepedia.
So, really, is a decrease in editors really a bad thing? Does it decrease the quality of already-existing pages? No. Do we need articles being written at a high rate? Not anymore.
Probably pretty low. Athlon 1200, GeForce 2, 512MB of SDR RAM (1x 256MB, 2x 128MB), and a 120GB hard drive. In the other one, there's a Pentium II (300mHz, IIRC), and 256MB SDR RAM. No hard drive, although I could put in the 20GB hard drive I salvaged elsewhere...
Let me calculate mine. Laptop, worth $1200 when I bought it 2 years ago. Wii - $200. Desktop - built from parts, about $600. Spare desktop - salvaged, $100. Most of a desktop - salvaged, $50. Pile of dismantled parts that I'm trying to sell to a metal recycler - $50. 3 TVs, 720p - $400 each. DVD player - $50. DVD/VHS combo player - $50. Old Gamecube - $20. Old GBA SP - $20.
Yeah, that's only up to $3450. That's not everything, exhaustively, but that's the biggest, most expensive ones. The kinds of things you'd confiscate. No way I'm getting up to $14K.
True, but do YOU even have $14K in electronics? I don't think I do, unless that ancient Pentium II is a rare collectable model. While it's possible that's the case, it requires an additional set of conditions. Occam's Razor - the simplest explanation is that he was doing it commercially. Later evidence supports that.
Slashdot Mirror
Hmmm.... is there some financial equivelant to Hawking Radiation? And can we calculate the length of time it would take such a singularity to evaporate?
Wait, a piece of software moving towards a slower, more enterprise-friendly release system, in direct contradiction of recent trends (see: Firefox 10)?
One cent is effectively nothing on most stock transactions. Remember, that's regardless of shares transferred or their total value - selling 10,000 shares of Google's $500 stock (total value: $5M) would still get only a one-cent tax. Even small stuff, it's not a barrier - I could sell my ten shares of the now-bankrupt Circuit City for $0.04 (current price), and have made a one-cent profit after that tax (I bought it at $0.0021)
Hell, lower it to a tenth of a cent if you still think it's a problem. For million-transactions-a-day HFT, that's still a huge cost. But for common trades that last more than a minute, it's negligible.
They're not a threat to the US. South Korea and Japan (NK still hasn't gotten over that) are still rather threatened - a 10 kiloton warhead hitting a major city would still kill tens of thousands, minimum. Might hit the millions, if it's the right city (Tokyo is within range, and has 13 million people). That's hardly a bloody nose - more like a knife to the leg.
True, if NK did that, they would get bombed back to the Precambrian era, but they'd have still done a rather nasty amount of damage.
It's been shown repeatedly that conversing on a phone requires substantially more attention than talking to someone face-to-face.
Plus, if the person you're talking to is physically in the car, they know instantly "crap something's going on I need to shut up to let the driver focus". Hell, they might spot a potential problem before the driver does and alert them ("watch out, that moron's trying to cut you off").
Better idea - add a low fixed tax on any stock trade, regardless of value or size. Just a $0.01 per transaction would make HFT extremely costly- if you're executing millions of trades per second, that quickly brings you up to billions of dollars per trading day. HFT can't be that profitable. Yet it wouldn't really affect people making actual investments - ones where you actually investigate the company to make sure it knows how to earn a profit.
Uh, GP never said that. You're taking a reasonable position ("distracted driving is dangerous") and taking it to absurd extremes. Nobody has ever proposed banning talking or radios, to my knowledge, and pretending that's even relevant to the discussion at hand is bad debate form.
No no no! Don't introduce those two organizations! You'll reach a critical mass of evil and the world will collapse into a singularity!
Are cheesy "collectible" coins the new push on Slashdot?
The Slashdot Coin Collection - now featuring Richard Stallman, Linus Torvalds, and more! Now with limited-edition $2.56 Donal Knuth coin!
Nanotechnology is no more a risk than any other technology. Seriously, look at some of the things that are classified as nanotech. If you make a powder fine enough it can be classified as nanotech.
There's two reasons people are so irrationally afraid. First, nanotech is, by definition, invisible to the human eye. Since they can't see how it works, they're afraid.
Second comes all the media, both news and entertainment, that uses nanotech the same way the 60s used radiation or the same way the 1800s used electricity. Quoth TVTropes: "Nanotechnology has become an all-purpose magic substitute for soft science fiction and sci-fi-flavored fantasy. Nano is the latest sci-fi name buzzword".
Anyone with a modicum of education in the matter can tell you that nanotechnology, as it now stands, is completely unable to destroy the world. That famous "grey goo" scenario? Yeah, that's not only extremely unlikely to ever happen (comparable to a virus taking over every machine on the Internet and turning it against us), but completely impossible with current engineering. We don't have nano-scale robots. That's probably further away than fusion power, honestly. The most advanced nanotech we have now is the processor in your computer - the actual transistors and wires and such are made at 100nm, possibly as small as 32nm. That qualifies it as nanotechnology. And it's as likely to destroy the world as the chair you're sitting in.
Fact: The Predator drone was originally created to defend against the flying squirrel menace. However, due to normal government incompetence, it was instead used in the War on Terror.
Soldiers at the front continued to be baffled by why the drone's decoy launcher is filled with acorns instead of flares, something they, too, attribute to typical government incompetence.
Even engine developers don't want to do that unless necessary for some really, really cool feature (realtime ray-tracing, maybe). It's just far too much work.
HAHAHAHA!
Wait, you're serious?
HAHAHAHAHAHA!
NO games programmer wants to get involved in bare hardware coding. That would require so much redundant code to be written, and testing would be an absolute nightmare. Even the vaunted Intel Larrabee design was going to have drivers and code so that it would appear to games as a regular OpenGL/DirectX card. You could write your own code, sure, but it would default to acting just like any other card (as far as the software can tell).
Uh, quite a few Windows games use OpenGL, even the forthcoming Rage (so OpenGL on Windows is hardly "dead"). A lot of them (especially older ones) even offer both - I can set Half-Life or Unreal Tournament to use OpenGL, Direct3D, or even software rendering.
Most popular engines support both. UE3, used by about half the games on the market, uses OpenGL on the PS3, Wii, Mac, iPhone and Linux, and D3D on the XBox and Windows.
See, you're thinking too much about Windows VS Linux VS Mac, when the developers are thinking PS3 VS XBox VS Wii VS Windows. Coincidentally, half of those ONLY support OpenGL, while only one is pure Direct3D. Since every developer big enough that isn't owned by a console maker targets as many platforms as possible, most games end up having multiple renderers.
Problem is that free-to-play only works for certain kinds of games. It took a few years of work to convert TF2 from the traditional release to the current F2P form, and the game changed so much that they lost quite a few of their early players (myself included).
Basically, it ONLY works in multiplayer (many people either can't, or won't, play online), and it ONLY works in high-variety games. You couldn't make a F2P version of Portal, for instance - there's not enough variety of useful items to sell, and you can't go far at all selling cosmetic-only items.
You can't have the entire game industry based around F2P. At least, not one that was previously established using the "traditional" model. You'll lose a significant chunk of your market just by making a game that [em]could[/em] be F2P (from the Portal 2 stats, about 40% of players never even start online mode), and you'll lose another major chunk if you do make it F2P (many long-time gamers refuse to play "free" games as they've been conditioned to expect it to be low-quality, others will refuse because quite often bought items DO trash the game balance). Even in China, F2P is a trend, not a rule. There are many, many Chinese games made according to the traditional formula.
Except that most of the latency is inherent to DRAM - in the past decade, average latencies have dropped from 10-20ns to 6-7ns, while total bandwidth has skyrocketed from 1600 MB/s per channel to 10666MB/s per channel and capacities have surged from average systems having 128MB total to average systems reaching 2GB or 4GB.
Basically, DRAM latency isn't a problem that can be fixed by moving it on-die. Nobody's found a solution yet, other than piling on more and more cache and hoping your branch predictor works well. Eliminating the standard interface in order to plop it on-die won't do anything to change that.
Uh, you seem to be confused on your units. RAM latencies are measured in clock cycles - a CAS of 11 means 11 clock cycles for a content address strobe. The exact time will vary depending on the actual clock speed. It comes out to be 6.875 nanoseconds for 1600 mHz. Meanwhile, that CAS of 9 on the 1333 MT/s comes out to 6.75 nanoseconds - pretty much the same.
This is why "faster" RAM has higher latency numbers, and slower RAM has lower latency numbers. DDR3-2400 normally has latencies in the double-digits, despite often costing more than some computers and requiring liquid cooling. And that's also why you can go down all the way to DDR-400 and see latencies of 2-3-2 - the units have "changed" from 2400 to 400.
A latency of 11 clocks is a bit low for DDR3-1600 (average seems to be 9-9-9, but 7-7-7 isn't unheard of), but not terrifyingly slow.
Cache almost always uses SRAM, which uses a whopping six transistors per bit instead of one (although it's MUCH faster, latency-wise). Thus you get much higher memory densities from DRAM, with the corresponding price advantage (which is further inflated by economies-of-scale, as DRAM is a much higher-volume product than SRAM). So that's not really what he's saying.
It's still not really a good idea, though. RAM is generally one of the cheapest parts in a computer, and is often the only thing upgraded before fully replacing the computer. Merging the CPU and RAM would just make people upgrade the entire CPU more often (since CPU sockets change so often, even for AMD).
Funny. I recall reading a report a few months ago about shockingly poor build quality on the new wave of MacBooks. I seem to recall the soldering being described as "what an amateur would do in shop class" and "surprised that the thing even works without frying itself or the user".
Apple outsources to the same company pretty much everyone else does. Their designs may be better (debatable), but your build quality is going to be identical to any other laptop.
I've done 3D modeling and photo editing using a 1.2gHz Athlon, 384MB of RAM and a CRT heavier than some UPS units. It's slow, but definitely possible.
So Wayland is more like X12, then.
The thing is, there's not much important left to write about. All the things people generally go onto Wikipedia for are well-covered - there's data on every country, language, mountain range, planet, president, prime minister, prince and poet. There's a ton of placeholder articles, yeah, but does anybody really want to write an article on a Venezuelan political party from the '40s, a minor asteroid of no special significance, a particular bird species (already well-documented in the family page), or an early-90s Congressman from Ohio? Those are all real examples, by the way.
There's still current events, yeah, but history isn't being made that quickly. And the rise of topic-specific wikis is draining Wikipedia of otherwise useless articles (and their authors). Why have a page for every Pokemon on Wikipedia when you can have a page for every Pokemon on a Pokemon-specific Wikipedia? That's actually a good thing - it lets Wikipedia be an introductory course to pretty much everything, and more specific wikis can be more thorough and detailed. But this does mean that all the obsessive fan-nerds will be moving their Star Wars expanded-universe character bios from Wikipedia to Wookiepedia.
So, really, is a decrease in editors really a bad thing? Does it decrease the quality of already-existing pages? No. Do we need articles being written at a high rate? Not anymore.
Probably pretty low. Athlon 1200, GeForce 2, 512MB of SDR RAM (1x 256MB, 2x 128MB), and a 120GB hard drive.
In the other one, there's a Pentium II (300mHz, IIRC), and 256MB SDR RAM. No hard drive, although I could put in the 20GB hard drive I salvaged elsewhere...
Let me calculate mine. Laptop, worth $1200 when I bought it 2 years ago. Wii - $200. Desktop - built from parts, about $600. Spare desktop - salvaged, $100. Most of a desktop - salvaged, $50. Pile of dismantled parts that I'm trying to sell to a metal recycler - $50. 3 TVs, 720p - $400 each. DVD player - $50. DVD/VHS combo player - $50. Old Gamecube - $20. Old GBA SP - $20.
Yeah, that's only up to $3450. That's not everything, exhaustively, but that's the biggest, most expensive ones. The kinds of things you'd confiscate. No way I'm getting up to $14K.
True, but do YOU even have $14K in electronics? I don't think I do, unless that ancient Pentium II is a rare collectable model. While it's possible that's the case, it requires an additional set of conditions. Occam's Razor - the simplest explanation is that he was doing it commercially. Later evidence supports that.