Apparently I'm bad at math. 2123 is blatantly wrong. I what I meant to say was 2043. That'd be really good, but there's no way photovoltaics would grow that fast.
First you say that people are going to put in new photoelectric arrays that get them power cheaper than off-grid. Then you say that people are going to use their power more efficiently. People buy SUVs and need to have their arm twisted to use obviously-superior compact fluorescent bulbs - there's no way in hell they're going to use energy more efficiently as it gets cheaper. Cost down, demand up - that's basic economics.
Electricity is naturally something that people will keep wanting to use more of. Currently, that is resulting in 25 gigawatts/year of new generating capacity being installed. That rate of increase is at least holding steady, but demand is probably accelerating.
You seem convinced that photovoltaics are going to overtake all existing centralized power generating technologies, and that we just have to wait for it to happen. How long will that take? Can we afford to dump that much more carbon into the atmosphere?
I can imagine no scenario that would make uptake fast enough to have a significant impact on global warming. Today the entire world only has 4 gigawatts of photovoltaic generation capacity. I'm sure that the industry can't produce that much capacity in a year, but let's say they can. And let's say their production capacity doubles insanely fast - every five years. And let's say all those new panels get sold in the USA. That means that in 2017 photovoltaic installations will be producing more new power than fossil fuels. In 2023 we can stop building new coal plants in the United States and start shutting down coal plants. Almost sounds good... but it's not until 2123 that we can close the last Coal plant... in the USA... while people in China are starting to notice that it really hurts when you go outside in the hard acid rain.
extend the grid so that the Sun is always shining on it
That's... not going to work. Unless you have superconducting long distance transmission lines. Which are free. And you don't. Storage might be a solvable problem - high temperature low price superconductors isn't something we'll see this century, at least not *that* low price.
You seem to be assuming some sort of radical new electricity storage technology that takes away the whole issue of "base load". Storing electricity is hard, to the point that some energy companies are resorting to shit like flywheels. Maybe ultracapacitors will come down in price or something, but that's a technology that today is where photovoltaics were in 1975. Until some sort of really cheap and efficient electricity storage technique exists, the "constant power output" property of nuclear power plants will remain an advantageous feature.
It'd be really nice if solar took a bite out of coal, but I think you're overestimating the adoption rate. Currently the entire install base of photovoltaics in the United States produces less power the new Coal plants coming online in 2007. Even if photovoltaics experience exponential growth until Coal is gone, we'll still be putting in new Coal plants for more than a decade. That's the problem I see, and the problem that everyone advocating renewables over nuclear seems to ignore - the growth rate of actual solar/wind/tidal power installations is far, far behind the growth rate of energy usage. The difference is, and will continue to be, new coal and natural gas plants - simply because modern nuclear plants are illegal.
I think we might be able to agree about the desirability of distributed power generation.
Distributed power generation is definitely preferable, simply from a political perspective - decentralizing money makes lobbying less annoying. Unfortunately, I'm still not convinced that the technology really exists to compete economically with just letting the electric meter spin forwards.
That was my point. I'm pretty sure they'd have to to cause a meltdown in an Sodium-cooled fast reactor.
It is basically normal operation in the absence of coolant. The fuel cannot sustain the heat, melts, flows together and continues to react as is pools generating more heat.
If reactors like the IFR couldn't handle operation in the absence of coolant, there would have been a meltdown when they did a test and disconnected the entire cooling system when the reactor was running in the early 90's. What happens with that design is that when the fuel gets hotter than normal operating temperature, it expands. This decreases the number of neutrons emitted by volume below the level required to sustain a reaction, causing the reaction to stop.
The issue with sodium as a coolant is that is can oxidize very rapidly if there is an oxygen leak, leaving no coolant while at the same time producing quite a bit of kinetic energy of its own, leaving little confidence that specially arranged reactor geometries could be preserved.
Given that the reactor design will safely shut down in the absence of coolant, this is a simple question of math: If the sodium explodes as the result of any possible scenario, can it do enough damage to the reactor to interfere with safe shutdown? Sodium isn't C4. The coolant pipes aren't that thick. The claim that the Sodium wouldn't be able to damage the core enough to significantly change the material properties of the fuel mixture seems entirely reasonable.
It would be nice if gas-cooled or helium-cooled fast reactor designs were a bit more advanced. That would significantly reduce the amount of "OMG sodium blows up if it gets wet" clouding the issues related to nuclear power. Heck, it would be nice if we knew the Polywell worked - that would solve our power / nuclear waste problems. None of this wishful thinking changes the fact that we have a fully functional power plant design, Sodium-cooled Fast Reactors, available today that looks to be strictly better than any other large scale centralized power generation technology.
It looks quite a bit like the entire problem of cost is a side effect of hysteria over safety. If anyone were ever allowed to just take a good reactor design and build it rather than squabbling for years over the politics of nuclear safety, I'm sure the costs would be much more reasonable.
meltdown is a risk for any reactor
Perhaps you don't understand the energies involved.
You apparently don't appreciate the chemistry and engineering involved. How these reactions work is pretty well understood, and this specific problem has been worked on for 40 years now. In most of the modern reactor designs, you couldn't get them to melt down if the engineers on site tried to make it happen. The fuel mixes and/or chamber geometry simply won't sustain a runaway reaction.
As for sodium as a coolant, they were running the IFR for years - including a test where they had the thing running at full bore and just went and disconnected the sodium cooling system. Even if someone dumped a bunch of water in the cooling system, it's not like a sodium fire produces anywhere near enough neutrons to sustain a sub-critical fission reaction.
Hysteria is absolutely the right word to describe at least some of the public demand for more safety in nuclear plants. Activists have been complaining about the dangers of nuclear power for 40+ years, and engineers have been responding by making new designs more foolproof. The thing is - the Activists either don't know about or discount those improvements (possibly because the actual plants in production are old junk and don't have many of the well-known design improvements) and continue to complain about safety issues that largely don't exist in modern designs.
The assumption that anti-nuclear activists seem to make is that nuclear accidents look *worse* than Chernobyl, that *any* nuclear plant is susceptible to drastic meltdown resulting in a containment breach and spreading high level radioactive waste around in a giant area. You seem to assume that all Nuclear Plant designs are the same, as when you dismissed my reference to the IFR (An early 1990's sodium-cooled fast reactor prototype with passive safety properties) with a reference to a 1959 meltdown at a reactor that happened to also be sodium-cooled. Are you really unwilling to consider the possibility that between 1959 and 1990 that nuclear engineers might have come up with better designs that avoid well-known problems?
The direction I'm interested in is a mainstream linux that I can deploy on joe sixpack's computer.
That niche is well supplied by Ubuntu. If you think the space needs competition to keep things fresh, it has it: Linspire and Mandriva.
Debian *doesn't* need to target that niche at all. The existing policy of slow, steady progress and periodic rock-solid server releases produces a distro that's an excellent basis for projects ranging from Ubuntu to Embedded Debian to build from. There are occasionally cases where Debian is a bit slower than it should be (multiarch for example), but that doesn't mean that what Debian is should be thrown away to make an Ubuntu replica.
So, what you're saying is that - at worst - Ubuntu is as bad as Microsoft Windows. In order to fully use recently-released supported hardware, you need to download a driver package from the manufacturer and install it.
Intel suggests getting a binary version from the distro maintainers, but a search of Linux user forums suggests that finding such a beast has not proved easy for people using a wide range of distros.
My guess is that this is your problem. It's great that you have the option of using whatever distro you want, but not every distro will provide the same level of user experience. Ubuntu provides binary blobs for wireless drivers. Other distros like Fedora do not. For Fedora, an Intel wireless card isn't supported hardware.
My suggestion: Use Ubuntu, it works. If that's too hard, use Linspire. But don't complain when the distro you chose doesn't work with the hardware you chose.
Copyrightists are approximately as scummy as your average politician. Mostly, they don't understand exactly what the economic effects of the policy they are pushing are compared to the alternatives, but they think they'll get big payouts themselves if they get it implemented.
I suggest telling Adobe and Microsoft to die in a fire and using Free Software that can be fully supported locally instead. These harsh copyright laws are doing you a favor - why let yourself get sucked in to the upgrade treadmill and culture-crushing poor local language support that is proprietary software?
What does "piracy" have to do with rape and murder? Pirates are *really cool* guys like Captain Jack Sparrow who try to look tough but are really good at heart. "Pirate" is also a somewhat amusing term for people who like to share data that other people are trying to sell.
Trying to fight against the terminology at this point is silly. It's much faster and easier to just accept the term and run with "Pirates are cool" - it's not like the term really has any negative connotation at all among people young enough to make a difference culturally.
Since I'm not persuaded that the breeder cycle is a good idea at all, I don't see it as a solution to the waste problem.
I suggest you actually look into nuclear fuel cycles in a bit more detail. When you actually look at the nuclear chemistry of the processes, it's pretty obvious that you put in nuclear fuel (including "spent fuel") and you get out energy and low level waste. The two most mentioned issues, cost and proliferation, are both seriously overhyped - it's not like you can make a nuclear bomb out of Plutonium 240 mixed with random Actinides.
I've got the impression there are patches for Linux which does the same + more, and what about say trusted Solaris or similair? Are OpenBSD really better than those?
You can't patch in security, nor do you get it simply by complying with some FIPS standard.
Linux is maintained optimizing for features. If you want support for every USB scanner and 15 different file systems, Linux has that. Security and stability are sacrificed, but not too much.
FreeBSD is maintained largely for performance. Security and features are sacrificed, but not too much. It probably won't support your digital camera or encrypted HFS-Formatted USB key.
OpenBSD is maintained for security. Performance and features are sacrificed, and you'll notice if you try to use it. Stability and security are synergistic, so OpenBSD is rock solid. For some applications (network appliances and hardened servers), stability and security are strictly more important than performance or features. In those cases, OpenBSD is the obvious choice.
If you are using hardware that should be supported and having problems, that's a bug. You should report it, and it will get fixed.
As for
But if Linux distributors want to get their product used, they need to make sure it works out-of-the-box on Dells.
It *should* work out of the box on Dells, except for "Dell Wireless" - which is illegal to provide out of the box support for because it is based on a Broadcom chipset. If you can manage to click "Intel Wireless" when you buy your Dell, there's no reason it shouldn't work.
More relevantly though... operating systems should really come pre-installed. Dell desktops *are* off-brand for Linux, because Dell doesn't support Linux on them. If you buy from someone like http://www.system76.com/everything will Just Work, or at least Just Work better than Windows does.
Hey, if you're right and we can really make and install enough solar panels to put a dent in our coal usage, that'd be really nice. In that case, we'd still want to build some modern nuclear plants so that we can burn the "spent fuel" from all the PWRs, but there's really no hurry - that stuff's not going anywhere.
I'm still not convinced that the solar's coming fast enough though - I'm mostly seeing all the new coal plants that are getting built.
Chernobyl was an important lesson about fission plant design and safety. Since then, we've designed power plants that simply don't fail like that. Seriously - something like the IFR doesn't even damage itself (much less melt down) when you do ridiculous stuff like physically disconnecting the cooling system with the thing running.
Any industrial process is dangerous. Many of them involve dealing with highly toxic materials. With nuclear power, it's reasonable to accept the fact that if you eat the fuel rods or something you're going to die of radiation poisoning. But, as a power generation technology, modern nuclear fission is damn safe - at this point, far safer than it really needs to be because engineers have worked overtime trying to combat irrational hysteria with safety improvements.
When I say "major failure", I mean something worthy of media coverage even without the hysteria surrounding "radioactivity". You know, where the result of the accident was somewhat more drastic than "At least one of the engineers on-site at the time, Phil Harrigan, later died in his '50's of leukemia."
Even Three Mile Island, hyped as some sort of horrific disaster, caused so much damage that "statistically, someone will probably die of cancer" from it.
I bet more people die every year because they get hit by windmills than because of the US nuclear power industry.
One of the big problems with your argument is that "renewables" aren't really just one thing. If you look at them that way, you'd see "10% renewables" in the USA, which is strictly misleading. That's all hydropower. Your link about Germany mentions Wind and Hydropower.
There are places that are good for Wind. Great, we should use them. After that, you have "slightly worse" places, but the energy of wind varies the cube of the speed, so slightly slower wind means significantly less power - quickly falling to the point where its uneconomical. As for hydropower, there are only so many dammable rivers; in the USA, we've already dammed them all. And... we're still building more coal plants.
The point is, until I see a technology like photoelectric cells or ethanol-generating bacteria really take off, putting in modern nuclear plants really looks like the only thing that can actually slow down our carbon emissions. And new reactors are the only good answer to the nuclear waste issue.
The nuclear power industry in the USA isn't the greatest, sure. But it's actually pretty amazing that outdated junkers they're running haven't had a really major failure yet. Those plants are 40 years old, and they have no innate design protection against meltdown.
The point is, there have been 25 or 30 years worth of design improvements since those plants were built. We'd have gotten the full 40 years worth of development, but some damn protesters got the USA out of the game 15 years ago. Sure, there have been experimental failures getting to where we are today - but now we have reasonably good designs that support a useful fuel cycle and don't melt into a puddle when the control computer breaks.
We can't "end the production of nuclear waste". These old, junky PWRs are providing 22% of the power in the United States. Power consumption is going up, and we're building new coal plants to compensate. The *only* sane course of action (short of spending two years seeing if Robert Bussard is right) is to put some of our existing Fast Reactor designs into production and start burning that waste for power.
Since renewables can provide all the power we use and more at lower cost than a pie in the sky breeder program
Since breeders can provide all the power we use and more at lower cost than a pie in the sky renewables program...
I guess we'd need hard numbers on both sides to figure out who's blowing smoke where, but... the fact of the matter is that junky old nuclear power plants are providing 22% of the electricity used in the United States today and renewables (other than hydropower, which we've maxed out) don't even show up on the chart. If we don't build modern nuclear plants that can use all that "spent" fuel sitting around, what the hell else are we going to do with it? Dump it in a really expensive hole where it will stay dangerously radioactive for a hundred thousand years?
The reactor designs necessary to allow the use of a reasonable nuclear fuel cycle exist. All I want to see is for it to be legally easier to install a new modern nuclear reactor than it is to install yet another coal plant. Because, let's face it, in the time we've spent arguing another coal plant got built.
USB devices mostly Just Work under Linux. Device support for Cameras, Scanners, and Printers is so good that it's reasonably safe to go to the store, buy a random device, and expect it to work out of the box. In the rare case where something doesn't work (mostly cheap printers), you can return the thing and buy a different brand, and that'll work.
Slashdot Mirror
Apparently I'm bad at math. 2123 is blatantly wrong. I what I meant to say was 2043. That'd be really good, but there's no way photovoltaics would grow that fast.
First you say that people are going to put in new photoelectric arrays that get them power cheaper than off-grid. Then you say that people are going to use their power more efficiently. People buy SUVs and need to have their arm twisted to use obviously-superior compact fluorescent bulbs - there's no way in hell they're going to use energy more efficiently as it gets cheaper. Cost down, demand up - that's basic economics.
Electricity is naturally something that people will keep wanting to use more of. Currently, that is resulting in 25 gigawatts/year of new generating capacity being installed. That rate of increase is at least holding steady, but demand is probably accelerating.
You seem convinced that photovoltaics are going to overtake all existing centralized power generating technologies, and that we just have to wait for it to happen. How long will that take? Can we afford to dump that much more carbon into the atmosphere?
I can imagine no scenario that would make uptake fast enough to have a significant impact on global warming. Today the entire world only has 4 gigawatts of photovoltaic generation capacity. I'm sure that the industry can't produce that much capacity in a year, but let's say they can. And let's say their production capacity doubles insanely fast - every five years. And let's say all those new panels get sold in the USA. That means that in 2017 photovoltaic installations will be producing more new power than fossil fuels. In 2023 we can stop building new coal plants in the United States and start shutting down coal plants. Almost sounds good... but it's not until 2123 that we can close the last Coal plant... in the USA... while people in China are starting to notice that it really hurts when you go outside in the hard acid rain.
That's... not going to work. Unless you have superconducting long distance transmission lines. Which are free. And you don't. Storage might be a solvable problem - high temperature low price superconductors isn't something we'll see this century, at least not *that* low price.
You seem to be assuming some sort of radical new electricity storage technology that takes away the whole issue of "base load". Storing electricity is hard, to the point that some energy companies are resorting to shit like flywheels. Maybe ultracapacitors will come down in price or something, but that's a technology that today is where photovoltaics were in 1975. Until some sort of really cheap and efficient electricity storage technique exists, the "constant power output" property of nuclear power plants will remain an advantageous feature.
It'd be really nice if solar took a bite out of coal, but I think you're overestimating the adoption rate. Currently the entire install base of photovoltaics in the United States produces less power the new Coal plants coming online in 2007. Even if photovoltaics experience exponential growth until Coal is gone, we'll still be putting in new Coal plants for more than a decade. That's the problem I see, and the problem that everyone advocating renewables over nuclear seems to ignore - the growth rate of actual solar/wind/tidal power installations is far, far behind the growth rate of energy usage. The difference is, and will continue to be, new coal and natural gas plants - simply because modern nuclear plants are illegal.
Reference? This PBS interview isn't 100% clear, but it strongly implies that coolant was completely shut off to the plant: http://www.pbs.org/wgbh/pages/frontline/shows/reac tion/interviews/till.html.
Distributed power generation is definitely preferable, simply from a political perspective - decentralizing money makes lobbying less annoying. Unfortunately, I'm still not convinced that the technology really exists to compete economically with just letting the electric meter spin forwards.
Here's what you're forgetting: http://en.wikipedia.org/wiki/Image:Global_energy_u se_from_1980_to_2004.jpg. Nuclear fission could provide every watt of the power on that graph and more, but instead we're talking about solar and building more coal plants.
That was my point. I'm pretty sure they'd have to to cause a meltdown in an Sodium-cooled fast reactor.
If reactors like the IFR couldn't handle operation in the absence of coolant, there would have been a meltdown when they did a test and disconnected the entire cooling system when the reactor was running in the early 90's. What happens with that design is that when the fuel gets hotter than normal operating temperature, it expands. This decreases the number of neutrons emitted by volume below the level required to sustain a reaction, causing the reaction to stop.
Given that the reactor design will safely shut down in the absence of coolant, this is a simple question of math: If the sodium explodes as the result of any possible scenario, can it do enough damage to the reactor to interfere with safe shutdown? Sodium isn't C4. The coolant pipes aren't that thick. The claim that the Sodium wouldn't be able to damage the core enough to significantly change the material properties of the fuel mixture seems entirely reasonable.
It would be nice if gas-cooled or helium-cooled fast reactor designs were a bit more advanced. That would significantly reduce the amount of "OMG sodium blows up if it gets wet" clouding the issues related to nuclear power. Heck, it would be nice if we knew the Polywell worked - that would solve our power / nuclear waste problems. None of this wishful thinking changes the fact that we have a fully functional power plant design, Sodium-cooled Fast Reactors, available today that looks to be strictly better than any other large scale centralized power generation technology.
It looks quite a bit like the entire problem of cost is a side effect of hysteria over safety. If anyone were ever allowed to just take a good reactor design and build it rather than squabbling for years over the politics of nuclear safety, I'm sure the costs would be much more reasonable.
You apparently don't appreciate the chemistry and engineering involved. How these reactions work is pretty well understood, and this specific problem has been worked on for 40 years now. In most of the modern reactor designs, you couldn't get them to melt down if the engineers on site tried to make it happen. The fuel mixes and/or chamber geometry simply won't sustain a runaway reaction.
As for sodium as a coolant, they were running the IFR for years - including a test where they had the thing running at full bore and just went and disconnected the sodium cooling system. Even if someone dumped a bunch of water in the cooling system, it's not like a sodium fire produces anywhere near enough neutrons to sustain a sub-critical fission reaction.
Hysteria is absolutely the right word to describe at least some of the public demand for more safety in nuclear plants. Activists have been complaining about the dangers of nuclear power for 40+ years, and engineers have been responding by making new designs more foolproof. The thing is - the Activists either don't know about or discount those improvements (possibly because the actual plants in production are old junk and don't have many of the well-known design improvements) and continue to complain about safety issues that largely don't exist in modern designs.
The assumption that anti-nuclear activists seem to make is that nuclear accidents look *worse* than Chernobyl, that *any* nuclear plant is susceptible to drastic meltdown resulting in a containment breach and spreading high level radioactive waste around in a giant area. You seem to assume that all Nuclear Plant designs are the same, as when you dismissed my reference to the IFR (An early 1990's sodium-cooled fast reactor prototype with passive safety properties) with a reference to a 1959 meltdown at a reactor that happened to also be sodium-cooled. Are you really unwilling to consider the possibility that between 1959 and 1990 that nuclear engineers might have come up with better designs that avoid well-known problems?
That niche is well supplied by Ubuntu. If you think the space needs competition to keep things fresh, it has it: Linspire and Mandriva.
Debian *doesn't* need to target that niche at all. The existing policy of slow, steady progress and periodic rock-solid server releases produces a distro that's an excellent basis for projects ranging from Ubuntu to Embedded Debian to build from. There are occasionally cases where Debian is a bit slower than it should be (multiarch for example), but that doesn't mean that what Debian is should be thrown away to make an Ubuntu replica.
So, what you're saying is that - at worst - Ubuntu is as bad as Microsoft Windows. In order to fully use recently-released supported hardware, you need to download a driver package from the manufacturer and install it.
My guess is that this is your problem. It's great that you have the option of using whatever distro you want, but not every distro will provide the same level of user experience. Ubuntu provides binary blobs for wireless drivers. Other distros like Fedora do not. For Fedora, an Intel wireless card isn't supported hardware.
My suggestion: Use Ubuntu, it works. If that's too hard, use Linspire. But don't complain when the distro you chose doesn't work with the hardware you chose.
Copyrightists are approximately as scummy as your average politician. Mostly, they don't understand exactly what the economic effects of the policy they are pushing are compared to the alternatives, but they think they'll get big payouts themselves if they get it implemented.
Get him to buy an Ubuntu desktop support package for $240 from Canonical and say that it includes unlimited desktop installs.
I suggest telling Adobe and Microsoft to die in a fire and using Free Software that can be fully supported locally instead. These harsh copyright laws are doing you a favor - why let yourself get sucked in to the upgrade treadmill and culture-crushing poor local language support that is proprietary software?
What does "piracy" have to do with rape and murder? Pirates are *really cool* guys like Captain Jack Sparrow who try to look tough but are really good at heart. "Pirate" is also a somewhat amusing term for people who like to share data that other people are trying to sell.
Trying to fight against the terminology at this point is silly. It's much faster and easier to just accept the term and run with "Pirates are cool" - it's not like the term really has any negative connotation at all among people young enough to make a difference culturally.
Why? Isn't that like leather seats in a Geo Metro?
I suggest you actually look into nuclear fuel cycles in a bit more detail. When you actually look at the nuclear chemistry of the processes, it's pretty obvious that you put in nuclear fuel (including "spent fuel") and you get out energy and low level waste. The two most mentioned issues, cost and proliferation, are both seriously overhyped - it's not like you can make a nuclear bomb out of Plutonium 240 mixed with random Actinides.
You can't patch in security, nor do you get it simply by complying with some FIPS standard.
Linux is maintained optimizing for features. If you want support for every USB scanner and 15 different file systems, Linux has that. Security and stability are sacrificed, but not too much.
FreeBSD is maintained largely for performance. Security and features are sacrificed, but not too much. It probably won't support your digital camera or encrypted HFS-Formatted USB key.
OpenBSD is maintained for security. Performance and features are sacrificed, and you'll notice if you try to use it. Stability and security are synergistic, so OpenBSD is rock solid. For some applications (network appliances and hardened servers), stability and security are strictly more important than performance or features. In those cases, OpenBSD is the obvious choice.
If you are using hardware that should be supported and having problems, that's a bug. You should report it, and it will get fixed.
As for
It *should* work out of the box on Dells, except for "Dell Wireless" - which is illegal to provide out of the box support for because it is based on a Broadcom chipset. If you can manage to click "Intel Wireless" when you buy your Dell, there's no reason it shouldn't work.
More relevantly though... operating systems should really come pre-installed. Dell desktops *are* off-brand for Linux, because Dell doesn't support Linux on them. If you buy from someone like http://www.system76.com/ everything will Just Work, or at least Just Work better than Windows does.
Hey, if you're right and we can really make and install enough solar panels to put a dent in our coal usage, that'd be really nice. In that case, we'd still want to build some modern nuclear plants so that we can burn the "spent fuel" from all the PWRs, but there's really no hurry - that stuff's not going anywhere.
I'm still not convinced that the solar's coming fast enough though - I'm mostly seeing all the new coal plants that are getting built.
Chernobyl was an important lesson about fission plant design and safety. Since then, we've designed power plants that simply don't fail like that. Seriously - something like the IFR doesn't even damage itself (much less melt down) when you do ridiculous stuff like physically disconnecting the cooling system with the thing running.
Any industrial process is dangerous. Many of them involve dealing with highly toxic materials. With nuclear power, it's reasonable to accept the fact that if you eat the fuel rods or something you're going to die of radiation poisoning. But, as a power generation technology, modern nuclear fission is damn safe - at this point, far safer than it really needs to be because engineers have worked overtime trying to combat irrational hysteria with safety improvements.
When I say "major failure", I mean something worthy of media coverage even without the hysteria surrounding "radioactivity". You know, where the result of the accident was somewhat more drastic than "At least one of the engineers on-site at the time, Phil Harrigan, later died in his '50's of leukemia."
Even Three Mile Island, hyped as some sort of horrific disaster, caused so much damage that "statistically, someone will probably die of cancer" from it.
I bet more people die every year because they get hit by windmills than because of the US nuclear power industry.
One of the big problems with your argument is that "renewables" aren't really just one thing. If you look at them that way, you'd see "10% renewables" in the USA, which is strictly misleading. That's all hydropower. Your link about Germany mentions Wind and Hydropower.
There are places that are good for Wind. Great, we should use them. After that, you have "slightly worse" places, but the energy of wind varies the cube of the speed, so slightly slower wind means significantly less power - quickly falling to the point where its uneconomical. As for hydropower, there are only so many dammable rivers; in the USA, we've already dammed them all. And... we're still building more coal plants.
The point is, until I see a technology like photoelectric cells or ethanol-generating bacteria really take off, putting in modern nuclear plants really looks like the only thing that can actually slow down our carbon emissions. And new reactors are the only good answer to the nuclear waste issue.
The nuclear power industry in the USA isn't the greatest, sure. But it's actually pretty amazing that outdated junkers they're running haven't had a really major failure yet. Those plants are 40 years old, and they have no innate design protection against meltdown.
The point is, there have been 25 or 30 years worth of design improvements since those plants were built. We'd have gotten the full 40 years worth of development, but some damn protesters got the USA out of the game 15 years ago. Sure, there have been experimental failures getting to where we are today - but now we have reasonably good designs that support a useful fuel cycle and don't melt into a puddle when the control computer breaks.
We can't "end the production of nuclear waste". These old, junky PWRs are providing 22% of the power in the United States. Power consumption is going up, and we're building new coal plants to compensate. The *only* sane course of action (short of spending two years seeing if Robert Bussard is right) is to put some of our existing Fast Reactor designs into production and start burning that waste for power.
Since breeders can provide all the power we use and more at lower cost than a pie in the sky renewables program...
I guess we'd need hard numbers on both sides to figure out who's blowing smoke where, but... the fact of the matter is that junky old nuclear power plants are providing 22% of the electricity used in the United States today and renewables (other than hydropower, which we've maxed out) don't even show up on the chart. If we don't build modern nuclear plants that can use all that "spent" fuel sitting around, what the hell else are we going to do with it? Dump it in a really expensive hole where it will stay dangerously radioactive for a hundred thousand years?
The reactor designs necessary to allow the use of a reasonable nuclear fuel cycle exist. All I want to see is for it to be legally easier to install a new modern nuclear reactor than it is to install yet another coal plant. Because, let's face it, in the time we've spent arguing another coal plant got built.
But... I've already written this rant. Check out http://www.cs.uml.edu/~ntuck/nuclear/
USB devices mostly Just Work under Linux. Device support for Cameras, Scanners, and Printers is so good that it's reasonably safe to go to the store, buy a random device, and expect it to work out of the box. In the rare case where something doesn't work (mostly cheap printers), you can return the thing and buy a different brand, and that'll work.