Decommissioning any industrial site is a nightmare. Nuclear power isn't special in that respect. Perhaps we should just be more reluctant to close these facilities.
Bethlehem Steel in Buffalo, NY shut down in the 1970s. Here is what it looks like today. 40 years later, and it's still a barren wasteland.
We want a decentralized energy production to become independent from big energy companies and to produce the energy more safely
Without realizing it, you've stuck upon the real psychological motivation behind the "decentralized everything" movement: it's political. It's a reflective reaction against the complexity of modern society, and against globalization.
Every honest intellectual person knows that sometimes centralization is desirable. Centralization is cheaper, more efficient, and often cleaner and safer as well. It's a lot cheaper for one building on campus to generate steam than for shack to have its own heater. It's easier to scrub the output of 100 coal plants than that of 10,000 automobiles.
Yet there are otherwise-intelligent people arguing for community-run, small, decentralized infrastructure even where it's batshit insane, like for nuclear power plants. This is not the product of honest reasoning, but an expression to live out the fantasy of living in a commune in the woods.
You want to stem the power of large corporations? I'm with you. Regulate them. But sometimes scaling up an operation is a no-brainer.
The attitude that small is always beautiful is the product of a small mind.
Sure, there's a chance of failure in every system, but good design can reduce it to an acceptable level. There's chance in everything: you could walk outside and be struck dead my a freak meteor.
As for the Titantic: how many passenger liner disasters have there been since her sinking?
Nuclear waste is hardly dangerous for millennia. Nuclear physics are such that a substance is either very radioactive for a short time, or slightly radioactive for a long time. After a few years in a cooling tank, the wastes will be of the latter form. They can then be vitrified (turned into glass) and put in a cave somewhere, where they'll be of no harm to anyone.
Yes, even low-radioactivity substances are still toxic heavy metals. But we know how to deal with heavy metals. If you don't agree, why aren't you up in arms about all industrial waste instead of focusing on the nuclear industry's?
That sounds like a fiendishly difficult problem. What materials are you using for it? It seems like you'd have to basically use ceramics and... tungsten.
there is a chance that some [radioactivity] will leak out.
Err, yes. Why didn't I realize that before? You've really opened my eyes. Radioactivity can get into the environment! OH MY GOD! LET'S BAN SMOKE DETECTORS. THEY CONTAIN TEH RADIOACTIVITY.
And how many genuinely foolproof and fail-safe machines do you use every day without noticing, because they work so well?
We can build nuclear reactors that are safe, and we don't need thorium to do it. We can build inherently safe nuclear reactors today using a variety of techniques. (See "void coefficient".)
But like I said above, if using thorium leads to new public acceptance of nuclear power, it's a win regardless of its technical merits.
On the one hand, modern uranium reactors (pebble bed, or even well-made light water reactors) are perfectly safe. Using thorium instead is at best a minor improvement.
On the other hand, if using a different fuel convinces members of the general public that nuclear power is safe, and allows the construction of new facilities in less than a decade, that's great, and worth it even if thorium is slightly inferior as a fuel. In short, it can be a PR win.
Markets happen whether they're intended or not. They're as natural as water flowing downhill, even in ostensibly destructive fields. Capitalism is not more a "choice" than gravity is: what matters is how you deal with it.
Clearly, we don't have enough incentives to either 1) discourage these people from writing malware, or 2) encouraging them to do other things.
Phase change memory is nothing like A CD-R. This stuff has the density of a hard drive, and the speed is very close to DRAM. It's non-volatile to boot. It's a serious contender to become universal memory.
Imagine how different operating systems and programs would be if we could make RAM non-volatile.
I hope you'll forgive me for using this last comment to illustrate why you're exactly the kind of person who should be given minimal privileges.
And I hope you'll forgive me for using your reply to illustrate why I would never want to work in an environment you administer. Programmers not cogs in a machine, and resent being treated as such. They chafe and strain against restrictions imposed by people who don't understand neither the problem domain nor the solution space of software development.
While you are correct in noting that individual choice and experimentation make an environment more akin to a research lab than a factory, but I would argue that every software development hop needs some elements of a research lab.
After all, programming is research and development. It's an inherently creative process, and nothing that can or should be mechanized like some garment factory. By allowing programmers to tailor their environment to their particular needs, you account for intrinsic individual differences and make everyone happier and more productive. It's not about using a workstation as a toy, but rather about finding the best way of parlaying my mental faculties into products for the company.
In short, management can dictate my job, yes: but they shouldn't dictate how I do it.
Regarding commonality and communication among software developers: good developers inform their associates about the techniques they use, and bad developers don't. Company policy should help this process. If you, as a manager, suddenly lose a skilled but uncommunicative developer, finding the programs he used to hack on the code will be least of your problems.
Are you trying to do a 360 on that claim and are you now going to another extreme and say that any administrator or support persons should also know their entire supported OSs inside out and also be able to do their task of administration and support?
Where do you get that idea? Making the source available as an aide to problem-solving is far different from demanding a complete and thorough knowledge of the entire codebase.
Actually, I do precisely that, and test via VMWare.
mingw works very well as a cross-compiler. Fedora even has ready-to-go RPM packages to get you set up. (And the packages have been backported to RHEL5 too.)
If I understand you correctly, your central assertion is that developers aren't paranoid enough. There are plenty of those people out there: but if someone doesn't take care to research the actual operation of HKLM\DoWhatIThinkLooksRight, then he's probably the kind of person to use an unchecked strcpy too, or to admit an XSS vulnerability to a website. You really don't want that kind of person working for you in any case.
Granted, though, it's difficult to screen out that kind of carelessness in advance.
As an aside, having done both systems administration and software development work, I love open-source operating systems: I can see exactly where the "evil spirits" are and step through their machinations with a debugger. With Windows, I'm reduced to sacrificing chickens.
Any developer who can't competently administer his own machine is incompetent. The kind of rigorous thinking required is identical. I'd be highly reluctant to work at a place that didn't let me install and manage the software packages I needed to do my job. I've used hundreds of small programs to help me in my work, along with kernel debuggers and other tools that require administrative privileges. Having to ask for approval and installation assistance for each of them would have been impractical.
If you're worried about developers screwing up their boxes, why aren't you more worried about these developers screwing up the their code?!
Yes, KASUMI has a 128-bit key. If it had a 64-bit key, that would clearly be a disaster.
But having a 64-bit block size is still problematic: a 64-bit block size is small enough to lead to practical birthday attacks in some applications, and increases the possibility of a short loop in OFB mode.
Slashdot Mirror
Decommissioning any industrial site is a nightmare. Nuclear power isn't special in that respect. Perhaps we should just be more reluctant to close these facilities.
Bethlehem Steel in Buffalo, NY shut down in the 1970s. Here is what it looks like today. 40 years later, and it's still a barren wasteland.
No nuclear power involved there.
Hey! Guess what? Everything is finite. What do you think you build solar panels and wind turbines from, pot smoke?
Without realizing it, you've stuck upon the real psychological motivation behind the "decentralized everything" movement: it's political. It's a reflective reaction against the complexity of modern society, and against globalization.
Every honest intellectual person knows that sometimes centralization is desirable. Centralization is cheaper, more efficient, and often cleaner and safer as well. It's a lot cheaper for one building on campus to generate steam than for shack to have its own heater. It's easier to scrub the output of 100 coal plants than that of 10,000 automobiles.
Yet there are otherwise-intelligent people arguing for community-run, small, decentralized infrastructure even where it's batshit insane, like for nuclear power plants. This is not the product of honest reasoning, but an expression to live out the fantasy of living in a commune in the woods.
You want to stem the power of large corporations? I'm with you. Regulate them. But sometimes scaling up an operation is a no-brainer.
The attitude that small is always beautiful is the product of a small mind.
Sure, there's a chance of failure in every system, but good design can reduce it to an acceptable level. There's chance in everything: you could walk outside and be struck dead my a freak meteor.
As for the Titantic: how many passenger liner disasters have there been since her sinking?
Nuclear waste is hardly dangerous for millennia. Nuclear physics are such that a substance is either very radioactive for a short time, or slightly radioactive for a long time. After a few years in a cooling tank, the wastes will be of the latter form. They can then be vitrified (turned into glass) and put in a cave somewhere, where they'll be of no harm to anyone.
Yes, even low-radioactivity substances are still toxic heavy metals. But we know how to deal with heavy metals. If you don't agree, why aren't you up in arms about all industrial waste instead of focusing on the nuclear industry's?
Not true.
That sounds like a fiendishly difficult problem. What materials are you using for it? It seems like you'd have to basically use ceramics and... tungsten.
Err, yes. Why didn't I realize that before? You've really opened my eyes. Radioactivity can get into the environment! OH MY GOD! LET'S BAN SMOKE DETECTORS. THEY CONTAIN TEH RADIOACTIVITY.
And how many genuinely foolproof and fail-safe machines do you use every day without noticing, because they work so well?
We can build nuclear reactors that are safe, and we don't need thorium to do it. We can build inherently safe nuclear reactors today using a variety of techniques. (See "void coefficient".)
But like I said above, if using thorium leads to new public acceptance of nuclear power, it's a win regardless of its technical merits.
On the one hand, modern uranium reactors (pebble bed, or even well-made light water reactors) are perfectly safe. Using thorium instead is at best a minor improvement.
On the other hand, if using a different fuel convinces members of the general public that nuclear power is safe, and allows the construction of new facilities in less than a decade, that's great, and worth it even if thorium is slightly inferior as a fuel. In short, it can be a PR win.
Markets happen whether they're intended or not. They're as natural as water flowing downhill, even in ostensibly destructive fields. Capitalism is not more a "choice" than gravity is: what matters is how you deal with it.
Clearly, we don't have enough incentives to either 1) discourage these people from writing malware, or 2) encouraging them to do other things.
Speak for yourself. Water in New York is great.
Phase change memory is nothing like A CD-R. This stuff has the density of a hard drive, and the speed is very close to DRAM. It's non-volatile to boot. It's a serious contender to become universal memory.
Imagine how different operating systems and programs would be if we could make RAM non-volatile.
And I hope you'll forgive me for using your reply to illustrate why I would never want to work in an environment you administer. Programmers not cogs in a machine, and resent being treated as such. They chafe and strain against restrictions imposed by people who don't understand neither the problem domain nor the solution space of software development.
While you are correct in noting that individual choice and experimentation make an environment more akin to a research lab than a factory, but I would argue that every software development hop needs some elements of a research lab.
After all, programming is research and development. It's an inherently creative process, and nothing that can or should be mechanized like some garment factory. By allowing programmers to tailor their environment to their particular needs, you account for intrinsic individual differences and make everyone happier and more productive. It's not about using a workstation as a toy, but rather about finding the best way of parlaying my mental faculties into products for the company.
In short, management can dictate my job, yes: but they shouldn't dictate how I do it.
Regarding commonality and communication among software developers: good developers inform their associates about the techniques they use, and bad developers don't. Company policy should help this process. If you, as a manager, suddenly lose a skilled but uncommunicative developer, finding the programs he used to hack on the code will be least of your problems.
Where do you get that idea? Making the source available as an aide to problem-solving is far different from demanding a complete and thorough knowledge of the entire codebase.
Not necessarily. You could just as easily test on a company-managed machines used remotely (say, via VNC).
Actually, I do precisely that, and test via VMWare.
mingw works very well as a cross-compiler. Fedora even has ready-to-go RPM packages to get you set up. (And the packages have been backported to RHEL5 too.)
If I understand you correctly, your central assertion is that developers aren't paranoid enough. There are plenty of those people out there: but if someone doesn't take care to research the actual operation of HKLM\DoWhatIThinkLooksRight, then he's probably the kind of person to use an unchecked strcpy too, or to admit an XSS vulnerability to a website. You really don't want that kind of person working for you in any case.
Granted, though, it's difficult to screen out that kind of carelessness in advance.
As an aside, having done both systems administration and software development work, I love open-source operating systems: I can see exactly where the "evil spirits" are and step through their machinations with a debugger. With Windows, I'm reduced to sacrificing chickens.
Gah! My eyes! My pre-caffeine grammar!
Any developer who can't competently administer his own machine is incompetent. The kind of rigorous thinking required is identical. I'd be highly reluctant to work at a place that didn't let me install and manage the software packages I needed to do my job. I've used hundreds of small programs to help me in my work, along with kernel debuggers and other tools that require administrative privileges. Having to ask for approval and installation assistance for each of them would have been impractical.
If you're worried about developers screwing up their boxes, why aren't you more worried about these developers screwing up the their code?!
Then you are merely working from your conclusion to your premises.
Point taken. In general, people should use entire off-the-shelf cryptosystems when they're available and applicable.
Yes, KASUMI has a 128-bit key. If it had a 64-bit key, that would clearly be a disaster.
But having a 64-bit block size is still problematic: a 64-bit block size is small enough to lead to practical birthday attacks in some applications, and increases the possibility of a short loop in OFB mode.
As another poster mentioned, the government can already get a wiretap easily enough without having to break the cipher.
I am sick and tired of conspiracy theories. Remember the sage advice to never attribute to malice what can be adequately explained by incompetence.
The other hemisphere? But there be dragons on them antipodes. :)