I, personally, find it astonishing how much it's asserted that video drivers contain proprietary software, when video cards are basically the other CPU in a computer, and actually faster for many things.
Why the hell aren't they running this stuff out of fireware on their GPU? What exactly is going on here? Why do they need a 1 meg binary blob loaded into kernel space to manage a damn processor? Can't they have that processor manage video stuff?
Their 'serious advantage' is THE major cause of their operating system's most notable problem, aka blue screens.
That sounds like a fucking stupid 'advantage' to me.
When a Linux kernel oopses because the USB-to-USB networking cable driver is broken, who's fault is it? The Linux people, and they fix it. As soon as possible.
In the worse case, it might be due to someone emailing patches to you and telling you to try them out, because no one has exactly the problem, but by gum they'll fix it.
When it blue screens in XP? Well, I dunno, because the damn company that made the cable went out of business before XP and their driver from 2000 doesn't work. Possibly that was a bad example.
Pretending they still existed, and you could still find anyone who gives a damn about hardware they aren't selling anymore and it's under warranty, I'd lay pretty good odds they'd just tell to 'upgrade to the newest driver', and that would be the entire solution...if it didn't work, you're screwed.
This is because producing drivers is not their business, whereas it is the 'business' of Linux kernel developers.
Erm...what the hell do you think kernel development is? The kernel is almost all drivers. There are a few parts of the kernel that aren't, like process management subsystem, but, on the whole, about 90% of it the kernel is a driver for something or another.
Does that explain why my fucking signed Logitech wireless keyboard driver bluescreens XP when it loads itself at any point except boot, or ever unloads itself. (I.e., if I ever plug or unplug it with the computer on.)
And, no, I don't mean sometimes, I mean every single damn time. I can demonstrate bluescreens for fun and profit. This is a wireless USB keyboard/mouse called iTouch, the older model with the ball in the mouse. (Why the hell I need drivers for a USB keyboard and mouse, even a wireless one, is beyond me.)
Not only is that signed, my USB drivers are signed, as is everything up the chain. Admittedly, I have some drivers that aren't...but it did this before I installed them, and I doubt Alcohol's virtual CD-ROMs are causing this, or my USB wifi, or my TAP virtual network adapter. (As an aside...MS, please come up with some sort of standard 'virtual device' for CD/disk/network/whatever. This is actually one of the things an OS is supposed to be able to do, device management, as opposed to sending instant messages or whatever you've decided to include in the OS today.)
OTOH, if something ever goes horribly wrong with my computer and I can't unplug it or turn it off, I can always unplug the keyboard, and know my computer will crash in about three seconds.
I'm just saying, if you're not moving the racks around, you're not moving the power around, assuming you ran enough in the first place. You, at minimum, have to add a single cable for each new server.
Ergo, you can run overhead trays with the power in them, but have them up high, where you need a stepladder to reach them but they aren't in the way.
That's one of the most uninformed post I've ever heard here.
Why, if someone was investing in steganography, would they go and put, say, an image of a kitten with a secret message in it hidden in a directory of spaceships?
What they would do is download a whole bunch of the same kind of images, resize them all or something so none of them exactly match the original, and put stuff in a few of them. Possibly even use a tool that pulls data out of an entire directory, including virtual filenames, and puts new data in the best place it can find.
And you can't run probabilities against encrypted messages to see which are important without decoding them unless the sole criteria is size. As you can't even see the message size before finding the message in steganography, that doesn't seem incredibly useful.
Steganography isn't the 'security through obscurity' that you seem to think it is. (Althoguh all encryption is that it some extent.) No one uses it to encrypt data by itself. Steganography programs already encrypt the data, using well-accepted algorythms. They just then store that data intermixed into an existing lossy format so that you can't see it's encrypted. Just like any other method of encryption, the only secret is the key.
It's just harder to see you should even try to decrypt a certain file, thus making the amount of files you have to process a few orders magnitude bigger. It's like broadcasting random data on 999 channels and perfectly good encrypted data on one. Have fun wasting CPU cycles on the extra channels.
Most normal encryption programs already do something like this. I have one, called Truecrypt, that has no header information or any identifable parts of the encryped files. The could just be random data....and, if my computer already had a lot of completely random unidentifably files, that would be steganography. Steganography just takes the result and hides it inside other things.
And why the hell would your hypothetical encryption program put stuff in the most recently used list? Or registry?
I don't know what you mean by this, but it's not anymore true of steganography than anything else.
The key, in steganography, tells you where and how the data is hidden. The algorythm obviously determine where data is put and gotten from, but it uses the key to figure this out.
Otherwise, the police could obviously just grab the steganography tools and run them on all the stuff.
However, you should encrypt on top of steganography, because of the risk of the people doing the decoding being able to find an unmodified copy of the file, and just XORing them. This is the only way to 'detect its use'..by finding unmodified files.
Which is also why you should 'crappify' whatever you use. If it's a JPEG, open it up, change one pixel in the corner to black, and save it at a slightly higher compression. If it's an MP3, downcode it by 16bps. Make it so no one has exactly whatever you're sticking info into.
Including you. You need to treat the originals like you treat unencrypted images...keep them only in RAM, or wipe them when done.
And, heck, delete the stuff that you degraded them from, too. Otherwise, they could just use whatever tools you used, which are presumably still on your computer, and get exactly the same thing you did. By deleting, you'll make it hard for them to figure out which rip of 'One Week' you degraded to 160k before possibly hiding stuff in it, or if this is an original rip that's simply not that easy to find on P2P networks.
Or you can go in the other direction, and simply use things that no one else could have a copy of, like images from your digital camera.
'Doesn't work' in the sense of 'They'll stupidly give up after one level', or 'doesn't work' in the sense of 'They will continue to torture you for more passwords long after you've revealed all of them'?
A lot of the old rules went out the window when the US started torturing people, and 'plausible deniability' is one of them. Plausible deniability is a legal term, it doesn't not related to torturability. All that means now is that you continue to get tortured for codes after you tell them everything, because you cannot prove you gave them everything.
I'm not talking about email. I don't keep most email either, although the stuff I do keep does end up getting indexed by google desktop. I've been tempted to run that off, because frankly it's not useful...if I saved the email I put it somewhere I can find it.
I'm talking about things like the W3C's HTML reference or this page. I'm talking about stuff you'd bookmark and then spend a minute finding and then three minutes looking around on the site to find the specific bit of information you wanted.
The point of the method I spoke of is not to keep 'your' data. I agree, people who search for their own files need to be a bit better organized. It's to have a full-text search of a tiny, self-collected part of the internet, from whole websites to single sentences, and it will never be screwed up by search engines reranking things or pages going dead. All you have to do is say 'Hey, that's a useful page, File, Capture page' and continue on your way, and never worry abut it again, until you think 'Hey, what was that page that had...'.
Incidentally, Google desktop tries to index your browser cache for exactly this reason, but that is rather idiotic and doesn't work well as pages disappear from that but not the index. Turn that feature off.
And an added bonus, my bookmarking has gone down immensely, now only to pages with changing content, so now I can actually find things there, too.
As for the other stupid stuff in google desktop? Well, you can easily turn it off. Or you can find another tool that does the same thing for desktop searches, I'm sure they exist. The indexing tool is not important.
Don't use it for the sidebar. Generic global sidebars never have what you want.
I use it solely for the search.
But first, get this. It's an extension for firefox, it lets you save parts of web pages, whole web pages, and even spider sites.
Then change the directory for the scrapbook to something you'd be able to find normally. You need to do this in case you reinstall Firefox or make a new profile, because it stupidly defaults to an impossible-to-find subdirectory of your profile. I use My Documents/info/scrapbook/. It's just the downloaded files and some metadata files that have the URL and whatnot, so even if the extension goes away you still have the data.
Whenever you find any interesting data (Instead of interesting changing sites.), don't bookmark it. Instead, scrapbook it. The scrapbook keeps a record of the original site so you can get back there anyway, and, best of all, you can delete parts of the page you don't need, and just keep the actual info.
Instant. Fulltext. Search. Of every bit of information you've ever come across, and it's as simple as choose 'File/Capture Page', or hitelighting and choosing 'Capture Selection', and, when you want to search, you go to google and click 'Desktop'. (Or you use the taskbar thingy, or bookmark it, or add it as a firefox search, or whatever.)
You can, if you want, organize the scrapbook in such a way as that you can browse it, with folders and whatnot. Or you can just completely ignore the layout, never bothering with the actual sidebar, just doing searchs. The scrapbook extension that lets you track a saved page back to a URL, and edit the document, works off the location, so it works even when google desktop finds it.
I've already got 111 megs of data, and I've only been doing this two or three months. And I still need to go to my old 'saved web pages' directories from before I did this and see if I can track down those pages again and save them with this method.
That's exactly what I'm always worried about with raised floor cooling thingies. One tile ajar and air flies out. Or one little hole punched somewhere in the tile to let a wire through and you've got a leak, and if that leak's easier to get air through than the servers...
I've never quite understood what's wrong with laying a duct or two into the room and hooking them to the racks. If your racks expect bottom cooling, well, hook it to the bottom, under the floor. If not, it'd be good to figure out a way to split up the air so you can cool all servers equally. They probably already make racks that you can hook up ducts to.
Subfloor wiring and subfloor AC just don't mix. You can't keep fiddling around down there with wires and not screw up the airflow.
And I love subfloor wiring, so would rather keep that...but one of them has got to go.
This is why power should not be under the floor like that. Power should come down from the ceiling, because that is completely unrelated to any other kind of cabling and won't require rerouting. (Unless you move the racks, in which case, duh, you have to change all cabling, period.)
Everything else should be under the floor, except the incoming wires, which should come in on the wall, and end up under the floor only if they need to. (Normally they should just connect to a panel and make a short hop to a router.)
This is how I would design a server room if I could. This is not how I've ever seen a server room.
How the room should be cooled...I dunno. I've always distrusted the forced floor cooling. It's always seemed to me that completely falls apart if you have any panels up. But I am not an expert on cooling.
If you put X cubic feet of air per second into something at one end, you're damn well going to get X cubic feet of air per second out at the other end. (Or it's going to explode, or has a leak.)
All that can happen that you might need a larger fan to push the same amount of air, and it might exit the other end more spread out, instead of in a stream.
The bad thing is that PETA takes animals from no-kill animal shelters and kills them. Animals that were perfectly safe, and usually fairly happy, or as happy as they can be in an animal shelter.
PETA kills 85% of all animals it takes in. This is a horrible track record compared to almost any other shelter in existance. Even the most uncaring city pound manages to get rid of more than 15% of their animals before having to kill them. PETA kills animals in vans as they drive away from picking them up, without making the slightest effort to give them good homes.
2 PETA employees were arrested on 62 counts of felon animal cruelty in June, for killing animals without a permit and dumping their bodies in a dumpster. The unusual thing here is not the killing, it's the dumping in the dumpster, which admittedly is not normal PETA practice.
This killing is because PETA lives in a universe where all animals should be free, and animals that can't be free, like cows, and animals that don't want to be free, like dogs, should die or be killed instead of being 'slaves'. The killing of temporarily unwanted pets is just taking this philosophy to the logical place.
PETA is scum. Don't have anything to do with them.
1 billion is nothing. Power plants are expensive undertakings. Even non-nuclear. We're talking hundreds of millions to start, so an accident costing ten times that isn't absurd.
Thus, it's rather stupid taxpayers are paying to contain it. Power plant operators should be able to cover it.
What we actually need is a law limiting the liability to nuclear power plant operators, if, and only if, their waste output of radioactive materials is nothing. (And their should be no 'accidents'. Nuclear power plants design has advanced to the point that there should never be a meltdown again.)
In other words, if they dump radioactive water for decades, hell, yes, every person can sue their ass off who was anywhere near them and got cancer.
But they don't get to assert there's some magical radiation coming from the plant that made their vegetables taste funny, if the plant can prove it is not leaking.
That's the real reason they can't get insurance. Because throw the word 'nuclear' in there and damages triple, and nuclear plants don't want bad press, so they settle.
So the law merely needs to say 'If you follow these standards, you are not liable for any damages, period. The case will be summarily dismissed unless they can prove you didn't follow those standards. Even if there is a magical failure and you somehow take out half the state, you will not be liable if you followed the standards.'. (Not that there actually can be nuclear accidents with intelligent plant design anymore.)
And then have really tight standards. And enforce them. 1/4th the people at the plant should be safety inspectors operating under government authority.
Actually, Iraq did have yellowcake. About 500 tons of it.
Buried in sealed plastic containers from last time they tried to build an atomic bomb, back in 91. They were allowed to have this, before anyone gets any ideas, and we knew they had this and let them keep it.
Note that 500 tons of yellowcake isn't anywhere as impressive as it sounds. Yellowcake's the stuff you get uranium from. And after you get uranium, you have to enrich it. It takes a lot of yellowcake to make a nuke.
But while 500 tons may or may not have been enough to make a nuke, and there are arguments on both sides about whether or not the yellowcake they had could have made an atomic bomb, much less a 'program', it obviously would have been in Iraq's best interest to use their own stuff first. We found no indications they had done so. None of the certifuges required, none of the labs, none of the many complicated things that turning huge mountains of yellowcake into tiny qualities of uranium 232 requires. Much less any of the complicated things it takes to turn uranium 232 into a bomb.
Which makes the 'Iraq was trying to buy yellowcacke from Niger' even more preposterious. They had yellowcake. They weren't doing anything with it, but they had it. You don't go shopping for things that will raise suspicious when you already have them. You build the facilities and use what you have first to refine the process, and then you go shopping.
About the only thing that Iraq had that they officially weren't allowed to have was some missiles that could apparently go like 5% farther than the weapons range they were supposed to be restricted to, and that's probably because someone screwed up the math somewhere, not because of some secret invasion plot. (Not that Iraq could attack the US with these missiles under any circumstances, the missile restrictions were to keep them from attacking Kuwait.) If Iraq was going to delibrately break the rules they would have bought missiles that flew a lot farther and actually hid them, instead of showing them off to various people.
Slashdot Mirror
Analog VGA? As opposed to what?
Why the hell aren't they running this stuff out of fireware on their GPU? What exactly is going on here? Why do they need a 1 meg binary blob loaded into kernel space to manage a damn processor? Can't they have that processor manage video stuff?
That sounds like a fucking stupid 'advantage' to me.
When a Linux kernel oopses because the USB-to-USB networking cable driver is broken, who's fault is it? The Linux people, and they fix it. As soon as possible.
In the worse case, it might be due to someone emailing patches to you and telling you to try them out, because no one has exactly the problem, but by gum they'll fix it.
When it blue screens in XP? Well, I dunno, because the damn company that made the cable went out of business before XP and their driver from 2000 doesn't work. Possibly that was a bad example.
Pretending they still existed, and you could still find anyone who gives a damn about hardware they aren't selling anymore and it's under warranty, I'd lay pretty good odds they'd just tell to 'upgrade to the newest driver', and that would be the entire solution...if it didn't work, you're screwed.
This is because producing drivers is not their business, whereas it is the 'business' of Linux kernel developers.
Erm...what the hell do you think kernel development is? The kernel is almost all drivers. There are a few parts of the kernel that aren't, like process management subsystem, but, on the whole, about 90% of it the kernel is a driver for something or another.
And, no, I don't mean sometimes, I mean every single damn time. I can demonstrate bluescreens for fun and profit. This is a wireless USB keyboard/mouse called iTouch, the older model with the ball in the mouse. (Why the hell I need drivers for a USB keyboard and mouse, even a wireless one, is beyond me.)
Not only is that signed, my USB drivers are signed, as is everything up the chain. Admittedly, I have some drivers that aren't...but it did this before I installed them, and I doubt Alcohol's virtual CD-ROMs are causing this, or my USB wifi, or my TAP virtual network adapter. (As an aside...MS, please come up with some sort of standard 'virtual device' for CD/disk/network/whatever. This is actually one of the things an OS is supposed to be able to do, device management, as opposed to sending instant messages or whatever you've decided to include in the OS today.)
OTOH, if something ever goes horribly wrong with my computer and I can't unplug it or turn it off, I can always unplug the keyboard, and know my computer will crash in about three seconds.
I'm just saying, if you're not moving the racks around, you're not moving the power around, assuming you ran enough in the first place. You, at minimum, have to add a single cable for each new server.
Ergo, you can run overhead trays with the power in them, but have them up high, where you need a stepladder to reach them but they aren't in the way.
What are those?
Because the attackers are complete morons, I guess, and don't have the ability to detect the fact you've written to the disk.
Why, if someone was investing in steganography, would they go and put, say, an image of a kitten with a secret message in it hidden in a directory of spaceships?
What they would do is download a whole bunch of the same kind of images, resize them all or something so none of them exactly match the original, and put stuff in a few of them. Possibly even use a tool that pulls data out of an entire directory, including virtual filenames, and puts new data in the best place it can find.
And you can't run probabilities against encrypted messages to see which are important without decoding them unless the sole criteria is size. As you can't even see the message size before finding the message in steganography, that doesn't seem incredibly useful.
Steganography isn't the 'security through obscurity' that you seem to think it is. (Althoguh all encryption is that it some extent.) No one uses it to encrypt data by itself. Steganography programs already encrypt the data, using well-accepted algorythms. They just then store that data intermixed into an existing lossy format so that you can't see it's encrypted. Just like any other method of encryption, the only secret is the key.
It's just harder to see you should even try to decrypt a certain file, thus making the amount of files you have to process a few orders magnitude bigger. It's like broadcasting random data on 999 channels and perfectly good encrypted data on one. Have fun wasting CPU cycles on the extra channels.
Most normal encryption programs already do something like this. I have one, called Truecrypt, that has no header information or any identifable parts of the encryped files. The could just be random data....and, if my computer already had a lot of completely random unidentifably files, that would be steganography. Steganography just takes the result and hides it inside other things.
And why the hell would your hypothetical encryption program put stuff in the most recently used list? Or registry?
I don't know what you mean by this, but it's not anymore true of steganography than anything else.
The key, in steganography, tells you where and how the data is hidden. The algorythm obviously determine where data is put and gotten from, but it uses the key to figure this out.
Otherwise, the police could obviously just grab the steganography tools and run them on all the stuff.
However, you should encrypt on top of steganography, because of the risk of the people doing the decoding being able to find an unmodified copy of the file, and just XORing them. This is the only way to 'detect its use'..by finding unmodified files.
Which is also why you should 'crappify' whatever you use. If it's a JPEG, open it up, change one pixel in the corner to black, and save it at a slightly higher compression. If it's an MP3, downcode it by 16bps. Make it so no one has exactly whatever you're sticking info into.
Including you. You need to treat the originals like you treat unencrypted images...keep them only in RAM, or wipe them when done.
And, heck, delete the stuff that you degraded them from, too. Otherwise, they could just use whatever tools you used, which are presumably still on your computer, and get exactly the same thing you did. By deleting, you'll make it hard for them to figure out which rip of 'One Week' you degraded to 160k before possibly hiding stuff in it, or if this is an original rip that's simply not that easy to find on P2P networks.
Or you can go in the other direction, and simply use things that no one else could have a copy of, like images from your digital camera.
At which point they cut off two toes, restore the hard drive image they were playing with, and start over.
A lot of the old rules went out the window when the US started torturing people, and 'plausible deniability' is one of them. Plausible deniability is a legal term, it doesn't not related to torturability. All that means now is that you continue to get tortured for codes after you tell them everything, because you cannot prove you gave them everything.
Is that anything like the old charge of 'Being Willfully and Persistantly a Negro' in the US?
I'm talking about things like the W3C's HTML reference or this page. I'm talking about stuff you'd bookmark and then spend a minute finding and then three minutes looking around on the site to find the specific bit of information you wanted.
The point of the method I spoke of is not to keep 'your' data. I agree, people who search for their own files need to be a bit better organized. It's to have a full-text search of a tiny, self-collected part of the internet, from whole websites to single sentences, and it will never be screwed up by search engines reranking things or pages going dead. All you have to do is say 'Hey, that's a useful page, File, Capture page' and continue on your way, and never worry abut it again, until you think 'Hey, what was that page that had...'.
Incidentally, Google desktop tries to index your browser cache for exactly this reason, but that is rather idiotic and doesn't work well as pages disappear from that but not the index. Turn that feature off.
And an added bonus, my bookmarking has gone down immensely, now only to pages with changing content, so now I can actually find things there, too.
As for the other stupid stuff in google desktop? Well, you can easily turn it off. Or you can find another tool that does the same thing for desktop searches, I'm sure they exist. The indexing tool is not important.
I use it solely for the search.
But first, get this. It's an extension for firefox, it lets you save parts of web pages, whole web pages, and even spider sites.
Then change the directory for the scrapbook to something you'd be able to find normally. You need to do this in case you reinstall Firefox or make a new profile, because it stupidly defaults to an impossible-to-find subdirectory of your profile. I use My Documents/info/scrapbook/. It's just the downloaded files and some metadata files that have the URL and whatnot, so even if the extension goes away you still have the data.
Whenever you find any interesting data (Instead of interesting changing sites.), don't bookmark it. Instead, scrapbook it. The scrapbook keeps a record of the original site so you can get back there anyway, and, best of all, you can delete parts of the page you don't need, and just keep the actual info.
Instant. Fulltext. Search. Of every bit of information you've ever come across, and it's as simple as choose 'File/Capture Page', or hitelighting and choosing 'Capture Selection', and, when you want to search, you go to google and click 'Desktop'. (Or you use the taskbar thingy, or bookmark it, or add it as a firefox search, or whatever.)
You can, if you want, organize the scrapbook in such a way as that you can browse it, with folders and whatnot. Or you can just completely ignore the layout, never bothering with the actual sidebar, just doing searchs. The scrapbook extension that lets you track a saved page back to a URL, and edit the document, works off the location, so it works even when google desktop finds it.
I've already got 111 megs of data, and I've only been doing this two or three months. And I still need to go to my old 'saved web pages' directories from before I did this and see if I can track down those pages again and save them with this method.
However, that's not really relevant to the privacy policy, as it doesn't send that anywhere outside your computer.
I don't know where you've been, but an abacus's power supply operates a 98.6 degrees.
If not, what on earth was the point of that?
I've never quite understood what's wrong with laying a duct or two into the room and hooking them to the racks. If your racks expect bottom cooling, well, hook it to the bottom, under the floor. If not, it'd be good to figure out a way to split up the air so you can cool all servers equally. They probably already make racks that you can hook up ducts to.
Subfloor wiring and subfloor AC just don't mix. You can't keep fiddling around down there with wires and not screw up the airflow.
And I love subfloor wiring, so would rather keep that...but one of them has got to go.
Everything else should be under the floor, except the incoming wires, which should come in on the wall, and end up under the floor only if they need to. (Normally they should just connect to a panel and make a short hop to a router.)
This is how I would design a server room if I could. This is not how I've ever seen a server room.
How the room should be cooled...I dunno. I've always distrusted the forced floor cooling. It's always seemed to me that completely falls apart if you have any panels up. But I am not an expert on cooling.
If you put X cubic feet of air per second into something at one end, you're damn well going to get X cubic feet of air per second out at the other end. (Or it's going to explode, or has a leak.)
All that can happen that you might need a larger fan to push the same amount of air, and it might exit the other end more spread out, instead of in a stream.
The bad thing is that PETA takes animals from no-kill animal shelters and kills them. Animals that were perfectly safe, and usually fairly happy, or as happy as they can be in an animal shelter.
PETA kills 85% of all animals it takes in. This is a horrible track record compared to almost any other shelter in existance. Even the most uncaring city pound manages to get rid of more than 15% of their animals before having to kill them. PETA kills animals in vans as they drive away from picking them up, without making the slightest effort to give them good homes.
2 PETA employees were arrested on 62 counts of felon animal cruelty in June, for killing animals without a permit and dumping their bodies in a dumpster. The unusual thing here is not the killing, it's the dumping in the dumpster, which admittedly is not normal PETA practice.
This killing is because PETA lives in a universe where all animals should be free, and animals that can't be free, like cows, and animals that don't want to be free, like dogs, should die or be killed instead of being 'slaves'. The killing of temporarily unwanted pets is just taking this philosophy to the logical place.
PETA is scum. Don't have anything to do with them.
And, yes, this is completely offtopic. Bite me.
Thus, it's rather stupid taxpayers are paying to contain it. Power plant operators should be able to cover it.
What we actually need is a law limiting the liability to nuclear power plant operators, if, and only if, their waste output of radioactive materials is nothing. (And their should be no 'accidents'. Nuclear power plants design has advanced to the point that there should never be a meltdown again.)
In other words, if they dump radioactive water for decades, hell, yes, every person can sue their ass off who was anywhere near them and got cancer.
But they don't get to assert there's some magical radiation coming from the plant that made their vegetables taste funny, if the plant can prove it is not leaking.
That's the real reason they can't get insurance. Because throw the word 'nuclear' in there and damages triple, and nuclear plants don't want bad press, so they settle.
So the law merely needs to say 'If you follow these standards, you are not liable for any damages, period. The case will be summarily dismissed unless they can prove you didn't follow those standards. Even if there is a magical failure and you somehow take out half the state, you will not be liable if you followed the standards.'. (Not that there actually can be nuclear accidents with intelligent plant design anymore.)
And then have really tight standards. And enforce them. 1/4th the people at the plant should be safety inspectors operating under government authority.
What you didn't mention is that this was because the White House delibrately distorted evidence and conflated Iraq and 9-11.
Buried in sealed plastic containers from last time they tried to build an atomic bomb, back in 91. They were allowed to have this, before anyone gets any ideas, and we knew they had this and let them keep it.
Note that 500 tons of yellowcake isn't anywhere as impressive as it sounds. Yellowcake's the stuff you get uranium from. And after you get uranium, you have to enrich it. It takes a lot of yellowcake to make a nuke.
But while 500 tons may or may not have been enough to make a nuke, and there are arguments on both sides about whether or not the yellowcake they had could have made an atomic bomb, much less a 'program', it obviously would have been in Iraq's best interest to use their own stuff first. We found no indications they had done so. None of the certifuges required, none of the labs, none of the many complicated things that turning huge mountains of yellowcake into tiny qualities of uranium 232 requires. Much less any of the complicated things it takes to turn uranium 232 into a bomb.
Which makes the 'Iraq was trying to buy yellowcacke from Niger' even more preposterious. They had yellowcake. They weren't doing anything with it, but they had it. You don't go shopping for things that will raise suspicious when you already have them. You build the facilities and use what you have first to refine the process, and then you go shopping.
About the only thing that Iraq had that they officially weren't allowed to have was some missiles that could apparently go like 5% farther than the weapons range they were supposed to be restricted to, and that's probably because someone screwed up the math somewhere, not because of some secret invasion plot. (Not that Iraq could attack the US with these missiles under any circumstances, the missile restrictions were to keep them from attacking Kuwait.) If Iraq was going to delibrately break the rules they would have bought missiles that flew a lot farther and actually hid them, instead of showing them off to various people.