Yeah, watching nudist beaches is indeed more specialized, but I dispute the claim that this makes them less useful to scientists. If anything, geeks need devices like that more.
Last I heard, the Swedes were working on a Stealth Frigate, a ship effectively invisible to radar and sonar (carbon fibre hull, ultra-dark materials, non-ferrous engines, that sort of thing). Haven't heard anything on that since. If they actually developed such a vessel, then between that and Google Earth, I'm not sure they'd need any spy planes.
The underlying problem, though, is that you can never know if something is secret, you can only know if something is not secret. Thus, you have a paradox - the only way to know if something was secret is to share it and see if anyone else already knew.
As such, any system based on secrets of any kind whatsoever is inherently flawed because it is dependent on an assumption that is provably unprovable.
This is why you will see the phrase "security through obscurity is no security at all".
The catch is that published crypto systems rely on a problem being irreducably hard (exponential or hyperexponential difficulty). There are very few hard problems for which there is a non-existance proof of a solution simpler than the best solution hitherto known. This means that opponents could be keeping secret a solution to the problem you do not possess. However, they too are subject to the secrets paradox and therefore their ability to crack the cypher in better-than-expected time may be known to others besides themselves.
(This has been used in wartime, when Allies and the Axis powers deliberately sent messages in encrypted forms that they knew the other side could break in order to confuse them.)
The first of the removed paragraphs could be considered "original research" (banned on Wikipedia). I'm of the opinion that linear deductions are not research, but automatically follow. However, I've had a few entries edited out as "original research" myself and know that Wikipedia takes the rule extremely seriously even if it is to the point of absurdity.
The rest of the paragraphs are more inflamatory/op-ed and don't belong in an encyclopedia setting. They may be technically correct (only RSA knows) but they are most certainly not neutral POV and not useful in understanding the event.
I'd interpret it as "our firewall AV isn't stopping it", which is fine because AV software isn't a generic solution but one that detects specific, well-defined viruses. And when you shove it onto a firewall, it can't do much checking if you don't want horrible packet loss.
What it does mean, though, is that whoever wrote the article doesn't use NIDS or HIDS (the former will detect cross-scripting attacks, the latter will detect changes to files that aren't supposed to change) but relies entirely on anti-virus software on a (probably) mis-configured firewall that (likely) is running obsolete software.
Anyways, as said before, there's plenty of guides (including by the NSA) on how to not suffer cross-scripting attacks. That anyone still suffers from them is not through a lack of resources.
Well, not entirely sure how much precision is required. Each reactor has its own signature radiochemistry (which is why you can identify which reactor produced any given sample of material). In turn, you know the chemistry of the radionucleotides, the daughter radionucleotides, the containment vessel and the concrete. Any given reaction is impossible below the energy required to initiate it or above the energy at which the reaction produces the lowest-energy state. You only need to know if the energy is inside the range for that reaction, since you don't care why something won't happen.
From that, it's certainly possible to model different scenarios on a decent computer. (I'd consider the Earth Simulator pretty decent and I'd expect TEPCO can get time on it.) You then look for the scenario that matches the chemistry as observed from air pollution and water pollution. Since you know what chemistry HAS occurred, you can infer a lot about what it would take to have made it occur in the first place. Since you know the ratios of radionucleotides, you can infer something about the rate of reaction and therefore the mix of fuel and contaminants (and therefore the degree to which meltdown would have been required in order for the contamination to have reached that point).
So there's a lot you can do by indirect observation and deduction.
Let's put this in context. My A-Level computing project was an expert system to identify the most probable radionucleotides present in any given sample by looking at the energies detected, adjusting probabilities based on the probability of the presence or absence of associated daughter radionucleotides given the age of the sample. (Since iterating this could change whether something was considered detected or not, each iteration could branch off multiple ways. It was an interesting study in herustics.) It was calibrated using Chernobyl fallout and tested using historical samples collected from other incidents and achieved a success rate comparable with any of the commercial systems at the time and superior to the software BNFL was using for the same purpose. (I happen to know this because the BNFL software was used in a court case regarding leukemia deaths in Seascale and the raw data from Seascale was available to me. To be fair, BNFL's software was listed as version 0.8, so was likely not even out of beta testing and may well have been buggy.)
Now, what do you learn from this? Well, (a) I've studied the subject well enough to match the nuclear industry experts on analysis when I was 18, and (b) I had the means to actually obtain the raw data, the actual lab analysis of what was present in the samples and enough books on radiochemistry to write my own solution from the ground up. In other words, not exactly on the outside.
It hasn't kept it cool - the temperatures rose much more sharply after spraying began and the partial meltdown is likely more a product of the panic reaction than anything. The use of seawater, initiallly, also will have complicated things. Far too many minerals will now be baked onto the sides of things, it's far too corrosive, and it's generally really not well-suited to this kind of stuff.
The Soviet core was not on fire. Part of the Soviet core was in the stratosphere, the other part was melting its way through the concrete. There was no fire. The Soviet response was, in essence, to seal up the molten core. Sensible, but diluting it first (anything that is liquid at that temperature and absorbs neutrons - to any extent - would work) would have been better.
The only serious core fire that has ever happened was at Windscale, and that was England. Made a real mess of the local area, that did, with the soil contamination being a major problem.
Why on earth would you need to upgrade the nuclear portion of the plant? Ten years is plenty of time to build watertight doors for the emergency generator rooms (whose flooding caused many subsequent issues) and watertight doors leading into the occupied portions of the facilities (cut down on water pooling).
They most certainly do not have things under control - the head of TEPCO is hospitalized due to stress, the leaks haven't been located (never mind closed), the groundwater levels are far higher than anticipated and the engineers there are concerned that the partial meltdown may progress to a total meltdown that breaks out of the containment.
If that's your idea of control, I'd hate to see your idea of a breakdown.
As for understanding what's going on, I probably have a better grasp of nuclear reactor systems, their technology and their history than anyone else on Slashdot.
I take it that is sarcasm. Not very good at it, are you?
Nuclear power is perfectly safe, if done properly. So is coal mining. Both become extremely dangerous when not done properly - nuclear power due to the risk of reactions getting too hot (decay causes heat, accelerated decay through neutron emissions causes a lot more heat), which can lead to a failure of the structural integrity of the system or - worse - uncontrolled chemical reactions resulting in a chemical explosion (essentially a "dirty bomb"), coal mining due to the risk of coal gas (methane) igniting, resulting in an explosion and/or an uncontrolled fire within the coal seam itself, beyond the confines of the mine.
In both cases, the problem is not so much the initial event (although nobody likes fatalities - unless they're newsreaders or Fox TV presenters), the problem is that event zero can lead to the problem spreading in a way that cannot be controlled or stopped.
In both cases, competent design and competent management can make the probability of event zero happening at all virtually zero. The number of nuclear reactors worldwide is extremely high, but other than the Windscale core fire, Three Mile Island, Chernobyl and the Fukushima complex, there really hasn't been any major accident in the industry in 50 years. That's not bad, given that our knowledge of the physics is the same age.
(The Windscale core fire was an interesting piece of history. A graphite nuclear reactor core was allowed to burn for 3 days before anyone even thought to check why the temperature gagues were showing excessive readings. This was not corruption or greed, and even calling it incompetence is a stretch as maintenance was due over that time.)
The fact of the matter is this: in ALL of these accidents, there was a VERY long chain of events from the the initial point that turned towards disaster and the disaster actually happening. ANY person along that chain COULD have broken that chain at ANY time. They failed to do so.
Typically in disasters, this is because the difference between what they should have done and what they did do was so small that the person disregarded the difference. A very large sum of very small deltas will eventually add up.
(Even the Titanic's sinking was not due to a single person's failure or a single event, but a laundry list of very tiny deltas from the time the iron was first processed to the time the helmsman mistook what sterring order the helm was set to. R101, likewise, failed because of an incredibly large number of people making incredibly tiny errors.)
Yours is a common, and pitiful, belief that criticism of a sequence means criticism of the entire world in which that sequence lives. I've noticed such a mindset most amongst the right-wing and the libertarians. There is, sadly, no known cure and they are doomed to live in a world that really doesn't exist outside of their own interpretation.
I'm not going to repeat the BBC, the Guardian, or indeed the scientists who have published papers on the flaws in the design and positioning of the reactors - back in 2001. I'm also not going to repeat classical demonstrations of combusion (taught to me by the superb Dr John Salthouse of the University of Manchester), though I've given you one of the basic experiments demonstrating what essentially happened in a prior posting.
The workers aren't idiots, they're Japanese - taught to obey orders without question, taught that honor is worth more than life, and exposed to a situation where if they didn't damn well go in there (regardless of the inept management) a lot more people could die. You can look up the reports on the needless and senseless radiation burns they've suffered, the IAEA commentary, the level of badging - this is all public knowledge. But, no, you'd rather question someone who has the integrity to not AC, the age and experience to be a 4UID, and the knowledge of the basic chemistry and physics involved (and it isn't hard). Questioning irrationally to cover up your own fears is something that deserves scorn, not respect. Maybe that is exactly why you posted as AC.
The papers were published in 2001, so we have already seen. The accident was NOT stopped and is still continuing. And, FYI, I seem to have had better chemistry lecturers since you are the one guessing and I'm the one using experimental data.
A paper presented in 2001, I believe, demonstrated by means of sand deposits that tsunamis in Fukushima have historically been around 21 meters high and have typically travelled about 4 kilometers inland. TEPCO's assessment ignored all data before 1898, which meant it ignored essentially all tsunamis.
You don't need power. Hydrogen will not combust outside the presence of oxygen and a sufficiently powerful jet of hydrogen (such as superheated hydrogen from inside a compact concrete structure) would happily burn in the air outside with no possibility of the flame descending inside even if there was any oxygen present. The best thing they could have done would have been to open up the vents that they EVENTUALLY used to their maximum as early as possible, then ignite the hydrogen jet coming out.
(You want to burn the hydrogen rather than just vent it, as you really want some measure of control.)
If you happen to have a hydrogen cylinder handy, do the following experiment. If not, look it up on YouTube. Take a tin can and drill two holes in it. One in the lid, one in the base. Connect the hydrogen cylinder to the base and open the valve. Light the hydrogen stream at the top of the can. The hydrogen will burn cleanly and brightly. And, most importantly, in an EXTREMELY stable and safe way. Now remove the hose connecting the cylinder to the can. You will observe the flame start to descend into the can. What you will NOT observe is the air being pulled in at the base. When the flame enters the can, the mix of oxygen and hydrogen inside will ignite and the top of the can will explode.
It's hard to say for certain, but I would be willing to bet that something analogous happened inside the three concrete structures that exploded. Air must have been pulled in, as the zirconium reacted with all the oxygen from the water (which is why there was hydrogen there in the first place). My guess would be that the vents being opened and closed the way they were would be suspect as the three vented structures were the three that exploded. There must also have been some point of ignition. The heat is of no interest - if that had caused combustion, it would have burned the oxygen at the rate it was being pulled in, resulting in a slow burn. That ignition must have taken place after hydrogen and oxygen mixed in significant quantity.
If the vents had been left open, there would have been no significant hydrogen left, the ignition would likely not have detonated anything of significance, and we wouldn't have an international nuclear catastrophe.
The first problem is that TEPCO isn't telling anyone what they know (to save face and because they're freaking out) The second problem is that whatever they are telling, they're telling to the Japanese government and no-one else (even their own workers, who they convinced to wade through radioactive water without boots, go into radioactive buildings without radiation badges or suitable gear, etc). The third problem is that the experts are working with minimal data - and what they do have is suspect The fourth problem is that TEPCO has been trying to salvage the reactors at the same time as spraying them with seawater (which would be corrosive) and after the outer shell had exploded on three of them (causing untold damage to electronics, shock-proofing, etc)
On top of all that, TEPCO allowed the hydrogen build-up in the first place. They could have burned it off with a controlled burn. This would have prevented the explosions, reduced the spillage and possibly prevented the fuel leak. (Reducing pressure may have reduced water temperature and may have conserved some of the cooling pools.)
As for building the reactors ALONG the fault-line, despite advice not to by their own chief scientists, and building a tsunami wall far lower than the historic tsunami wave-heights....
This accident was stoppable at so many points in so many ways. The problem wasn't so much the reactor alone as the mindset together with the reactor.
I suggested to the Blender folks that they might want to look at elastons, as animation is a key area for Blender but they didn't have many ideas for how to improve it. Got no response and they didn't update their page, but I still think it'd be a worthy extension that could very well be helpful in their movie series (as rigid models only go so far).
Trying to emulate Windows will only ever produce a second-rate Windows (because the original will always be ahead of you by definition). That can't really let Linux become mainstream, because nobody is going to want second-rate. The only way Linux can become acceptable on the mainstream is to do things differently. It's why Apple still exists at all - if MacOS or OS/X were mere clones of Windows, Apple would be long-dead. They hold a small but very respectable market share because they solve things differently. They've only lost out in those areas they tried to copy Microsoft - if they'd kept with solving desktop problems distinctly but usably, they'd probably be doing better today.
I agree, however, that you'll only go so far with altering the underlying code. "Mucking about with code" is why there are Linux variants that are FAA-rated and others that are Carrier Grade, demonstrating incredible robustness. I'd consider that a high achievement. Users probably don't care for such certifications, but they certainly do care that their system doesn't crash when they're 5 points off the high score or half-way through a term paper they've stupidly not been saving. What it can NEVER do, and this is - I think - the point you're making, is make the software any more usable, any more learnable or any more efficient for getting things done.
My contention is that the front-end part, the user part, can't be solved by throwing people at it until a genuinely new way of approaching this whole "user interface" problem exists. The desktop metaphor has been done to death by Microsoft and Apple, and most of the recent advances have moved away from that entire concept. If Linux wants to be mainstream, it has to anticipate and get to where GUIs are moving before the mainstream players get there. Have Microsoft play catch-up for a change.
Nobody ever won a race by aiming to be where their opponents were. You win by being where your opponents haven't yet got.
Excellent to see scientific stuff like OPeNDAP get in there. (The stats on Freshmeat show that it's not a particularly well-known project, but it's an important one for distributed data.) Globus (a fascinating package for developing grids) also made it. Climate Code Foundation, Genome Informatics, the Marine Biological Laboratory and CERN have also had projects accepted, boding well for a boost to seriously complex computing. Not only that, but internships on heavy-duty projects may well produce programmers with a far healthier respect for specifications and documentation - often lacking in industry - because even looking at some of these mega-complex projects without specs will cause your brain to explode.
OS projects:
Nice range of OS' - Linux, Plan9, FreeBSD, Haiku and HelenOS amongst others. Hard to tell what the outcome will be, but since OS theory tends to be taught along very narrow lines that all these projects violate profusely and in ways that would shock the prudish, I'd expect that this might get people thinking through problems in new ways
There were typically three grades of certificate in the Old Days - personal certificates (which is what you're describing), level 2 (where there were basic background checks) and level 3 (where they made the NSA's Top Secret clearance look trivial).
These days, I'd extend the range but I'd say there should be an absolute minimum level for certain types of activity and that this should be enforceable in some way. (We know damn well that if it was voluntary, every bank and retailer would still go for the personal certs because they're dirt cheap, eliminating any real choice or any real security, with no alternative for consumers.)
Slashdot Mirror
Yeah, watching nudist beaches is indeed more specialized, but I dispute the claim that this makes them less useful to scientists. If anything, geeks need devices like that more.
Last I heard, the Swedes were working on a Stealth Frigate, a ship effectively invisible to radar and sonar (carbon fibre hull, ultra-dark materials, non-ferrous engines, that sort of thing). Haven't heard anything on that since. If they actually developed such a vessel, then between that and Google Earth, I'm not sure they'd need any spy planes.
The French, having forced everyone to adopt their units, have now redefined the meter in order to differ from everyone else.
When Wonder Woman sues you for patent infringement. Duh!
Hey, don't knock it! In a hundred years time, that could be the best-selling soft drink for electronics!
The underlying problem, though, is that you can never know if something is secret, you can only know if something is not secret. Thus, you have a paradox - the only way to know if something was secret is to share it and see if anyone else already knew.
As such, any system based on secrets of any kind whatsoever is inherently flawed because it is dependent on an assumption that is provably unprovable.
This is why you will see the phrase "security through obscurity is no security at all".
The catch is that published crypto systems rely on a problem being irreducably hard (exponential or hyperexponential difficulty). There are very few hard problems for which there is a non-existance proof of a solution simpler than the best solution hitherto known. This means that opponents could be keeping secret a solution to the problem you do not possess. However, they too are subject to the secrets paradox and therefore their ability to crack the cypher in better-than-expected time may be known to others besides themselves.
(This has been used in wartime, when Allies and the Axis powers deliberately sent messages in encrypted forms that they knew the other side could break in order to confuse them.)
The first of the removed paragraphs could be considered "original research" (banned on Wikipedia). I'm of the opinion that linear deductions are not research, but automatically follow. However, I've had a few entries edited out as "original research" myself and know that Wikipedia takes the rule extremely seriously even if it is to the point of absurdity.
The rest of the paragraphs are more inflamatory/op-ed and don't belong in an encyclopedia setting. They may be technically correct (only RSA knows) but they are most certainly not neutral POV and not useful in understanding the event.
I'd interpret it as "our firewall AV isn't stopping it", which is fine because AV software isn't a generic solution but one that detects specific, well-defined viruses. And when you shove it onto a firewall, it can't do much checking if you don't want horrible packet loss.
What it does mean, though, is that whoever wrote the article doesn't use NIDS or HIDS (the former will detect cross-scripting attacks, the latter will detect changes to files that aren't supposed to change) but relies entirely on anti-virus software on a (probably) mis-configured firewall that (likely) is running obsolete software.
Didn't we already see this article?
Anyways, as said before, there's plenty of guides (including by the NSA) on how to not suffer cross-scripting attacks. That anyone still suffers from them is not through a lack of resources.
Well, not entirely sure how much precision is required. Each reactor has its own signature radiochemistry (which is why you can identify which reactor produced any given sample of material). In turn, you know the chemistry of the radionucleotides, the daughter radionucleotides, the containment vessel and the concrete. Any given reaction is impossible below the energy required to initiate it or above the energy at which the reaction produces the lowest-energy state. You only need to know if the energy is inside the range for that reaction, since you don't care why something won't happen.
From that, it's certainly possible to model different scenarios on a decent computer. (I'd consider the Earth Simulator pretty decent and I'd expect TEPCO can get time on it.) You then look for the scenario that matches the chemistry as observed from air pollution and water pollution. Since you know what chemistry HAS occurred, you can infer a lot about what it would take to have made it occur in the first place. Since you know the ratios of radionucleotides, you can infer something about the rate of reaction and therefore the mix of fuel and contaminants (and therefore the degree to which meltdown would have been required in order for the contamination to have reached that point).
So there's a lot you can do by indirect observation and deduction.
Let's put this in context. My A-Level computing project was an expert system to identify the most probable radionucleotides present in any given sample by looking at the energies detected, adjusting probabilities based on the probability of the presence or absence of associated daughter radionucleotides given the age of the sample. (Since iterating this could change whether something was considered detected or not, each iteration could branch off multiple ways. It was an interesting study in herustics.) It was calibrated using Chernobyl fallout and tested using historical samples collected from other incidents and achieved a success rate comparable with any of the commercial systems at the time and superior to the software BNFL was using for the same purpose. (I happen to know this because the BNFL software was used in a court case regarding leukemia deaths in Seascale and the raw data from Seascale was available to me. To be fair, BNFL's software was listed as version 0.8, so was likely not even out of beta testing and may well have been buggy.)
Now, what do you learn from this? Well, (a) I've studied the subject well enough to match the nuclear industry experts on analysis when I was 18, and (b) I had the means to actually obtain the raw data, the actual lab analysis of what was present in the samples and enough books on radiochemistry to write my own solution from the ground up. In other words, not exactly on the outside.
It hasn't kept it cool - the temperatures rose much more sharply after spraying began and the partial meltdown is likely more a product of the panic reaction than anything. The use of seawater, initiallly, also will have complicated things. Far too many minerals will now be baked onto the sides of things, it's far too corrosive, and it's generally really not well-suited to this kind of stuff.
The Soviet core was not on fire. Part of the Soviet core was in the stratosphere, the other part was melting its way through the concrete. There was no fire. The Soviet response was, in essence, to seal up the molten core. Sensible, but diluting it first (anything that is liquid at that temperature and absorbs neutrons - to any extent - would work) would have been better.
The only serious core fire that has ever happened was at Windscale, and that was England. Made a real mess of the local area, that did, with the soil contamination being a major problem.
Why on earth would you need to upgrade the nuclear portion of the plant? Ten years is plenty of time to build watertight doors for the emergency generator rooms (whose flooding caused many subsequent issues) and watertight doors leading into the occupied portions of the facilities (cut down on water pooling).
They most certainly do not have things under control - the head of TEPCO is hospitalized due to stress, the leaks haven't been located (never mind closed), the groundwater levels are far higher than anticipated and the engineers there are concerned that the partial meltdown may progress to a total meltdown that breaks out of the containment.
If that's your idea of control, I'd hate to see your idea of a breakdown.
As for understanding what's going on, I probably have a better grasp of nuclear reactor systems, their technology and their history than anyone else on Slashdot.
In the sense that retro is chic. Besides, the 80 megapixel camera won't fit in that small a case.
I take it that is sarcasm. Not very good at it, are you?
Nuclear power is perfectly safe, if done properly. So is coal mining. Both become extremely dangerous when not done properly - nuclear power due to the risk of reactions getting too hot (decay causes heat, accelerated decay through neutron emissions causes a lot more heat), which can lead to a failure of the structural integrity of the system or - worse - uncontrolled chemical reactions resulting in a chemical explosion (essentially a "dirty bomb"), coal mining due to the risk of coal gas (methane) igniting, resulting in an explosion and/or an uncontrolled fire within the coal seam itself, beyond the confines of the mine.
In both cases, the problem is not so much the initial event (although nobody likes fatalities - unless they're newsreaders or Fox TV presenters), the problem is that event zero can lead to the problem spreading in a way that cannot be controlled or stopped.
In both cases, competent design and competent management can make the probability of event zero happening at all virtually zero. The number of nuclear reactors worldwide is extremely high, but other than the Windscale core fire, Three Mile Island, Chernobyl and the Fukushima complex, there really hasn't been any major accident in the industry in 50 years. That's not bad, given that our knowledge of the physics is the same age.
(The Windscale core fire was an interesting piece of history. A graphite nuclear reactor core was allowed to burn for 3 days before anyone even thought to check why the temperature gagues were showing excessive readings. This was not corruption or greed, and even calling it incompetence is a stretch as maintenance was due over that time.)
The fact of the matter is this: in ALL of these accidents, there was a VERY long chain of events from the the initial point that turned towards disaster and the disaster actually happening. ANY person along that chain COULD have broken that chain at ANY time. They failed to do so.
Typically in disasters, this is because the difference between what they should have done and what they did do was so small that the person disregarded the difference. A very large sum of very small deltas will eventually add up.
(Even the Titanic's sinking was not due to a single person's failure or a single event, but a laundry list of very tiny deltas from the time the iron was first processed to the time the helmsman mistook what sterring order the helm was set to. R101, likewise, failed because of an incredibly large number of people making incredibly tiny errors.)
Yours is a common, and pitiful, belief that criticism of a sequence means criticism of the entire world in which that sequence lives. I've noticed such a mindset most amongst the right-wing and the libertarians. There is, sadly, no known cure and they are doomed to live in a world that really doesn't exist outside of their own interpretation.
I'm not going to repeat the BBC, the Guardian, or indeed the scientists who have published papers on the flaws in the design and positioning of the reactors - back in 2001. I'm also not going to repeat classical demonstrations of combusion (taught to me by the superb Dr John Salthouse of the University of Manchester), though I've given you one of the basic experiments demonstrating what essentially happened in a prior posting.
The workers aren't idiots, they're Japanese - taught to obey orders without question, taught that honor is worth more than life, and exposed to a situation where if they didn't damn well go in there (regardless of the inept management) a lot more people could die. You can look up the reports on the needless and senseless radiation burns they've suffered, the IAEA commentary, the level of badging - this is all public knowledge. But, no, you'd rather question someone who has the integrity to not AC, the age and experience to be a 4UID, and the knowledge of the basic chemistry and physics involved (and it isn't hard). Questioning irrationally to cover up your own fears is something that deserves scorn, not respect. Maybe that is exactly why you posted as AC.
The papers were published in 2001, so we have already seen. The accident was NOT stopped and is still continuing. And, FYI, I seem to have had better chemistry lecturers since you are the one guessing and I'm the one using experimental data.
A paper presented in 2001, I believe, demonstrated by means of sand deposits that tsunamis in Fukushima have historically been around 21 meters high and have typically travelled about 4 kilometers inland. TEPCO's assessment ignored all data before 1898, which meant it ignored essentially all tsunamis.
You don't need power. Hydrogen will not combust outside the presence of oxygen and a sufficiently powerful jet of hydrogen (such as superheated hydrogen from inside a compact concrete structure) would happily burn in the air outside with no possibility of the flame descending inside even if there was any oxygen present. The best thing they could have done would have been to open up the vents that they EVENTUALLY used to their maximum as early as possible, then ignite the hydrogen jet coming out.
(You want to burn the hydrogen rather than just vent it, as you really want some measure of control.)
If you happen to have a hydrogen cylinder handy, do the following experiment. If not, look it up on YouTube. Take a tin can and drill two holes in it. One in the lid, one in the base. Connect the hydrogen cylinder to the base and open the valve. Light the hydrogen stream at the top of the can. The hydrogen will burn cleanly and brightly. And, most importantly, in an EXTREMELY stable and safe way. Now remove the hose connecting the cylinder to the can. You will observe the flame start to descend into the can. What you will NOT observe is the air being pulled in at the base. When the flame enters the can, the mix of oxygen and hydrogen inside will ignite and the top of the can will explode.
It's hard to say for certain, but I would be willing to bet that something analogous happened inside the three concrete structures that exploded. Air must have been pulled in, as the zirconium reacted with all the oxygen from the water (which is why there was hydrogen there in the first place). My guess would be that the vents being opened and closed the way they were would be suspect as the three vented structures were the three that exploded. There must also have been some point of ignition. The heat is of no interest - if that had caused combustion, it would have burned the oxygen at the rate it was being pulled in, resulting in a slow burn. That ignition must have taken place after hydrogen and oxygen mixed in significant quantity.
If the vents had been left open, there would have been no significant hydrogen left, the ignition would likely not have detonated anything of significance, and we wouldn't have an international nuclear catastrophe.
The first problem is that TEPCO isn't telling anyone what they know (to save face and because they're freaking out)
The second problem is that whatever they are telling, they're telling to the Japanese government and no-one else (even their own workers, who they convinced to wade through radioactive water without boots, go into radioactive buildings without radiation badges or suitable gear, etc).
The third problem is that the experts are working with minimal data - and what they do have is suspect
The fourth problem is that TEPCO has been trying to salvage the reactors at the same time as spraying them with seawater (which would be corrosive) and after the outer shell had exploded on three of them (causing untold damage to electronics, shock-proofing, etc)
On top of all that, TEPCO allowed the hydrogen build-up in the first place. They could have burned it off with a controlled burn. This would have prevented the explosions, reduced the spillage and possibly prevented the fuel leak. (Reducing pressure may have reduced water temperature and may have conserved some of the cooling pools.)
As for building the reactors ALONG the fault-line, despite advice not to by their own chief scientists, and building a tsunami wall far lower than the historic tsunami wave-heights....
This accident was stoppable at so many points in so many ways. The problem wasn't so much the reactor alone as the mindset together with the reactor.
I suggested to the Blender folks that they might want to look at elastons, as animation is a key area for Blender but they didn't have many ideas for how to improve it. Got no response and they didn't update their page, but I still think it'd be a worthy extension that could very well be helpful in their movie series (as rigid models only go so far).
I agree but also disagree.
Trying to emulate Windows will only ever produce a second-rate Windows (because the original will always be ahead of you by definition). That can't really let Linux become mainstream, because nobody is going to want second-rate. The only way Linux can become acceptable on the mainstream is to do things differently. It's why Apple still exists at all - if MacOS or OS/X were mere clones of Windows, Apple would be long-dead. They hold a small but very respectable market share because they solve things differently. They've only lost out in those areas they tried to copy Microsoft - if they'd kept with solving desktop problems distinctly but usably, they'd probably be doing better today.
I agree, however, that you'll only go so far with altering the underlying code. "Mucking about with code" is why there are Linux variants that are FAA-rated and others that are Carrier Grade, demonstrating incredible robustness. I'd consider that a high achievement. Users probably don't care for such certifications, but they certainly do care that their system doesn't crash when they're 5 points off the high score or half-way through a term paper they've stupidly not been saving. What it can NEVER do, and this is - I think - the point you're making, is make the software any more usable, any more learnable or any more efficient for getting things done.
My contention is that the front-end part, the user part, can't be solved by throwing people at it until a genuinely new way of approaching this whole "user interface" problem exists. The desktop metaphor has been done to death by Microsoft and Apple, and most of the recent advances have moved away from that entire concept. If Linux wants to be mainstream, it has to anticipate and get to where GUIs are moving before the mainstream players get there. Have Microsoft play catch-up for a change.
Nobody ever won a race by aiming to be where their opponents were. You win by being where your opponents haven't yet got.
There are indeed two winners - the users and the developers.
Scientific projects:
Excellent to see scientific stuff like OPeNDAP get in there. (The stats on Freshmeat show that it's not a particularly well-known project, but it's an important one for distributed data.) Globus (a fascinating package for developing grids) also made it. Climate Code Foundation, Genome Informatics, the Marine Biological Laboratory and CERN have also had projects accepted, boding well for a boost to seriously complex computing. Not only that, but internships on heavy-duty projects may well produce programmers with a far healthier respect for specifications and documentation - often lacking in industry - because even looking at some of these mega-complex projects without specs will cause your brain to explode.
OS projects:
Nice range of OS' - Linux, Plan9, FreeBSD, Haiku and HelenOS amongst others. Hard to tell what the outcome will be, but since OS theory tends to be taught along very narrow lines that all these projects violate profusely and in ways that would shock the prudish, I'd expect that this might get people thinking through problems in new ways
Dunno how expensive dirt is where you live, but it's free here. :)
Ok, yes, personal certs were offered free by Thawte and - I think - even Verisign for a bit.
There were typically three grades of certificate in the Old Days - personal certificates (which is what you're describing), level 2 (where there were basic background checks) and level 3 (where they made the NSA's Top Secret clearance look trivial).
These days, I'd extend the range but I'd say there should be an absolute minimum level for certain types of activity and that this should be enforceable in some way. (We know damn well that if it was voluntary, every bank and retailer would still go for the personal certs because they're dirt cheap, eliminating any real choice or any real security, with no alternative for consumers.)