Microsoft is dying (damnit BSD Trolls:-). You can't see it now, but you can see it in the intertia. Everybody knows that if you partner with them successfully as did IBM, Novell, Borland, the anti-virus companies, Netscape and so on then eventually Microsoft will get interested in your business and try to take it from you. Sensible developers try to keep a distance, continually thinking about how to move to other platforms. At the same time all of their new products (Zune, Azure, WP6, WP7, Kindle etc.) fail to get real traction. It is not an accident that Apple has a higher valuation; anyone can see that Microsoft is disfunctional.
There are lots of institutional investors with vast amounts of shares in both Microsoft and Nokia. I think that a bunch of the have decided that Nokia is worth sacrificing in an attempt to save Microsoft. They figure that it might not work, but if it does they will protect a much bigger investmen.
I'm guessing these companies made it "worth while" for the senior management at Nokia to bring in an MS guy.
There have been a bunch of interesting paleontological discoveries by Chinese scientists in the past few years. These werereported in western media. Generally it's not surprising if they end up attracting good people from elsewhere and if there weren't scientists from other countries getting involved then you would begin to be concerned. International collaboration is a crucial element of scientific credibility. China probably (rightly) wants that more than it wants credit for any particular dinosaurs. From a long term economic point of view this should probably be more important in China's attempt to overtake the USA economically. There is no way that research like this is going to be properly funded by private companies but you need it to get the really bright fundamental science people to come and visit and that, long term, is what drives real invention, not just thousands of patents on minor variations of the same idea.
The edit was incorrect in any case. There are pretty clear Wikipedia policies limiting editing of your right to edit articles about yourself. The edit didn't clearly state who it was from. The editor should have copied the text to the talk page for discussion. There were facts which have been referred to elsewhere on news sites (e.g. the existence of an RSA letter to customers) which were simply deleted. Most importantly, all of the speculation referred to in the edit does exist in widely known sources. At most adding a "citation needed" tag of some sort would be the right thing to do. The best thing would be to link to the correct sources and rewrite to state exactly what they say.
There's a more important thing here though. The edit came from EMC. As a company EMC knows the current state of the investigation. If they know that the secrets have been stolen then the edit was disingenuous. If they know that the secrets have not been stolen then they could also say so. If they refuse to answer that then they shouldn't be messing around manipulating alternative media. If it was a rogue employee who doesn't know the status of the investigation made the edit they should also say so. They can only expect to have correct coverage to the extent
And just to amplify this with a bit of Wikipedia manipulation; have a look at this edit which comes from 128-221-197-57.emc.com, Where EMC is RSA's parent company, which I found from this article which also includes an RSA letter which they are supposedly sending out to customers.
Full disclosure to all affected users; it shouldn't be a matter of dispute. It should be the law.
Yes; fun fun fun. It's good the way they let a mafia of MSCE certified IT administrators pretend they didn't screw up by choosing SecurID and letting them keep the seed info whilst their real customers, the people who have their systems and data secured with SecurID, don't know squat about what's going on.
This is an understatement. We continually hear that "Sun was OSS friendly"; "Sun contributed more than anyone else" etc. etc. The thing was, they did put stuff out, but they almost always did it in a way that meant they didn't join the community. Their choice to put out Solaris under the CCDL was deliberately done to make it incompatible with the Linux community and the FSF. They took ages to open up Java, and a truly GPL Java only came about because of the efforts of RedHat and others on IcedTea.
The thing about RedHat is that they manage to be consistent (even boring). They make some money related to non open source, but they always try to move in the direction of FOSS. This means that gradually, steadily more and more people start to trust them and build solutions on their stuff. Red Hat even manages to get away with not opening up things for ages simply because everyone knows they will eventually (I'm talking about their own software; not license breaches). Sun could never manage more than two months without an about-face. Look at the difference between Ubuntu and RedHat and you can see that ever since they took on their idiot former COO, Ubuntu have been losing credibility. His hire was probably more a symptom than a cause, but you could see immediately that he started blogging for open-core Ubuntu started to lose devotees.
Sun was actually worse than Oracle. With Oracle it's pretty clear where you stand. Make it financially worth their while (==Millions minimum) and they will negotiate what you want. Otherwise give only exactly what you agree to give and only in return for exactly what they promise. You are never going to end up in the situation of being screwed over by surprise.
<\rant>
Actually, as a person who remembers 6800 Suns, I just remembered what it was. The bastards who killed SunOS to inflict Solaris on us. Why couldn't you just free the source. Sun Management; AT' Never forgive. Never forget.
There is nothing wrong with the fact that many people can sign certificates. What is wrong is that there's no easy way to mark that up and control it and there are no ways to have multiple independent signing bodies. E.g. for financial transactions I would only want to trust a bank signed by an extended verification certificate from at least two registries + the government regulatory body of the country where the bank is registered. When I'm browsing slashdot I would probably be happy just to have a self signed certificate and get warned if it changed. What is needed is essentially a web of trust like PGP with a pre-loaded set of trusted bodies which varies according to the configuration of the user. There is no reason for a Chinese user to trust an American bank or the other way round.
With sufficiently clever defaults this could add quite a bit of security without any interaction or thinking from the user. They probably have to learn more about the colours of the address bar or something however.
For emergency calls you do not need to record the cell phone location for any more time than after the next few locations come in. We have already established that cell phone companies need to know where the subscriber is in order to route calls to them. The question is, why do they need to store that information. The answer, if you had just bothered to RTFA, is that they don't need to. The German company in the article has completely stopped doing so now.
There are some mobile companies which have location dependent billing (e.g. use of your phone is free if you are in your home or close). This means that location becomes a legitimate part of billing data. The equipment manufacturers have to include the possibility of gathering it. For the bills of most customers on most networks this data isn't used, but you can never tell when someone from the marketing department is going to start such a special offer. Also you can never tell when some customer is going to turn out to have had the special offer active, but the customer service people put it on the wrong subscriber number. For this reason the technical people just keep the whole load of data and then sort it all out later. This is the way that you make sure that you can safely generate the customer's correct bill even in cases of error.
"The road to hell is paved with good intentions" (interestingly enough, if you look up Wikipedia's page on the subject they say explicitly that "Studies of business ethics indicate that most wrongdoing is not due directly to wickedness but is performed by people who did not plan to err". To get this stopped there has to be a positive inducement to destroy the data which is not actually needed.
That's the trouble. Nuclear plants are held to a massively high standard of "safe" already.
Did you know, for example, that Coal kills 4,000 (not a typo) more people per wattHour than Nuclear does? But its a slow, boring kind of killed, like the 40,000+ who die every year in automobile accidents in the US alone, not the fun exciting kind of killed that you get every couple of years when an airliner crashes and kills 200 folk halfway around the globe, making national news.
To have a meaningful discussion you need to compare nuclear safety to other power-generation mechanisms (more people fall off roofs installing solar panels and die every year than have been killed by nuclear power generation disasters). And then scale them to account for the power generated. Once you do so, you realize just how unsafe many of the alternatives actually are.
The interesting thing is to follow up your numbers (you don't give a citation and nor does this seth guy). Looking at them, we see that he includes a low figure for Chernobyl deaths (4000) and even then discounts those saying "cause and effect becomes more tenuous" whilst completely failing to take into account other studies. Now, as this article points out it's incredibly hard to work out which numbers are correct, but even the WHO which put out the 4000 death number has since published a correction which says that there have been more than 9000 deaths. Numbers of deaths go up to a million in other studies of varied credibility.
My fundamental conclusion has to be that we have yet another nuclear safety "expert" who's telling us that it's safe, but turns out to have no clue what he's talking about. Numbers of deaths go up to a million, and studies like that seem to be deeply dubious too, but I'm really not going to accept these numbers without a big bit more clarity. The worst thing about this is the open admission by the IAEA that "the total of 4,000 deaths was highlighted to counter much higher figures" - in other words they set out to manipulate perception.
Going on from there; I agree with your point about visible public death being more worrying than hidden death. If the death rate in planes was anything like the rate in cars they would be banned. However, there's another bias factor in here. The small nuclear incidents are not really so worrying. What is worrying is outlier major events. We seem to have come close to a reactor melt down; the reactor containment failed but still we have been lucky. If cooling had failed for longer what might have happened? I see lots of assurances that nothing. These come from the same people who tell us that exposure to beta decay and gamma rays is the same as exposure to UV light (see recent articles in the Register)
No you don't. One of the original nuclear bombs worked by putting two non critical bits of metal together. When they are separated they are non critical since the neutrons are lost to the environment. When you bring them together they go critical since the neutron exchanage from one to the other is enough to allow the chain reaction to continue.
I'm sorry to pick on your post which seems to be merely factually incorrect and not malicious, but it's exactly comments like this where someone assures me 100% that things are safe whilst making basic errors in nuclear physics which worry me most of all about the nuclear industry. The perfect example is the person above who claimed that radioactive cesium decays away in hours.
Look, I'm sure the risk is small; it may even be "infinitesimal" but it isn't zero. Consider for example a situation where the containment vessel is cracked (as has already happened) and that crack leads to the uranium going down into one particular area of the vessel. Consider also the situation where the bottom of the vessel has been filled in with other material (dried up salt from sea water?) forcing a change in configuration.
There are reasons to believe that a stable critical mass will not be reached; it's incredibly difficult to do and the uranium will tend to blow it's self apart immediately that happens meaning that no real nuclear explosion will happen. However, a little more humility and a whole bunch more circumspection would really help you maintain credibility for when you want to persuade people that the modern "safe" nuclear plants really are safe.
RH requires a ridiculous support contract for access to patches.
Patches (as in source) are available for free; That's what CentOS bases off. It's the updates (as in binaries) which you pay for. If RedHat ceases to do this then I, and many like me, will start releasing them myself.
Finally, RH has only 4.8% of the packages Debian has, in a comparison of official stable repositories
Your comparison really isn't fair for several reasons. You are comparing only the core repositories of RH which are at a higher quality level than Debian's default repository precisely because they are more filtered. Try doing the same comparison with EPEL and/or RPMForge (which is where "community" packages are) included and you will see a more even story.
Red Hat sells support, not software, they would do well to remember that.
Both; to be honest; their Release CD images are still valuable. The fact that they limit access to the open access to the support forums is a complete pain even when we are fully paid up because most of our internal users aren't registered with RedHat and so we will never give direct access to all the people who might need it.
I'm not yet at all sure that this RedHat thing is a big problem; if it really is, then the correct solution is a license change at the kernel level. If RedHat can do this, then so can anyone else. That means that the Linux kernel people have to initiate, support and get involved in a GPLv4 update, using their "diplomatic skills" to persuade the FSF they are a partner worth working with. They have to concentrate on getting most known kernel developers to agree to a future license update and eliminating those that don't.
In the meantime, the key thing is to make sure that upstream of RedHat, more of these patches get merged in more effectively. This will have the benefit for RedHat of taking work off RedHat; will put more work into a better condition, will make the upstream old kernels more valuable and will provide wider hardware support for all Linux based on old kernels.. Oracle could put a bunch of kernel developers and testers on the case. If nobody is willing to do this, then, to be honest, RedHat is doing the right thing
In the meantime; even if you are a RedHat customer, concentrate on putting answers you find to problems onto Linux Questions Wiki or forums where they will be available under the CC-SA for everyone.
Because the grandparent was completely Astroturf and moderated down for a good reason. By replying to it you are increasing the discussion about it and doing the work for the astroturfer. You are directing the discussion away from Microsoft's need to use lawsuits in a market where they can't compete through lack of platform and towards a minor issue; the quality of the development tools on WP7; which almost nobody here, let alone normal users, will ever care about.
Trolling is exactly trying to misdirect discussion; you have succeeded (I'm even being trolled here, replying to you, but it's an important point as long as people like your other respondents try to pretend that this is some kind of anti-Microsoft campaign). Your moderation as a troll was 100% correct, whatever your intentions.
Is it? I think that people know how to do forwarding etc. etc.
It seems to me that it's actually easier to block all executable content (flash / javascript etc) and then block file upload/download to / from the browser than it would be to find every possible https based mail service (including my own secret one; which is used only by me personally and even that almost never) which is what you would have to do in order for this to make sense.
No; this is fundamentally different. Consecutive life sentences affect a person. This is a sentence against a company. There was no problem charging some poor ignorant file sharer millions, but over-finling a company is an insult to justice.
So; The grandparent arrives suddenly having never posted before, is spending enough time refreshing to get a first post. I think we find that suspicious. We think that accusing WrongSizeGlass of trolling must be unfair and outrageous
But then let's look at the moderation on this (if you don't have an account you will want to sign up for one to see this stuff).
Starting Score: 1 point
Moderation -2
70% Troll
30% Insightful
Extra 'Troll' Modifier
If the moderation is 70/30 then the only possible solutions are multiples of 7 and 3 so at least three insightful moderations have been given. Now, the first post has some interesting comment. It's definitely a reflection of Microsoft's paranoia. However, there are too many things which are wrong which have been covered too often on Slashdot for any serious mod to mod it insightful. In any case, almost all legitimate positive moderation shows variety (underrated / interesting / insightful).
Looking at the facts there seem to be two possible explanations; a) the GNAA has returned having learned to troll subtly, pretending to be Microsoft Astroturfers and never ever linking to Goatse whilst building an elite undercover uber-posting super-moderators who can afford to give them mod points without risking losing the ability to moderate b) Microsoft is an immoral deceptive company using publicity people to spread lies and astroturfing like mad.
Obviously b) which would mean we do need a way to mod astroturfers is untenable so only a) can be true. Microsoft is an upstanding tax paying pillar of society. They would never ever ever astroturf and lie and cheat. WrongSizeGlass's moderation as troll must be justified.
Your post is insightful but I'll pick one little nit.
Copyright is the basis of the software industry, both closed and open source.
The basis of the software industry is people's need for software. If there was no copyright protection business models would have to change, but there would definitely still be customers paying people to create software. Free software would likely do fine, not on the basis of copyleft but simply on the basis of fair sharing. Software users would pay for maintenance and quality control or would learn to download from specific sites and do their own.
Definitely there would be some disadvantages; proprietary software companies would likely rely much more on copy-protection software (unless that was made illegal; a definite thought) and a small number of companies that contribute to software due to legal pressure would not. On the other hand there would be advantages; reverse engineering would advance much faster since there would be no legal restrictions on it and the end results would be immediately usable. This would in turn lead to much more secure software since even closed modules would be subject to serious scrutiny.
I'm not sure that total elimination of copyright is a good idea, but it's very important to remember that copyright is a bargain restrictions on freedom of speech are paid for by a promise that in future there will be more varied speech available. When one side stops keeping up their end of the bargain by using DRM which keeps works locked in even after the term of copyright expires, by continually extending copyright terms, by not delivering source code to proprietary programs even after they have ceased to be fully maintained and by attempting to use other limits on free speech such as patents then the time has arrived to question copyright full stop.
The "general public" may not care, but, before they buy they normally ask those of us who know and do care. When we tell them to avoid a "locked down" phone they probably don't even know what it means, they just know I think it's bad. Previously these people bought HTC instead, now I will direct them elsewhere.
No, your understanding of cancer is far too simple to be useful in this application.
Even my understanding of radiation was too simple/simplified to correctly model that. When one single particle of radiation passes through the body it is possible for it to cause multiple interactions; e.g. a gamma ray could hit one single cell's DNA at multiple points. That's before we take into account that a high energy particle, e.g. a single cosmic ray, could cause a cascade of different particles.
My point; my application; was not to say that any radiation above two strikes is "dangerous". It was to say that a) there is no threshold level; risk starts from almost the beginning and b) the risk is likely not linear and certainly depends on the way the dose is delivered. Statement a) matches the 2005 NAS report which states clearly:
There is no safe level or threshold of ionizing radiation exposure.
As for your comment:
There was a crazy guy who thought exactly two "strikes" would make a huge difference, but he was debunked in a series of studies years ago.
Direct links to actual scientific studies or even just some names to search for really greatly appreciated. Most of us have a scientifically educated layman's interest in this subject but not a postgrad student's time to research this ourselves.
Definitely a really nice chart. It's good to see something so easy to read and quantitative that helps people debate with some level of knowledge. The main problems for me with it are that it doesn't really do a good job on the time axis, spacial axes and the probabalistic risk. For example:
50mSv absorbed in one year is probably completely safe. 50mSv absorbed in one second is quite likely to be bad, even if that's the only radiation absorbed in that whole year.
50mSv spread through all issue types is likely no problem. Even though skin is normally considered less important in calculating Sieverts, 50mSv concentrated on a small area of skin can be a real problem .
What makes this all difficult is that it seems the mechanisms are random. E.g. most of the time a particle of radiation does nothing. It dissociates a water molecule which soon after re-associates. Even if it does cause a mutation, that likely doesn't cause cancer because the body copes with mutation all the time and genetic codes self correct. However, if two or more mutations happen in close together / related genetic material in the same cell, that is reasonably likely to cause cancer as the cell is no longer able to self-correct. Now of course, this means that the "Lowest one-year dose clearly linked to increased cancer risk" is actually incorrect; that minimum ("clearly linked to increased", not to "noticeable") is about two particles of radiation where clearly is defined as "we clearly understand that this is so and "increased" is defined as "greater than would be otherwise. However, the minimum yearly dose "linked to a worrying increase according to a reasonable probabalistic model" is what we really want to know and is completely missing from the chart.
Since the location of radiation damage is entirely random, that can mean that millions of particles could cause no damage to one person whilst just three could damage another very unlucky person. This risk gets higher the more concentrated in space and time a dose of radiation is. When you think about it, the reason is obvious. The chance of a repeat strike in the same cell goes up quadratically as the volume shrinks and factorially as the dosage increases. These are the crucial things which mean that radioactive iodine and back scatter scanners are likely to be much more dangerous than e.g. cosmic ray exposure at altitude or through body X-rays. They are also mean that having a back scatter X-ray just before or after travelling is (I have no idea exactly how much) worse than having the X-ray on its own.
It would be really great if xkcd could do something which did a comparison of the dangers of different kinds of radiation exposure in different circumstances. Very important would be to leave in the ares of doubt where we actually don't know.
Slashdot Mirror
Microsoft is dying (damnit BSD Trolls :-). You can't see it now, but you can see it in the intertia. Everybody knows that if you partner with them successfully as did IBM, Novell, Borland, the anti-virus companies, Netscape and so on then eventually Microsoft will get interested in your business and try to take it from you. Sensible developers try to keep a distance, continually thinking about how to move to other platforms. At the same time all of their new products (Zune, Azure, WP6, WP7, Kindle etc.) fail to get real traction. It is not an accident that Apple has a higher valuation; anyone can see that Microsoft is disfunctional.
There are lots of institutional investors with vast amounts of shares in both Microsoft and Nokia. I think that a bunch of the have decided that Nokia is worth sacrificing in an attempt to save Microsoft. They figure that it might not work, but if it does they will protect a much bigger investmen.
I'm guessing these companies made it "worth while" for the senior management at Nokia to bring in an MS guy.
There have been a bunch of interesting paleontological discoveries by Chinese scientists in the past few years. These were reported in western media. Generally it's not surprising if they end up attracting good people from elsewhere and if there weren't scientists from other countries getting involved then you would begin to be concerned. International collaboration is a crucial element of scientific credibility. China probably (rightly) wants that more than it wants credit for any particular dinosaurs. From a long term economic point of view this should probably be more important in China's attempt to overtake the USA economically. There is no way that research like this is going to be properly funded by private companies but you need it to get the really bright fundamental science people to come and visit and that, long term, is what drives real invention, not just thousands of patents on minor variations of the same idea.
The edit was incorrect in any case. There are pretty clear Wikipedia policies limiting editing of your right to edit articles about yourself. The edit didn't clearly state who it was from. The editor should have copied the text to the talk page for discussion. There were facts which have been referred to elsewhere on news sites (e.g. the existence of an RSA letter to customers) which were simply deleted. Most importantly, all of the speculation referred to in the edit does exist in widely known sources. At most adding a "citation needed" tag of some sort would be the right thing to do. The best thing would be to link to the correct sources and rewrite to state exactly what they say.
There's a more important thing here though. The edit came from EMC. As a company EMC knows the current state of the investigation. If they know that the secrets have been stolen then the edit was disingenuous. If they know that the secrets have not been stolen then they could also say so. If they refuse to answer that then they shouldn't be messing around manipulating alternative media. If it was a rogue employee who doesn't know the status of the investigation made the edit they should also say so. They can only expect to have correct coverage to the extent
And just to amplify this with a bit of Wikipedia manipulation; have a look at this edit which comes from 128-221-197-57.emc.com, Where EMC is RSA's parent company, which I found from this article which also includes an RSA letter which they are supposedly sending out to customers.
Full disclosure to all affected users; it shouldn't be a matter of dispute. It should be the law.
Yes; fun fun fun. It's good the way they let a mafia of MSCE certified IT administrators pretend they didn't screw up by choosing SecurID and letting them keep the seed info whilst their real customers, the people who have their systems and data secured with SecurID, don't know squat about what's going on.
Don't keep your database of nuclear launch codes on your gaming PC. Use a non networked computer instead.
has the securid seeds database been compromised?
anything else you announce is fluff.
This is an understatement. We continually hear that "Sun was OSS friendly"; "Sun contributed more than anyone else" etc. etc. The thing was, they did put stuff out, but they almost always did it in a way that meant they didn't join the community. Their choice to put out Solaris under the CCDL was deliberately done to make it incompatible with the Linux community and the FSF. They took ages to open up Java, and a truly GPL Java only came about because of the efforts of RedHat and others on IcedTea.
The thing about RedHat is that they manage to be consistent (even boring). They make some money related to non open source, but they always try to move in the direction of FOSS. This means that gradually, steadily more and more people start to trust them and build solutions on their stuff. Red Hat even manages to get away with not opening up things for ages simply because everyone knows they will eventually (I'm talking about their own software; not license breaches). Sun could never manage more than two months without an about-face. Look at the difference between Ubuntu and RedHat and you can see that ever since they took on their idiot former COO, Ubuntu have been losing credibility. His hire was probably more a symptom than a cause, but you could see immediately that he started blogging for open-core Ubuntu started to lose devotees.
Sun was actually worse than Oracle. With Oracle it's pretty clear where you stand. Make it financially worth their while (==Millions minimum) and they will negotiate what you want. Otherwise give only exactly what you agree to give and only in return for exactly what they promise. You are never going to end up in the situation of being screwed over by surprise.
<\rant>
Actually, as a person who remembers 6800 Suns, I just remembered what it was. The bastards who killed SunOS to inflict Solaris on us. Why couldn't you just free the source. Sun Management; AT' Never forgive. Never forget.
<\rant>
There is nothing wrong with the fact that many people can sign certificates. What is wrong is that there's no easy way to mark that up and control it and there are no ways to have multiple independent signing bodies. E.g. for financial transactions I would only want to trust a bank signed by an extended verification certificate from at least two registries + the government regulatory body of the country where the bank is registered. When I'm browsing slashdot I would probably be happy just to have a self signed certificate and get warned if it changed. What is needed is essentially a web of trust like PGP with a pre-loaded set of trusted bodies which varies according to the configuration of the user. There is no reason for a Chinese user to trust an American bank or the other way round.
With sufficiently clever defaults this could add quite a bit of security without any interaction or thinking from the user. They probably have to learn more about the colours of the address bar or something however.
For emergency calls you do not need to record the cell phone location for any more time than after the next few locations come in. We have already established that cell phone companies need to know where the subscriber is in order to route calls to them. The question is, why do they need to store that information. The answer, if you had just bothered to RTFA, is that they don't need to. The German company in the article has completely stopped doing so now.
There are some mobile companies which have location dependent billing (e.g. use of your phone is free if you are in your home or close). This means that location becomes a legitimate part of billing data. The equipment manufacturers have to include the possibility of gathering it. For the bills of most customers on most networks this data isn't used, but you can never tell when someone from the marketing department is going to start such a special offer. Also you can never tell when some customer is going to turn out to have had the special offer active, but the customer service people put it on the wrong subscriber number. For this reason the technical people just keep the whole load of data and then sort it all out later. This is the way that you make sure that you can safely generate the customer's correct bill even in cases of error.
"The road to hell is paved with good intentions" (interestingly enough, if you look up Wikipedia's page on the subject they say explicitly that "Studies of business ethics indicate that most wrongdoing is not due directly to wickedness but is performed by people who did not plan to err". To get this stopped there has to be a positive inducement to destroy the data which is not actually needed.
That's the trouble. Nuclear plants are held to a massively high standard of "safe" already.
Did you know, for example, that Coal kills 4,000 (not a typo) more people per wattHour than Nuclear does? But its a slow, boring kind of killed, like the 40,000+ who die every year in automobile accidents in the US alone, not the fun exciting kind of killed that you get every couple of years when an airliner crashes and kills 200 folk halfway around the globe, making national news.
To have a meaningful discussion you need to compare nuclear safety to other power-generation mechanisms (more people fall off roofs installing solar panels and die every year than have been killed by nuclear power generation disasters). And then scale them to account for the power generated. Once you do so, you realize just how unsafe many of the alternatives actually are.
The interesting thing is to follow up your numbers (you don't give a citation and nor does this seth guy). Looking at them, we see that he includes a low figure for Chernobyl deaths (4000) and even then discounts those saying "cause and effect becomes more tenuous" whilst completely failing to take into account other studies. Now, as this article points out it's incredibly hard to work out which numbers are correct, but even the WHO which put out the 4000 death number has since published a correction which says that there have been more than 9000 deaths. Numbers of deaths go up to a million in other studies of varied credibility.
My fundamental conclusion has to be that we have yet another nuclear safety "expert" who's telling us that it's safe, but turns out to have no clue what he's talking about. Numbers of deaths go up to a million, and studies like that seem to be deeply dubious too, but I'm really not going to accept these numbers without a big bit more clarity. The worst thing about this is the open admission by the IAEA that "the total of 4,000 deaths was highlighted to counter much higher figures" - in other words they set out to manipulate perception.
Going on from there; I agree with your point about visible public death being more worrying than hidden death. If the death rate in planes was anything like the rate in cars they would be banned. However, there's another bias factor in here. The small nuclear incidents are not really so worrying. What is worrying is outlier major events. We seem to have come close to a reactor melt down; the reactor containment failed but still we have been lucky. If cooling had failed for longer what might have happened? I see lots of assurances that nothing. These come from the same people who tell us that exposure to beta decay and gamma rays is the same as exposure to UV light (see recent articles in the Register)
No you don't. One of the original nuclear bombs worked by putting two non critical bits of metal together. When they are separated they are non critical since the neutrons are lost to the environment. When you bring them together they go critical since the neutron exchanage from one to the other is enough to allow the chain reaction to continue.
I'm sorry to pick on your post which seems to be merely factually incorrect and not malicious, but it's exactly comments like this where someone assures me 100% that things are safe whilst making basic errors in nuclear physics which worry me most of all about the nuclear industry. The perfect example is the person above who claimed that radioactive cesium decays away in hours.
There is zero risk of supercriticality.
Look, I'm sure the risk is small; it may even be "infinitesimal" but it isn't zero. Consider for example a situation where the containment vessel is cracked (as has already happened) and that crack leads to the uranium going down into one particular area of the vessel. Consider also the situation where the bottom of the vessel has been filled in with other material (dried up salt from sea water?) forcing a change in configuration.
There are reasons to believe that a stable critical mass will not be reached; it's incredibly difficult to do and the uranium will tend to blow it's self apart immediately that happens meaning that no real nuclear explosion will happen. However, a little more humility and a whole bunch more circumspection would really help you maintain credibility for when you want to persuade people that the modern "safe" nuclear plants really are safe.
a couple of small nits.
RH requires a ridiculous support contract for access to patches.
Patches (as in source) are available for free; That's what CentOS bases off. It's the updates (as in binaries) which you pay for. If RedHat ceases to do this then I, and many like me, will start releasing them myself.
Finally, RH has only 4.8% of the packages Debian has, in a comparison of official stable repositories
Your comparison really isn't fair for several reasons. You are comparing only the core repositories of RH which are at a higher quality level than Debian's default repository precisely because they are more filtered. Try doing the same comparison with EPEL and/or RPMForge (which is where "community" packages are) included and you will see a more even story.
Red Hat sells support, not software, they would do well to remember that.
Both; to be honest; their Release CD images are still valuable. The fact that they limit access to the open access to the support forums is a complete pain even when we are fully paid up because most of our internal users aren't registered with RedHat and so we will never give direct access to all the people who might need it.
I'm not yet at all sure that this RedHat thing is a big problem; if it really is, then the correct solution is a license change at the kernel level. If RedHat can do this, then so can anyone else. That means that the Linux kernel people have to initiate, support and get involved in a GPLv4 update, using their "diplomatic skills" to persuade the FSF they are a partner worth working with. They have to concentrate on getting most known kernel developers to agree to a future license update and eliminating those that don't.
In the meantime, the key thing is to make sure that upstream of RedHat, more of these patches get merged in more effectively. This will have the benefit for RedHat of taking work off RedHat; will put more work into a better condition, will make the upstream old kernels more valuable and will provide wider hardware support for all Linux based on old kernels.. Oracle could put a bunch of kernel developers and testers on the case. If nobody is willing to do this, then, to be honest, RedHat is doing the right thing
In the meantime; even if you are a RedHat customer, concentrate on putting answers you find to problems onto Linux Questions Wiki or forums where they will be available under the CC-SA for everyone.
How about, when stupid people put most of android under a license other than the GPL and they will release the GPL parts?
Because the grandparent was completely Astroturf and moderated down for a good reason. By replying to it you are increasing the discussion about it and doing the work for the astroturfer. You are directing the discussion away from Microsoft's need to use lawsuits in a market where they can't compete through lack of platform and towards a minor issue; the quality of the development tools on WP7; which almost nobody here, let alone normal users, will ever care about.
Trolling is exactly trying to misdirect discussion; you have succeeded (I'm even being trolled here, replying to you, but it's an important point as long as people like your other respondents try to pretend that this is some kind of anti-Microsoft campaign). Your moderation as a troll was 100% correct, whatever your intentions.
Is it? I think that people know how to do forwarding etc. etc.
It seems to me that it's actually easier to block all executable content (flash / javascript etc) and then block file upload/download to / from the browser than it would be to find every possible https based mail service (including my own secret one; which is used only by me personally and even that almost never) which is what you would have to do in order for this to make sense.
No; this is fundamentally different. Consecutive life sentences affect a person. This is a sentence against a company. There was no problem charging some poor ignorant file sharer millions, but over-finling a company is an insult to justice.
So; The grandparent arrives suddenly having never posted before, is spending enough time refreshing to get a first post. I think we find that suspicious. We think that accusing WrongSizeGlass of trolling must be unfair and outrageous
But then let's look at the moderation on this (if you don't have an account you will want to sign up for one to see this stuff).
Starting Score: 1 point
Moderation -2
70% Troll
30% Insightful
Extra 'Troll' Modifier
If the moderation is 70/30 then the only possible solutions are multiples of 7 and 3 so at least three insightful moderations have been given. Now, the first post has some interesting comment. It's definitely a reflection of Microsoft's paranoia. However, there are too many things which are wrong which have been covered too often on Slashdot for any serious mod to mod it insightful. In any case, almost all legitimate positive moderation shows variety (underrated / interesting / insightful).
Looking at the facts there seem to be two possible explanations; a) the GNAA has returned having learned to troll subtly, pretending to be Microsoft Astroturfers and never ever linking to Goatse whilst building an elite undercover uber-posting super-moderators who can afford to give them mod points without risking losing the ability to moderate b) Microsoft is an immoral deceptive company using publicity people to spread lies and astroturfing like mad.
Obviously b) which would mean we do need a way to mod astroturfers is untenable so only a) can be true. Microsoft is an upstanding tax paying pillar of society. They would never ever ever astroturf and lie and cheat. WrongSizeGlass's moderation as troll must be justified.
Your post is insightful but I'll pick one little nit.
Copyright is the basis of the software industry, both closed and open source.
The basis of the software industry is people's need for software. If there was no copyright protection business models would have to change, but there would definitely still be customers paying people to create software. Free software would likely do fine, not on the basis of copyleft but simply on the basis of fair sharing. Software users would pay for maintenance and quality control or would learn to download from specific sites and do their own.
Definitely there would be some disadvantages; proprietary software companies would likely rely much more on copy-protection software (unless that was made illegal; a definite thought) and a small number of companies that contribute to software due to legal pressure would not. On the other hand there would be advantages; reverse engineering would advance much faster since there would be no legal restrictions on it and the end results would be immediately usable. This would in turn lead to much more secure software since even closed modules would be subject to serious scrutiny.
I'm not sure that total elimination of copyright is a good idea, but it's very important to remember that copyright is a bargain restrictions on freedom of speech are paid for by a promise that in future there will be more varied speech available. When one side stops keeping up their end of the bargain by using DRM which keeps works locked in even after the term of copyright expires, by continually extending copyright terms, by not delivering source code to proprietary programs even after they have ceased to be fully maintained and by attempting to use other limits on free speech such as patents then the time has arrived to question copyright full stop.
The "general public" may not care, but, before they buy they normally ask those of us who know and do care. When we tell them to avoid a "locked down" phone they probably don't even know what it means, they just know I think it's bad. Previously these people bought HTC instead, now I will direct them elsewhere.
No, your understanding of cancer is far too simple to be useful in this application.
Even my understanding of radiation was too simple/simplified to correctly model that. When one single particle of radiation passes through the body it is possible for it to cause multiple interactions; e.g. a gamma ray could hit one single cell's DNA at multiple points. That's before we take into account that a high energy particle, e.g. a single cosmic ray, could cause a cascade of different particles.
My point; my application; was not to say that any radiation above two strikes is "dangerous". It was to say that a) there is no threshold level; risk starts from almost the beginning and b) the risk is likely not linear and certainly depends on the way the dose is delivered. Statement a) matches the 2005 NAS report which states clearly:
There is no safe level or threshold of ionizing radiation exposure.
As for your comment:
There was a crazy guy who thought exactly two "strikes" would make a huge difference, but he was debunked in a series of studies years ago.
Direct links to actual scientific studies or even just some names to search for really greatly appreciated. Most of us have a scientifically educated layman's interest in this subject but not a postgrad student's time to research this ourselves.
Definitely a really nice chart. It's good to see something so easy to read and quantitative that helps people debate with some level of knowledge. The main problems for me with it are that it doesn't really do a good job on the time axis, spacial axes and the probabalistic risk. For example:
What makes this all difficult is that it seems the mechanisms are random. E.g. most of the time a particle of radiation does nothing. It dissociates a water molecule which soon after re-associates. Even if it does cause a mutation, that likely doesn't cause cancer because the body copes with mutation all the time and genetic codes self correct. However, if two or more mutations happen in close together / related genetic material in the same cell, that is reasonably likely to cause cancer as the cell is no longer able to self-correct. Now of course, this means that the "Lowest one-year dose clearly linked to increased cancer risk" is actually incorrect; that minimum ("clearly linked to increased", not to "noticeable") is about two particles of radiation where clearly is defined as "we clearly understand that this is so and "increased" is defined as "greater than would be otherwise. However, the minimum yearly dose "linked to a worrying increase according to a reasonable probabalistic model" is what we really want to know and is completely missing from the chart.
Since the location of radiation damage is entirely random, that can mean that millions of particles could cause no damage to one person whilst just three could damage another very unlucky person. This risk gets higher the more concentrated in space and time a dose of radiation is. When you think about it, the reason is obvious. The chance of a repeat strike in the same cell goes up quadratically as the volume shrinks and factorially as the dosage increases. These are the crucial things which mean that radioactive iodine and back scatter scanners are likely to be much more dangerous than e.g. cosmic ray exposure at altitude or through body X-rays. They are also mean that having a back scatter X-ray just before or after travelling is (I have no idea exactly how much) worse than having the X-ray on its own.
It would be really great if xkcd could do something which did a comparison of the dangers of different kinds of radiation exposure in different circumstances. Very important would be to leave in the ares of doubt where we actually don't know.