Sure, but I assume you're going to read it, right?
But something's really wrong if there are authors who don't even read the paper that's going out with their name on it.
There are two issues in authorship, first there is the issue of credit, second there is the issue of endorsement. To be an author you should qualify on both counts.
The real problem at Lucent was that there were people on the author list who did not qualify on either ground. They did not contribute to the work and they afterwards denied that they endorsed it.
Fraud is rare, but experimental error is common. The real issue is whether you have someone reading the paper and looking at the experimental results the equipment and the method with a skeptical eye.
When I was a postdoc at Argonne National Lab, my group's policy was that every group member got his name on every paper. On this [arxiv.org] paper, one of my coauthors refused to read the paper before publication. He said he was busy, and it was a long paper.
The people who insist on such models tend to have a very parochial view of science.
If I spend 5 years designing an experimental apparatus and gather data then a collegue (or more likely his grad students) takes that data and produces an analysis I have the right to have my name on the paper, the analysis is only a part of the work.
Science is based on trust. Consider the case where a physicist wants to measure some effect but does not know how to build the apparatus to test the theory. I might well design the apparatus to test his theory even though I don't fully understand the theory he is testing, ultimately I have to trust him on that point. Equally if I provide him with a bunch of experimental data he trusts me not to have fabricated it.
On the large experiments (500+ authors) I have worked on there has been a review committee that checked over the paper in detail of 30 or so people.
You cannot get any kind of legal protection for ideas. You can only protect the tangible expression of specific ideas.
The idea of one click shopping has been patented. As have many ideas. The theory of patent law bears no relationship to the actual corrupt practice.
This is a problem with patent EXAMINATION, not with patents themselves.
No, it is a problem with the USpTO, I have no problem with any of the European or in ternational patent offices. The specific problem is the US examination system which uniquely has no period of public review and opposition.
Two points of issue here. First, particle physics research over the last fifty years is what made the solid-state revolution
That is untrue. There are no sub-nuclear effects that are employed in semi-conductors. Rutherford and Chadwick's work on the structure of the atom was all pre-war and that is what the physics of semiconductors is based on. I have degrees in electronics and particle physics, the quantum mechanics used in solid state uses a completely different notation to that used in particle physics.
Only through linear accelerators was the crystal structure and properties of semiconductors properly worked out
You were almost close. Accelerators are used for chrystalography but not lineacs, It is the bremstralung radiation that you get from accelerating a charged particle arround a curve that provides the high powered radiation. The crystaline structure of silicon GaAs etc are all very simple and were deduced long before quantum mechanics, let alone particle physics. I very much doubt that any crystal structure that had such a high degree of complexity it could only be deduced using those techniques would be far too complex to be useful for VLSI.
As for your fusion power comment, I must say that, for one thing, there hasn't been too much funding because there hasn't been too much progress.
How much progress has there been in particle physics? Why would basic research into the fusion process be intrinsically less interesting than finding out the structure function of the z0 etc?
Trade secrets do little to promote the progress of science. They work more of a hindrance.
A trade secret on an obvious idea does not impede my use of that obvious idea. A wrongly issued patent does.
The problem with the bulk of the software patents issued by the USPTO in recent years (and by bulk I mean 95%+ of those I have read) is that they are completely obvious to anyone who has an understanding of the field.
The legal standard of obvious is different - except of course when the USPTO attempts to justify its racket when the 'non-obvious' standard is held up as the guarantee of fairness.
I have never once read a patent to get a good idea. The only reason I read patents is to make sure that I do not use the technology described by mistake.
One interesting element about these three chaps is that when they had their great ideas there was no way to make money from it so no-one is interested.
Darwin's return from his voyage on the Beagle was anticipated by the leading naturalists of his day. He was an instant celebrity and rapidly became one of the leading naturalists of his day. The origin of species had the effect it did largely because it came from someone who was already established as a major scientific author.
While Darwin was financially secure, Larmark whose work he often criticised had made his living from science and a good one too.
Newton was also well off, but his recent predecessors Gallileo and daVinci had made their living from patronage.
Einstein worked at a patent office for a short time but rapidly found a university position.
All three were known internationally within a few years of their first major publications.
Your post was going so well... but then you had to bring up cold fusion. If it has been replicated, can you give references? And not just a google search, but preferably point to a properly peer-reviewed journal article.
Actually there is a genuine 'cold fussion' effect, just not the bogus Fleicheman and Pons variety.
If you take a muon and put it in orbit arround a tritium atom in place of the electron it is much heavier and thus orbits much closer... Ahh just read the article...
The real outrage that Fleiechman and Pons did was to discredit a whole line of research with their actions. There is no reason why cold fusion should be rejected as impossible just because the field has attracted cranks. Before Harrison the search for longitude was the domain of cranks and lunatics.
BTW the big problem with muon catalysed fusion is that creating muons is an energy intensive process and the muons decay rapidly so getting one to decay is hard. However one could imagine using a muon catalysed reaction as the ignition chain for a self-sustaining reaction.
If your science is purely defined by public popularity, though, you'd better hope that the public stays interested in biology.
We haven't had the equivalent of a public relations disaster for biology yet, which would cause public opinion to turn against it. All you need is a biological Chernobyl and you'll be tarred with the same brush that physicists have had applied to them.
Well if scientists want public funding for billion dollar experiments they had better have public support...
However one of the interesting aspects of the physics funding situation is the way that there is a seemingly inexhaustible supply of funds for large scale particle physics colliders which have no practical application I am aware of while fusion research has to scramble for every dollar while trying to solve the energy problem. This is in large part due to the fission mafia's attempt to sink every competing energy source the way they killed public funding for alternative energy research in the 80s (see Salter's duck).
I think though that Chernobyl has less to do with the problems of physics than the end of the cold war. The gravy train for physics research and in particular particle physics had everything to do with the national prestige attached to 'nuclear' research which had everything to do with the bomb. That is why the US just had to build their own SSC and build it in Texas rather than work on the LHC at CERN or site the project close to Canada who had offered to provide the power if that happened.
Nulcear power was a spent commercial force after three mile island. The incredible stupidity of siting a nuclear plant that close to Manhattan island exposed the industry as negligent and careless. By the time that Chernobyl happened nuclear power was already dead.
The problem with the life sciences is finding out if a Chernobyl has occurred. It took several decades before DDT was identified as the cause of the declining populations of perdatory birds. It took even longer to connect smoking to cancer and heart disease. The problem with genetically modified foods is that nobody knows what adverse effects may be linked to them in 20 years time.
This link is of course unproven, but it is not exactly disproven either, nor is the alternative cut hunter theory particularly persuasive given that there have been cut hunters eating monkeys since Homo Spaiens appeared on the planet but AIDS has only appeared in the past 50 years.
The 'scientific' response to this theory would be to examine it as a matter of urgency. Instead the reponse from the biologists has been pretty much the response of the physicists to Chernobyl; no not us, could not possibly happen here, no three mile island was not comparable, it was the fault of those heathen communists, etc.
Instead of examining the polio vaccine theory it was silenced by means of a law suit brought by Koprowski, the leader of the polio trial. I do not consider that to be an adequate standard of scientific proof.
The author was of the opinion that the life sciences are not as rigorous in testing the veracity of research results.
That is untrue. The physicists I worked with were extreemly negligent in their verification procedures. As my college tutor, Tony Hoare put is, Physicists used to repeat each other's experiments, now they run each other's code.
If you take a hard look at the quality of the creaky FORTRAN decks used to analyse the results of the billion dollar experiments the physicists get you will see what I mean. All four experiments at LEP used the same GEANT monte-carlo simulation code and PAW analysis code. This at a time the code would not even compile unless the compiler warnings were turned off.
Sorry, until the physicists put their house in order they simply have no right to go attacking any other groups.
What physicists are annoyed at is that their research interests are soley judged by the potential amount of money it can make.
And the LHC is expected to make how much money? The claim is completely ridiculous. Only physicists get billion dollar experiments. I am not aware of any billion dollar experiment that was justified on a commercial basis. In comparison intellectual property claims are often the alpha and omega in the life sciences.
I don't think so. I know that physics labs have less money now than before and that the public grants have gone to life sciences. And some people at (for instance) the CERN are quite pissed off by that. And I understand them. But then I work in the life sciences, so...
I worked at CERN for 2 years. Let us get something straight here, the lab budget was over a billion dollars a year. LEP had not one but four versions of the same experiment, each costing over a billion to build and far more to operate.
The only reason there were four experiments at LEP rather than two was politics.
The fact that CERN now trumpets itself as the birthplace of the Web rings hollow to those of us who worked on the Web at CERN. The Web reseach at CERN was closed down by the physicists because they were jealous of the press the Web got for a project outside CERN's 'core mission'. So much for the value of interdisciplinary research!
Life sciences have incredible potential, anti-cancer drugs, anti-viral drugs, gene therapy, replacement parts. Why shouldn't that potential be measured against the value of measuring Z0 structure functions to an extra decimal place?
Harassing the lawyer for doing his job is another step altogether. If he himself is harassing people, that's one thing; if he's just protecting the spammer's rights, he's doing his job. For that matter bear in mind that the law frowns on self-help generally.
I don't buy the argument that people are not responsible for what they do for a living.
Lawyers who file suits to help the clients deprive others of rights puts them lower than the spamers in my view.
It is one thing to provide a defense for a scumbag client, quite another to help them attack.
Just my opinion on the profession as a whole rather than this instance.
In this case, no one is trying to obtain money or property. Hence, no mail fraud.
Bingo, fraud is not lying, fraud is lying with a very specific intent, material gain.
When a lawyer files a crank suit for someone it is rarely the case that they go file the wrong crank suit. Filing a civil crank suit is much less likely to lead to problems than filing a criminal one.
However the guy is undoubtedly full of it. How does he claim to know who put him on the mailing lists? OK he can file a suit against John Doe #1 through 69, but recovering damages against them is not going to be happening.
The real question that should be asked is 'is the space station justified at all', not merely whether it could be done slightly cheaper. The project would still be overpriced at $5 billion.
Consider that the SSC would have provided far more science for $10 billion. Or for that matter consider how much science we could get by sending up a duplicate of Hubble - many of the parts exist already as test pieces for the orbitting Hubble, the test mirror made by Kodak was actually done right.
Or consider what a boost to the economy we could get by giving the same money to rich corporate campaign contributors. $40 billion is more than the retrospective tax handouts that Bush wanted to give Enron.
Or even (gasp) think what could be done if the same amount had gone into other research areas such as biotech or the Internet. There is a reason the Web was born at CERN, they had the resources to do that type of work.
The economist had a good article recently where they speculat that NASA asked Nixon for funding for a mars mission and got rejected, so they split the mission into three parts, first a reusable space shuttle, then a space station, finally a mars mission.
Since then the obvious conclusion to draw from the success of the unmanned missions is that they are cheaper and result in more science.
On the other hand though, how hard would it be for Microsoft to just give the option upon install of not installing these components?
I suggested to Bill's people that they produce a version of Windows called Windows DS (Dissenting States) Edition. This would be exactly the same as normal XP except that the media player, browser etc shells would not be there (but the dlls they access would be since they are pretty fundamental).
My guess is that absolutely noone would buy it since the idea of getting half a loaf was never something the consumers were demanding, it was the software houses.
No response yet on that one, although Bill did tell me in an email that in a move to demonstrate his appretiation of the open source movement, Melinda is going to cook a penguin for Christmass dinner.
Wrong! The reason PCs are cheap today is because of IBM cloning and has nothing to do with Microsoft.
Cloning is only possible because IBM lost control over the software platform. If IBM had succeeded in foisting OS/2 and microchannel on the industry the clones would have been killed.
Of course nobody was going to let OS/2 win for exactly that reason. The industry choose Microsoft because IBM was a much bigger threat at the time.
If Lotus and Wordperfect had figured this out a few years before it happened rather than many years later they might have survived.
When PC's cost $4000 Windows cost $100. Now that comodity PC costs ~$800 Windows still costs $100.
If there was competition in software as there is in hardware Microsoft would have had to bring their price down.
Or add a lot more to the product which is exactly what they did, unless you have the quaint belief that there is absolutely no difference between MSDOS and Windows.
Measured by features Windows is pretty cheap. And expect the price of Windows to start to decline as PCs drop further in price. The OEM price of windows is much less than the retail box price already. Microsoft is already seeing erosion in its prices as good mainstream PCs are now available at the $750 mark and discount models are $500 or less.
Well thats good to know. I'm sure that Sun is feeling pretty good about themselves right now, and at Microsoft's expense, who I bet is busy rethinking strategy after reading your comment.
I wrote most of the SAML specs which are the basis of the Liberty design. I really wish that people would stop trying to define the problem as one company bashing another.
I have absolutely no interest in the issue of whether Sun can stop Microsoft or Microsoft can stop Sun. I have been trying to deploy global authentication schemes for ten years now, I believe that the problem is sufficiently hard that it is not going to be solve by any party that makes its primary objective the defeat of another party.
First off lets recognize that companies working together can be a good thing for the consumer and can also be a bad thing. It is good when stuff works together, it is bad when working together effectively means a cartel.
I don't fault Microsoft for using their deployed base to build the user base for passport. After all AOL did the same thing by buying up rival instant messaging services.
What I do not see is how any party can reasonably expect the idea of global authentication to turn into some sort of monopoly. The competative forces involved are just too great.
Consider the problem of getting access to my frequent flyer plan at United. It would be pretty handy if I could simply log on to United transparently through my browser without having the browser store lots of personal data on my machine that could itself be a security vulnerability. On the other hand I don't see United paying anyone $10 per year for the privillege of offering this facility or anything like it.
Now consider what happens if we have 50 single sign on schemes, I don't see any advantage over having separate log ins.
So there has to be a critical mass for any of these schemes to be worthwhile, there has to be a reasonable cost structure and there has to be confidence that the operators of the scheme will not impose new costs or hidden restrictions at a future date.
I think that there is a value here but I think that both Liberty and Passport need to be radically rethunk before either can achieve the stated goals.
Before that happens however I think that there has to be a political realignment. In particular I think we need to get Liberty to stop promoting itself as a 'stop Microsoft' scheme and we need Passport and Liberty to agree to some form of convergence in the same way that Visa and Mastercard converged.
Specifically we should adopt SAML as the underlying architecture for global authentication. The ability to carry kerberos tickets and passport credentials is already designed into the SAML specs.
Once there is agreement on a technology base Liberty and Passport would both evolve into federated authentication brands in the same way that Mastercard and Visa have. There would be a strong assumption that merchants and web sites would support both brands rather than expecting consumers to cope with both sets of credentials.
Finally we need to work out who is going to actually pay for such a system to be established. Charging end users is really hard, charging merchants cuts out sites like slashdot. Where is the compelling value proposition? I believe that there is one to be found but we have not got there yet.
Anyway, if IBM can give away stuff like that (he got it at the end of a DB2 course), surely they could make some kind of simple USB smartcard ?
Rainbow have been making a smartcard in a USB form factor for several years. Thats why the SAML spec on which Liberty is based mentions 'hardware tokens' rather than smart cards.
As a browser plugin ? That way the person decides who can access their information ? The keys, credit card and personal information/etc. are stored in an encrypted file on the machine and only those with permission can get at the information ?
No way shoulod you do that unless you are running trusted hardware. I don't care what encryption is used for the file, the keys are still vulnerable, as is the data itself when it is decrypted.
So yes that is a great design if you are using a Palladium class machine, but there are not many of those arround at present.
It would eliminate the need for a hugemungous server (run by an evil corporation) and this way it would be pretty simple to access the information (with some authentication of course) and not need to pay an arm and a left testicle to an Evil Corporation..
I find it difficult to take this type of approach seriously. I am much more worried about John Piondexter than I am about corporations at this point. I can't think off hand of any corporation who illegally sold arms to terrorists in Iran to fund more terorrists in Latin America.
That aside, what you call 'authentication' is no more and no less than what Passport and Liberty both do. You are thinking about the problem from one angle alone, confidentiality. That is a bad mistake when talking about an authentication scheme.
Disclaimer: IANAL. Can any US or Danish lawyers clear this up?
IANAL either, however one salient fact is likely to be that Denmark is not a common law country. So while UK law is markedly similar to US law, extrapolation from US to Danish law is going to be problematic.
Another major issue that should be considered is that most countries other than the US have severe penalties for abusive lawsuits, ranging from payment of the winner's legal fees to fines. Certainly in English law an ill founded threat of litigation over an intellectual property claim is itself actionable - as Prince sports discovered when they tried some domain name grabbing.
As such I would be suspicious of cliams that the group would have made the threats lightly or that they are likely to be ill founded. The copyright owners have a pretty sound claim that the users of the P2P networks are infringing their copyright. Clearly there is a legitimate case that can be built, clearly the copyright owners are employing local lawyers with expertise in the field. I certainly would not place a great deal of faith in chat room lawyering working from third hand accounts...
The judge requested public comments, and of the non-trivial (non MS hate mail) comments that were kept, two thirds were against the settlement. Despite that, the judge decided in favor of the settlement, with only minor changes!
The public comment period is a feature of the anti-trust act, it is pretty irrelevant since the judge is meant to be rulling on the law, not conducting opinion polls.
They tried to appeal to the Supreme Court. The supremes refused to hear them and bundled them off to their new penalty phase judge the Appeals court had arranged for them.
That is normal practice with the supremes, they rarely take cases unless all the issues have been settled in the lower courts. Declining to take the case at that point is not as you attempt to imply a judgement on the merits of the appeal.
Had the case continued there is little doubt that the supremes would have eventually heard the case, probably arround 2006 or so by which time it is doubtful that anyone would place much store in Jackson's decision to decline to hear most of the evidence in order to accelerate the pace of the trial.
It would be much more reasonable, to simply have a digital camera, and to turn the light on with the door closed so it could feed you a snapshot of what's in the fridge.
I thought of that, you must have an amazingly empty fridge. And how do you look inside a pizza box to decide whether it has four slices or one? And how do you tell if the Chinees takeout was from yesterday or last thursday?
OK, you can get Kerberos v5 with any Linux distro and it works very well. Kerberos was designed for UNIX. The variant on Kerberos v5 that is used in Windows beginning in Windows 2000 is a bastardized version, so much so it can't even communicate with the open Kerberos standard.
OK, fire up telnet and log into another machine. Oh dear you password just went over the wire in the clear. Same for your POP connection.
Running ssh in place of telenet is a good idea, but it is certainly not a transparent replacement. You have to go round disabling the telnet deamons, install ssh and tel the users you made the switch.
Agreed, Linux/BSD needs a better security model, but it's in the works. The NSA's Secure Linux has a fully working version that is even better than Windows permissions. And you can set up groups and do limited group policy in xNIX even now.
The implicit assumption you make here is that military security models are the most challenging, actually that is completely untrue as anyone who has used a label based secure O/S knows. Military security models are actually considerably simpler to design because you can simply order people to use them and you get a high degree of compliance. It does not work like that in the real world. NSA secure Linux is far less sophisticated than the dotNET security framework.
All the NSA patches do is to bring UNIX up to the state of the art circa 1980s.
People aren't going to want their phones tell their friends where they are, anymore than they want radio tags to tell their employers where they are.
I had great difficulty trying to explain to mobile phone companies that the idea of selling data on where their customers were was a non-starter businesswise.
The business problem comes into focus when you consider that politicians tend to be avaracious consumers of wireless technology (both the Bush and Gore campaigns used RIM pagers), they also have serious security concerns. The ones I talked to imediately realized that the technology would be abused by stalkers, nutcases and assasins.
I sure don't want my washing machine reporting back to its manufacturer, but I have no qualms with it determing, for itself, that it is no longer functioning under normal operational parameters.
I don't want my washing machine to call the manufacturer, but loging a problem with the home maintenance center would be OK. This is the sort of thing that Web services will be big for.
I think that devices will talk but they will talk inside the home first and any external communication will be with permission from the owner.
Having the washing machine tell me when it needs repair is not a killer app. A status light on the front panel can achieve the same result. But having the thing tell me when the cycle has completed and the clothes are ready to go into the dryer, that is useful - I am writing this three floors above the washing machine in the basement.
Equally, Negroponte's fridge that orders stuff itself is inescapably clueless. There are in fact 'fridges' of that type, we call them vending machines and lots of them are now wired to report their inventory levels so the guy in the truck knows when to go and fill them up.
However it would be reasonably usefull to be able to check the contents of the fridge from my handheld PC when I am in the store and wondering if I need to buy more OJ, milk, and eggs. I suspect that a usable system would involve weight sensors and perhaps some sort of barcode printer / scanner built onto the side of the fridge for the frozen stuff. Problem with any such system however is that the discipline required to use it tends to be too much for most.
Linux security isn't competetive with Windows!? Cue Rod Serling.
Only if you define security in the exceptionally narrow sense of 'relatively free of security holes'.
In terms of support for security features Windows is much more advanced. Windows has Kerberos deeply embedded into the O/S, in Unix Kerberos is an optional afterthought. Windows has full support for PKI built into the core of the O/S, no UNIX has to date. Windows has features like an encrypting file system built into the core file system, on UNIX you have various ad hoc schemes.
dotNET has an exceptionaly sophisticated security model that provides for fine grained authorization on application defined permissions. Linux has the same old two level privillege system of user/root.
Palladium provides a means by which a client application on a remote computer can convince another system that it is running a specific version of an application and that the machine is secured to a specific profile. That is kinda useful if you want to use that machine to store a credit card number on or the like...
Of course you can if you chose simply stick your head in the sand the way that Lotus, Apple and Wordperfect did. Of course your system is perrfect and will never be challenged no matter how much time and effort Redmond put into overtaking you. Go ahead Mr Hare take a nice long nap, the armoured tank is way off in the distance and only moving at 50 mph or so.
Slashdot Mirror
But something's really wrong if there are authors who don't even read the paper that's going out with their name on it.
There are two issues in authorship, first there is the issue of credit, second there is the issue of endorsement. To be an author you should qualify on both counts.
The real problem at Lucent was that there were people on the author list who did not qualify on either ground. They did not contribute to the work and they afterwards denied that they endorsed it.
Fraud is rare, but experimental error is common. The real issue is whether you have someone reading the paper and looking at the experimental results the equipment and the method with a skeptical eye.
The people who insist on such models tend to have a very parochial view of science.
If I spend 5 years designing an experimental apparatus and gather data then a collegue (or more likely his grad students) takes that data and produces an analysis I have the right to have my name on the paper, the analysis is only a part of the work.
Science is based on trust. Consider the case where a physicist wants to measure some effect but does not know how to build the apparatus to test the theory. I might well design the apparatus to test his theory even though I don't fully understand the theory he is testing, ultimately I have to trust him on that point. Equally if I provide him with a bunch of experimental data he trusts me not to have fabricated it.
On the large experiments (500+ authors) I have worked on there has been a review committee that checked over the paper in detail of 30 or so people.
The idea of one click shopping has been patented. As have many ideas. The theory of patent law bears no relationship to the actual corrupt practice.
This is a problem with patent EXAMINATION, not with patents themselves.
No, it is a problem with the USpTO, I have no problem with any of the European or in ternational patent offices. The specific problem is the US examination system which uniquely has no period of public review and opposition.
That is untrue. There are no sub-nuclear effects that are employed in semi-conductors. Rutherford and Chadwick's work on the structure of the atom was all pre-war and that is what the physics of semiconductors is based on. I have degrees in electronics and particle physics, the quantum mechanics used in solid state uses a completely different notation to that used in particle physics.
Only through linear accelerators was the crystal structure and properties of semiconductors properly worked out
You were almost close. Accelerators are used for chrystalography but not lineacs, It is the bremstralung radiation that you get from accelerating a charged particle arround a curve that provides the high powered radiation. The crystaline structure of silicon GaAs etc are all very simple and were deduced long before quantum mechanics, let alone particle physics. I very much doubt that any crystal structure that had such a high degree of complexity it could only be deduced using those techniques would be far too complex to be useful for VLSI.
As for your fusion power comment, I must say that, for one thing, there hasn't been too much funding because there hasn't been too much progress.
How much progress has there been in particle physics? Why would basic research into the fusion process be intrinsically less interesting than finding out the structure function of the z0 etc?
A trade secret on an obvious idea does not impede my use of that obvious idea. A wrongly issued patent does.
The problem with the bulk of the software patents issued by the USPTO in recent years (and by bulk I mean 95%+ of those I have read) is that they are completely obvious to anyone who has an understanding of the field.
The legal standard of obvious is different - except of course when the USPTO attempts to justify its racket when the 'non-obvious' standard is held up as the guarantee of fairness.
I have never once read a patent to get a good idea. The only reason I read patents is to make sure that I do not use the technology described by mistake.
Darwin's return from his voyage on the Beagle was anticipated by the leading naturalists of his day. He was an instant celebrity and rapidly became one of the leading naturalists of his day. The origin of species had the effect it did largely because it came from someone who was already established as a major scientific author.
While Darwin was financially secure, Larmark whose work he often criticised had made his living from science and a good one too.
Newton was also well off, but his recent predecessors Gallileo and daVinci had made their living from patronage.
Einstein worked at a patent office for a short time but rapidly found a university position.
All three were known internationally within a few years of their first major publications.
Actually there is a genuine 'cold fussion' effect, just not the bogus Fleicheman and Pons variety.
If you take a muon and put it in orbit arround a tritium atom in place of the electron it is much heavier and thus orbits much closer... Ahh just read the article...
The real outrage that Fleiechman and Pons did was to discredit a whole line of research with their actions. There is no reason why cold fusion should be rejected as impossible just because the field has attracted cranks. Before Harrison the search for longitude was the domain of cranks and lunatics.
BTW the big problem with muon catalysed fusion is that creating muons is an energy intensive process and the muons decay rapidly so getting one to decay is hard. However one could imagine using a muon catalysed reaction as the ignition chain for a self-sustaining reaction.
Well if scientists want public funding for billion dollar experiments they had better have public support...
However one of the interesting aspects of the physics funding situation is the way that there is a seemingly inexhaustible supply of funds for large scale particle physics colliders which have no practical application I am aware of while fusion research has to scramble for every dollar while trying to solve the energy problem. This is in large part due to the fission mafia's attempt to sink every competing energy source the way they killed public funding for alternative energy research in the 80s (see Salter's duck).
I think though that Chernobyl has less to do with the problems of physics than the end of the cold war. The gravy train for physics research and in particular particle physics had everything to do with the national prestige attached to 'nuclear' research which had everything to do with the bomb. That is why the US just had to build their own SSC and build it in Texas rather than work on the LHC at CERN or site the project close to Canada who had offered to provide the power if that happened.
Nulcear power was a spent commercial force after three mile island. The incredible stupidity of siting a nuclear plant that close to Manhattan island exposed the industry as negligent and careless. By the time that Chernobyl happened nuclear power was already dead.
The problem with the life sciences is finding out if a Chernobyl has occurred. It took several decades before DDT was identified as the cause of the declining populations of perdatory birds. It took even longer to connect smoking to cancer and heart disease. The problem with genetically modified foods is that nobody knows what adverse effects may be linked to them in 20 years time.
We may even have seen a Chernobyl already. There is an interesting correlation between the sites of the earliest identified cases of AIDS and the testing of polio vaccines cultured on monkey kidneys
This link is of course unproven, but it is not exactly disproven either, nor is the alternative cut hunter theory particularly persuasive given that there have been cut hunters eating monkeys since Homo Spaiens appeared on the planet but AIDS has only appeared in the past 50 years.
The 'scientific' response to this theory would be to examine it as a matter of urgency. Instead the reponse from the biologists has been pretty much the response of the physicists to Chernobyl; no not us, could not possibly happen here, no three mile island was not comparable, it was the fault of those heathen communists, etc.
Instead of examining the polio vaccine theory it was silenced by means of a law suit brought by Koprowski, the leader of the polio trial. I do not consider that to be an adequate standard of scientific proof.
That is untrue. The physicists I worked with were extreemly negligent in their verification procedures. As my college tutor, Tony Hoare put is, Physicists used to repeat each other's experiments, now they run each other's code.
If you take a hard look at the quality of the creaky FORTRAN decks used to analyse the results of the billion dollar experiments the physicists get you will see what I mean. All four experiments at LEP used the same GEANT monte-carlo simulation code and PAW analysis code. This at a time the code would not even compile unless the compiler warnings were turned off.
Sorry, until the physicists put their house in order they simply have no right to go attacking any other groups.
What physicists are annoyed at is that their research interests are soley judged by the potential amount of money it can make.
And the LHC is expected to make how much money? The claim is completely ridiculous. Only physicists get billion dollar experiments. I am not aware of any billion dollar experiment that was justified on a commercial basis. In comparison intellectual property claims are often the alpha and omega in the life sciences.
I worked at CERN for 2 years. Let us get something straight here, the lab budget was over a billion dollars a year. LEP had not one but four versions of the same experiment, each costing over a billion to build and far more to operate.
The only reason there were four experiments at LEP rather than two was politics.
The fact that CERN now trumpets itself as the birthplace of the Web rings hollow to those of us who worked on the Web at CERN. The Web reseach at CERN was closed down by the physicists because they were jealous of the press the Web got for a project outside CERN's 'core mission'. So much for the value of interdisciplinary research!
Life sciences have incredible potential, anti-cancer drugs, anti-viral drugs, gene therapy, replacement parts. Why shouldn't that potential be measured against the value of measuring Z0 structure functions to an extra decimal place?
I don't buy the argument that people are not responsible for what they do for a living.
Lawyers who file suits to help the clients deprive others of rights puts them lower than the spamers in my view.
It is one thing to provide a defense for a scumbag client, quite another to help them attack.
Just my opinion on the profession as a whole rather than this instance.
Bingo, fraud is not lying, fraud is lying with a very specific intent, material gain.
When a lawyer files a crank suit for someone it is rarely the case that they go file the wrong crank suit. Filing a civil crank suit is much less likely to lead to problems than filing a criminal one.
However the guy is undoubtedly full of it. How does he claim to know who put him on the mailing lists? OK he can file a suit against John Doe #1 through 69, but recovering damages against them is not going to be happening.
The real question that should be asked is 'is the space station justified at all', not merely whether it could be done slightly cheaper. The project would still be overpriced at $5 billion.
Consider that the SSC would have provided far more science for $10 billion. Or for that matter consider how much science we could get by sending up a duplicate of Hubble - many of the parts exist already as test pieces for the orbitting Hubble, the test mirror made by Kodak was actually done right.
Or consider what a boost to the economy we could get by giving the same money to rich corporate campaign contributors. $40 billion is more than the retrospective tax handouts that Bush wanted to give Enron.
Or even (gasp) think what could be done if the same amount had gone into other research areas such as biotech or the Internet. There is a reason the Web was born at CERN, they had the resources to do that type of work.
The economist had a good article recently where they speculat that NASA asked Nixon for funding for a mars mission and got rejected, so they split the mission into three parts, first a reusable space shuttle, then a space station, finally a mars mission.
Since then the obvious conclusion to draw from the success of the unmanned missions is that they are cheaper and result in more science.
I suggested to Bill's people that they produce a version of Windows called Windows DS (Dissenting States) Edition. This would be exactly the same as normal XP except that the media player, browser etc shells would not be there (but the dlls they access would be since they are pretty fundamental).
My guess is that absolutely noone would buy it since the idea of getting half a loaf was never something the consumers were demanding, it was the software houses.
No response yet on that one, although Bill did tell me in an email that in a move to demonstrate his appretiation of the open source movement, Melinda is going to cook a penguin for Christmass dinner.
Cloning is only possible because IBM lost control over the software platform. If IBM had succeeded in foisting OS/2 and microchannel on the industry the clones would have been killed.
Of course nobody was going to let OS/2 win for exactly that reason. The industry choose Microsoft because IBM was a much bigger threat at the time.
If Lotus and Wordperfect had figured this out a few years before it happened rather than many years later they might have survived.
Or add a lot more to the product which is exactly what they did, unless you have the quaint belief that there is absolutely no difference between MSDOS and Windows.
Measured by features Windows is pretty cheap. And expect the price of Windows to start to decline as PCs drop further in price. The OEM price of windows is much less than the retail box price already. Microsoft is already seeing erosion in its prices as good mainstream PCs are now available at the $750 mark and discount models are $500 or less.
I wrote most of the SAML specs which are the basis of the Liberty design. I really wish that people would stop trying to define the problem as one company bashing another.
I have absolutely no interest in the issue of whether Sun can stop Microsoft or Microsoft can stop Sun. I have been trying to deploy global authentication schemes for ten years now, I believe that the problem is sufficiently hard that it is not going to be solve by any party that makes its primary objective the defeat of another party.
First off lets recognize that companies working together can be a good thing for the consumer and can also be a bad thing. It is good when stuff works together, it is bad when working together effectively means a cartel.
I don't fault Microsoft for using their deployed base to build the user base for passport. After all AOL did the same thing by buying up rival instant messaging services.
What I do not see is how any party can reasonably expect the idea of global authentication to turn into some sort of monopoly. The competative forces involved are just too great.
Consider the problem of getting access to my frequent flyer plan at United. It would be pretty handy if I could simply log on to United transparently through my browser without having the browser store lots of personal data on my machine that could itself be a security vulnerability. On the other hand I don't see United paying anyone $10 per year for the privillege of offering this facility or anything like it.
Now consider what happens if we have 50 single sign on schemes, I don't see any advantage over having separate log ins.
So there has to be a critical mass for any of these schemes to be worthwhile, there has to be a reasonable cost structure and there has to be confidence that the operators of the scheme will not impose new costs or hidden restrictions at a future date.
I think that there is a value here but I think that both Liberty and Passport need to be radically rethunk before either can achieve the stated goals.
Before that happens however I think that there has to be a political realignment. In particular I think we need to get Liberty to stop promoting itself as a 'stop Microsoft' scheme and we need Passport and Liberty to agree to some form of convergence in the same way that Visa and Mastercard converged.
Specifically we should adopt SAML as the underlying architecture for global authentication. The ability to carry kerberos tickets and passport credentials is already designed into the SAML specs.
Once there is agreement on a technology base Liberty and Passport would both evolve into federated authentication brands in the same way that Mastercard and Visa have. There would be a strong assumption that merchants and web sites would support both brands rather than expecting consumers to cope with both sets of credentials.
Finally we need to work out who is going to actually pay for such a system to be established. Charging end users is really hard, charging merchants cuts out sites like slashdot. Where is the compelling value proposition? I believe that there is one to be found but we have not got there yet.
Rainbow have been making a smartcard in a USB form factor for several years. Thats why the SAML spec on which Liberty is based mentions 'hardware tokens' rather than smart cards.
No way shoulod you do that unless you are running trusted hardware. I don't care what encryption is used for the file, the keys are still vulnerable, as is the data itself when it is decrypted.
So yes that is a great design if you are using a Palladium class machine, but there are not many of those arround at present.
It would eliminate the need for a hugemungous server (run by an evil corporation) and this way it would be pretty simple to access the information (with some authentication of course) and not need to pay an arm and a left testicle to an Evil Corporation..
I find it difficult to take this type of approach seriously. I am much more worried about John Piondexter than I am about corporations at this point. I can't think off hand of any corporation who illegally sold arms to terrorists in Iran to fund more terorrists in Latin America.
That aside, what you call 'authentication' is no more and no less than what Passport and Liberty both do. You are thinking about the problem from one angle alone, confidentiality. That is a bad mistake when talking about an authentication scheme.
IANAL either, however one salient fact is likely to be that Denmark is not a common law country. So while UK law is markedly similar to US law, extrapolation from US to Danish law is going to be problematic.
Another major issue that should be considered is that most countries other than the US have severe penalties for abusive lawsuits, ranging from payment of the winner's legal fees to fines. Certainly in English law an ill founded threat of litigation over an intellectual property claim is itself actionable - as Prince sports discovered when they tried some domain name grabbing.
As such I would be suspicious of cliams that the group would have made the threats lightly or that they are likely to be ill founded. The copyright owners have a pretty sound claim that the users of the P2P networks are infringing their copyright. Clearly there is a legitimate case that can be built, clearly the copyright owners are employing local lawyers with expertise in the field. I certainly would not place a great deal of faith in chat room lawyering working from third hand accounts...
The public comment period is a feature of the anti-trust act, it is pretty irrelevant since the judge is meant to be rulling on the law, not conducting opinion polls.
They tried to appeal to the Supreme Court. The supremes refused to hear them and bundled them off to their new penalty phase judge the Appeals court had arranged for them.
That is normal practice with the supremes, they rarely take cases unless all the issues have been settled in the lower courts. Declining to take the case at that point is not as you attempt to imply a judgement on the merits of the appeal.
Had the case continued there is little doubt that the supremes would have eventually heard the case, probably arround 2006 or so by which time it is doubtful that anyone would place much store in Jackson's decision to decline to hear most of the evidence in order to accelerate the pace of the trial.
I thought of that, you must have an amazingly empty fridge. And how do you look inside a pizza box to decide whether it has four slices or one? And how do you tell if the Chinees takeout was from yesterday or last thursday?
OK, fire up telnet and log into another machine. Oh dear you password just went over the wire in the clear. Same for your POP connection.
Running ssh in place of telenet is a good idea, but it is certainly not a transparent replacement. You have to go round disabling the telnet deamons, install ssh and tel the users you made the switch.
Agreed, Linux/BSD needs a better security model, but it's in the works. The NSA's Secure Linux has a fully working version that is even better than Windows permissions. And you can set up groups and do limited group policy in xNIX even now.
The implicit assumption you make here is that military security models are the most challenging, actually that is completely untrue as anyone who has used a label based secure O/S knows. Military security models are actually considerably simpler to design because you can simply order people to use them and you get a high degree of compliance. It does not work like that in the real world. NSA secure Linux is far less sophisticated than the dotNET security framework.
All the NSA patches do is to bring UNIX up to the state of the art circa 1980s.
I had great difficulty trying to explain to mobile phone companies that the idea of selling data on where their customers were was a non-starter businesswise.
The business problem comes into focus when you consider that politicians tend to be avaracious consumers of wireless technology (both the Bush and Gore campaigns used RIM pagers), they also have serious security concerns. The ones I talked to imediately realized that the technology would be abused by stalkers, nutcases and assasins.
I sure don't want my washing machine reporting back to its manufacturer, but I have no qualms with it determing, for itself, that it is no longer functioning under normal operational parameters.
I don't want my washing machine to call the manufacturer, but loging a problem with the home maintenance center would be OK. This is the sort of thing that Web services will be big for.
I think that devices will talk but they will talk inside the home first and any external communication will be with permission from the owner.
Having the washing machine tell me when it needs repair is not a killer app. A status light on the front panel can achieve the same result. But having the thing tell me when the cycle has completed and the clothes are ready to go into the dryer, that is useful - I am writing this three floors above the washing machine in the basement.
Equally, Negroponte's fridge that orders stuff itself is inescapably clueless. There are in fact 'fridges' of that type, we call them vending machines and lots of them are now wired to report their inventory levels so the guy in the truck knows when to go and fill them up.
However it would be reasonably usefull to be able to check the contents of the fridge from my handheld PC when I am in the store and wondering if I need to buy more OJ, milk, and eggs. I suspect that a usable system would involve weight sensors and perhaps some sort of barcode printer / scanner built onto the side of the fridge for the frozen stuff. Problem with any such system however is that the discipline required to use it tends to be too much for most.
Only if you define security in the exceptionally narrow sense of 'relatively free of security holes'.
In terms of support for security features Windows is much more advanced. Windows has Kerberos deeply embedded into the O/S, in Unix Kerberos is an optional afterthought. Windows has full support for PKI built into the core of the O/S, no UNIX has to date. Windows has features like an encrypting file system built into the core file system, on UNIX you have various ad hoc schemes.
dotNET has an exceptionaly sophisticated security model that provides for fine grained authorization on application defined permissions. Linux has the same old two level privillege system of user/root.
Palladium provides a means by which a client application on a remote computer can convince another system that it is running a specific version of an application and that the machine is secured to a specific profile. That is kinda useful if you want to use that machine to store a credit card number on or the like...
Of course you can if you chose simply stick your head in the sand the way that Lotus, Apple and Wordperfect did. Of course your system is perrfect and will never be challenged no matter how much time and effort Redmond put into overtaking you. Go ahead Mr Hare take a nice long nap, the armoured tank is way off in the distance and only moving at 50 mph or so.