My God! You are right! SCO have forced SGI to admit that core, critical components of Linux have actually been ripped out of System V! I mean, take one example from the massive 200 lines of offending code: the atoi(3) function. Yep, a function of such mind-blowing sophistication and complexity that it takes a NULL-terminated string argument and converts it to an integer. It is true, the Linux community lacks the technical ability to reproduce this function, hence the necessity to rip it from System V. Yes, this really is a tragic day for Linux.
Perhaps there is already a freely-available atoi(3) implementation that can be used by the Kernel? Yes, in fact there is! Even better, it coincides with the atoi(3) function that was removed! Woohoo! We are saved! There is not even eny need to look further afield, like the multitude of *other* free implementations around the place (say, glibc? BSD?), or (at last resort) actually finding a master C programmer to lock himself in a room for the long winter months in an effort to come up with an independent implementation.
The funniest thing is, that it probably WOULD take a huge effort by a master programmer to actually find an implementation of atoi(3) that was truely independent;-)
Linux has run natively on 64 bit architectures (Alpha, SPARC, MIPS etc etc) since almost the beginning. All of the GNU tools are 64-bit clean and always have been, AFAIK. I know for sure Gnome compiles without change for 64-bit arch, I would assume KDE does as well.
Each card has a unique ID number linked to the registered owner's name, which is recorded together with the location and time of the exchange every time the card is used.
As broken as the software patent system is, I find it hard to have much sympathy for Microsoft here. Bill Gates must know as well as any programmer (and I use the word 'programmer' in its loosest possible sense) that software patents are a flawed concept from the beginning, yet he deliberately set out to ride the system and push the patent boundaries well beyond resonable limits.
They are stuck between a rock and a hard place now. It sounds like there is a viable business model out there of "patent something really obvious then sue Microsoft for patent violation". As the biggest fish out there they are clearly the most attractive target, given a sufficently 'strong' case. On the other hand, Microsoft's patent portfolio is its ultimate trump card and which they are presumably saving for the final defence, in the event that open source starts to seriously affect their viability. They can't suddenly start lobbying to remove patent protection without invalidating their own portfolio too.
By default, duty is payable on anything worth more than $x (can't remember the exact figure). But it isn't payable on personal effects, so you could easily have claimed the duty back.
I am not sure what my response would have been. Saying "what the fuck is that?" on national TV is always going to be a risk, as you don't know the extent to which the interviewer is prepared to go to set you up. (Ignoring for the moment that I have been aware of the DHMO parody for several years now;) - when put on the spot, with the TV cameras there, and the implied threat that if you make a mess of it and say something silly it will be broadcast on national TV, I would probably just end up saying the first thing that comes into my head.
Perhaps it is a cultural difference (I am not from the US), but if I was at a protest rally, and someone was collecting signatures for a petition on a plausible-sounding environmental issue, I would assume that they were genuine. I don't mean that I would blindly sign my name on anything, but given a conversation like "Hi! Do you want to sign this petition to ban dihydrogen monoxide?" "What is dihydrogen monoxide?" "It is a dangerous chemical that kills many people every year, and bad companies like Monsanto are dumping it in rivers!" "OK! I'll sign!" I would assume that they were not deliberately setting out to mislead me.
I don't think it is valid to compare an activist with a used car salesman. The activist is doing what they think is right for some (presumably noble) cause. The used car salesman is doing what he thinks will give him a quick dollar.
Of course, that doesn't mean that any of the parties actually have a clue, but the intent is very different at least.
The injunction in Germany was about SCO making claims about linux but refusing to show any evidence to back those claims up.
Basically, they have been told to "put up or shut up", and they did neither, hence the fine.
It does not mean that a court has found SCO is not telling the truth, it just means that if SCO want to continue making big statement, they need to accompanying evidence.
Aww, c'mon this is pure troll, even if you didn't intend it to be.
Pick a person at random and ask them whether they support a ban on a very dangerous chemical substance named 'dihydrogen monoxide'. Anyone who isn't a chemistry graduate will have to think for at least a second or two to figure out what this is, and people with no chemistry background would likely not figure it out at all (not without more information, at least).
Now, practically all of what you consider 'activists' are ordinary people, who happen to believe that Greenpeace may, on some level, have a point. Why should these people be any more clued up on internet parody/hoax sites than the average person? It is very likely that the interviewed 'activists' had no prior knowledge of the dhmo paraody, and, in the fervour of a protest rally, unthinkingly said 'yes' to an interviewer asking if they wanted to sign a petition against a 'Very Dangerous Chemical'?
Can you honestly say that your response would have been any different?
Oops, yeah you are right. Disappointing no one else noticed. Still, treating 2^31 as negative is ultimately an arbitary choice, albeit a consistent one.
Sure, they just say "FET device". Possibly just a single FET. Early days yet, it is a long way to get from a single transistor to a microprocessor. The more immediate applications for this are high power high frequency devices which currently still use vacuum tubes.
I dunno, if it is Sun, then it is a desperate last-ditch bid to kill Linux, and maybe kill themselves in the process. If they are financing SCO (which is what this licencing amounts to: Sun more than most companies must know the worthlessness of SCO's claims), then their reputation is going to go down pretty fast once it becomes public.
Nice, but maybe the grammar is a bit too twisted? Here is my take:
Unstable, this software is. Coveted by the Dark Side* it is. Prepare for bugs or rampant filesystem corruption you must. Be sure young Jedi; a bug, to The Council, you must report. Ignoring bugs leads to apathy. Apathy leads to subservience. Subservience leads to EULA's. EULA's lead to slavery. Slavery leads to Microsoft.
*[SCO]
Re:Does Anyone Remember Cold Fusion?
on
More on Spintronics
·
· Score: 2, Interesting
It is not a transfer of energy, but a transfer of spin. Its essentially a transfer of information.
In general, electons exist in a superposition of two states, "up" and "down", with oppositely directed "spin" (which obeys almost the same mathematical formalism as angular momentum, with some interesting twists). For a free system of electrons, the up and down states occur with equal measure so that the resulting wavefunction is spinless (sum of spin over all electrons is zero) which implies it is rotationally invariant (ie, rotate all electrons by some angle and the system is indistinguishable from the original).
In the presence of interactions (ie. an E/M field), the alignment of the field specifies an axis, and the coupling to the spin means that up and down spins (with respect to this axis) have different energies. Thus there is no longer rotational symmetry and you can control the direction of spin.
This has nothing to do with the motion of the electrons themselves, but only the spin. One way to think about this is to think of each quantum number as corresponding to a different particle. An electron has spin 1/2 and charge 1 so you can consider an electron to be a bound state of a pure spin particle and a pure charge particle. In fact, in one-dimensional systems (and possibly sometimes in 2D) this is not even a mathematical trick, and it is possible to prove that the spin and charge components are no longer bound to each other! The system behaves as if it was composed of two separate species of particles, "spinons" (carrying the spin) and "holons" (carrying the charge). This is called "spin-charge separation", and it is a collective effect, it doesn't work in a few-body system.
This is not to say that Zhang's result has anything to do with spin-charge separation per se, just an example.
No, you are right, the parent comment is talking nonsense.
RFID tags are susceptible to everything that other electronics is - extreme high/low temperature, EM pulses, large currents ('large' in this context may in fact be rather small),cosmic rays (only an issue for very sensitive devices) etc etc.
Of course, you can try to build the circuit in such a way that it behaves well when subject to interference. Bypass capacitors and so on. Whether it is worth the expense (or even desirable) to do this for RFID tags is another story.
I think the danger is not su much that government(s) themselves will set this sort of thing up, but shops (especially large mega-chains) are undoubtedly very interested.
For example, put an RFID scanner at the door of the shop, supposedly to make sure the people walking out with a pair of jeans under their arm has actually paid for them. But an added bonus: you also know the instant someone walks into the store wearing a competitor's brand of jeans. This could easily be linked to in-store advertising (eg, anyone walking into the store wearing OTHER-brand jeans instantly gets bombarded by videos and promos for NAME-brand jeans). No civil rights abuse? Probably not, just more intrusive adversising, nothing that could not be done by a sufficiently pushy salesperson anyway.
But the problem is that if the store puts this stuff in a database, then the government has access to it (in the name of 'fighting terrorism' - if RFID tags are given the go-ahead based on the antiterrorism card, then this will probably be mandated). There is no reason for them to have this data, and it is hard to dream up a way that the current government (or even short/medium term governments) could abuse it.
But the point is that it allows very large scale and very scary abuse. For example, suppose you are at a G8 protest and the riot police charge into the crowd. You flee, only to get caught some minutes later by a cop with access to the real-time RFID database (yes, you were already listed in the database as a potential terrorist).
Unrealistic, but (1) technically possible, even in the near future, and (2) not illegal.
Once this type of surveilence is in place, copuled with strong powers to crack down on dissidents (which Ashcroft already has), there is very little protection to stop the state sliding into a 1984-type system. Yes, this is the worst-case scenario but to deny that it could happen is to ignore history.
And yes, the range of RFID tags is always going to be measured in inches or feet, rather than miles. Big technical limitation, but you can partially work around it by having more receivers.
Interesting, I did not know that they had done that. But, I can only find statements that they assigned S/390 port to FSF, I can't see anything about other stuff (NUMA, RCU etc).
I reckon IBM would be quite thankful for their own copyrights appearing in Linux just now. It gives them more grounds to countersue SCO;-)
So, any bets on when the FSF get involved legally? My guess is they won't unless they absolutely have to, I think most people trust IBM to set things straight for everyone. Surprising, given IBM's history!
LOL, I like the way they use the term "fear,
uncertainty and doubt" (BACKGROUND section E, paragraph 22), since it was IBM that pioneered the technique in the beginning (soon to be overtaken by the all-time master of the technique, Microsoft).
The term "Intellectual property" exists as a piece of language, but then so does the term "inedible food", or "military intelligence".
However, the concept of "intellectual properly", as an umbrella term for copyright, trademark, patents, and trade secrets, might be useful in some contexts, but as you admit it often leads to confused thinking, and isn't a good term to use purely on those grounds.
The concept of "intellectual properly", as an entity in its own right (ie, by taking the traditional meaning of "property" and applying it to something "intellectual" as opposed to "material"), I simply do not recognize as having any legitimacy. This is, ultimately, the biggest abuse of the term IMO.
What does the FSF have to do with this? The FSF holds no copyrights in the Linux kernel, so they have no right to sue on those grounds.
They could sue over GPL violations, but IBM already has that in hand, the FSF would only be interested if it was their code, or maybe as a community service, if no one else had any money to pay the lawyers.
Besides, the picture on exactly what GPL violations SCO has committed is not very clear at the moment. At least wait until there is some evidence!
Why would the GPL revert to a BSD-style licence? If the GPL was found to be improperly applied, then it reverts to COPYRIGHT law, exactly the same as if the MS EULA or the BSD licence was found to be invalid in some instance.
This means its just like a book, you have no rights to copy or make derivative works at all. None. Not unless all of the copyright holders agree to re-licence it in some way. In practice, if SCO manage to prove that some code was misappropriated into the Kernel, they are the true copyright holder of the offending code, and they choose not to licence the code under the GPL (hard to imagine, since they are already distributing said code under the GPL), then the simple answer is to remove all of the code that SCO prove copyright to. Instantly, the kernel becomes distributable again.
Of course, the kernel might not compile anymore, but fixing it is surely not too hard. The replacement code only needs to be different enough to avoid copyright problems, not patents or anything like that. A similar situation would be rewriting a chapter of a text-book.
But that statement is only there for clarification, even if it wasn't present then there would still be no restrictions on use. It also explicitly states that you don't have to agree to the licence to use it. That is true, and if you don't agree then your rights are exactly what you would have if it were, eg, a book. Free use for any purpose, but no distribution or derivative works.
That is in accord with copyright law. Licence agreements that put restrictions on use are completely different, and rely instead on contract law. Almost all such licences are on rather shakey ground in fact. I don't think the MS EULA has ever been tested in court, and the majority of clauses would surely be found unenforcable (well, I can't speak for the USA, but I'm fairly confident about Australia and Europe).
The GPL is in a different position, because it doesn't take away any rights at all. In the absence of GPL (ie, pure copyright law) you would have no rights to distribute, or create derivative works. The GPL allows this, but with restrictions on how you do it (ie, derivative works must also be GPL, no additional restictions beyond the GPL, etc).
You do. In Australia at least, scrutineers are allowed at every step of the process. While I suspect the counters would be a bit pissed if you stopped them in the middle of the count to give them a maths test, there is nothing stopping the scrutineers from repeating the calculation and comparing answers.
Mind you, I don't know how the scrutineers are selected. At minimum, each party can provide scrutineers, but I'm not sure whether the general public can. The system works well enough in Australia that there has never been any suggestions of systematic problems. (Well, Queensland might be an exception!)
Slashdot Mirror
Perhaps there is already a freely-available atoi(3) implementation that can be used by the Kernel? Yes, in fact there is! Even better, it coincides with the atoi(3) function that was removed! Woohoo! We are saved! There is not even eny need to look further afield, like the multitude of *other* free implementations around the place (say, glibc? BSD?), or (at last resort) actually finding a master C programmer to lock himself in a room for the long winter months in an effort to come up with an independent implementation.
The funniest thing is, that it probably WOULD take a huge effort by a master programmer to actually find an implementation of atoi(3) that was truely independent
Linux has run natively on 64 bit architectures (Alpha, SPARC, MIPS etc etc) since almost the beginning. All of the GNU tools are 64-bit clean and always have been, AFAIK. I know for sure Gnome compiles without change for 64-bit arch, I would assume KDE does as well.
Each card has a unique ID number linked to the registered owner's name, which is recorded together with the location and time of the exchange every time the card is used.
They are stuck between a rock and a hard place now. It sounds like there is a viable business model out there of "patent something really obvious then sue Microsoft for patent violation". As the biggest fish out there they are clearly the most attractive target, given a sufficently 'strong' case. On the other hand, Microsoft's patent portfolio is its ultimate trump card and which they are presumably saving for the final defence, in the event that open source starts to seriously affect their viability. They can't suddenly start lobbying to remove patent protection without invalidating their own portfolio too.
They brought it upon themselves.
By default, duty is payable on anything worth more than $x (can't remember the exact figure). But it isn't payable on personal effects, so you could easily have claimed the duty back.
I wouldn't be so sure of that, 10k is actually a significant fraction of their last years revenue....
Perhaps it is a cultural difference (I am not from the US), but if I was at a protest rally, and someone was collecting signatures for a petition on a plausible-sounding environmental issue, I would assume that they were genuine. I don't mean that I would blindly sign my name on anything, but given a conversation like "Hi! Do you want to sign this petition to ban dihydrogen monoxide?" "What is dihydrogen monoxide?" "It is a dangerous chemical that kills many people every year, and bad companies like Monsanto are dumping it in rivers!" "OK! I'll sign!" I would assume that they were not deliberately setting out to mislead me.
I don't think it is valid to compare an activist with a used car salesman. The activist is doing what they think is right for some (presumably noble) cause. The used car salesman is doing what he thinks will give him a quick dollar.
Of course, that doesn't mean that any of the parties actually have a clue, but the intent is very different at least.
Basically, they have been told to "put up or shut up", and they did neither, hence the fine.
It does not mean that a court has found SCO is not telling the truth, it just means that if SCO want to continue making big statement, they need to accompanying evidence.
Pick a person at random and ask them whether they support a ban on a very dangerous chemical substance named 'dihydrogen monoxide'. Anyone who isn't a chemistry graduate will have to think for at least a second or two to figure out what this is, and people with no chemistry background would likely not figure it out at all (not without more information, at least).
Now, practically all of what you consider 'activists' are ordinary people, who happen to believe that Greenpeace may, on some level, have a point. Why should these people be any more clued up on internet parody/hoax sites than the average person? It is very likely that the interviewed 'activists' had no prior knowledge of the dhmo paraody, and, in the fervour of a protest rally, unthinkingly said 'yes' to an interviewer asking if they wanted to sign a petition against a 'Very Dangerous Chemical'?
Can you honestly say that your response would have been any different?
Oops, yeah you are right. Disappointing no one else noticed. Still, treating 2^31 as negative is ultimately an arbitary choice, albeit a consistent one.
Which presumably happens when the bug count wraps around from 2^31 to -2^31+1 then up to zero...
Maybe this is the basis for Microsoft release schedules?
Sure, they just say "FET device". Possibly just a single FET. Early days yet, it is a long way to get from a single transistor to a microprocessor. The more immediate applications for this are high power high frequency devices which currently still use vacuum tubes.
I dunno, if it is Sun, then it is a desperate last-ditch bid to kill Linux, and maybe kill themselves in the process. If they are financing SCO (which is what this licencing amounts to: Sun more than most companies must know the worthlessness of SCO's claims), then their reputation is going to go down pretty fast once it becomes public.
Unstable, this software is. Coveted by the Dark Side* it is. Prepare for bugs or rampant filesystem corruption you must. Be sure young Jedi; a bug, to The Council, you must report. Ignoring bugs leads to apathy. Apathy leads to subservience. Subservience leads to EULA's. EULA's lead to slavery. Slavery leads to Microsoft.
*[SCO]
In general, electons exist in a superposition of two states, "up" and "down", with oppositely directed "spin" (which obeys almost the same mathematical formalism as angular momentum, with some interesting twists). For a free system of electrons, the up and down states occur with equal measure so that the resulting wavefunction is spinless (sum of spin over all electrons is zero) which implies it is rotationally invariant (ie, rotate all electrons by some angle and the system is indistinguishable from the original).
In the presence of interactions (ie. an E/M field), the alignment of the field specifies an axis, and the coupling to the spin means that up and down spins (with respect to this axis) have different energies. Thus there is no longer rotational symmetry and you can control the direction of spin.
This has nothing to do with the motion of the electrons themselves, but only the spin. One way to think about this is to think of each quantum number as corresponding to a different particle. An electron has spin 1/2 and charge 1 so you can consider an electron to be a bound state of a pure spin particle and a pure charge particle. In fact, in one-dimensional systems (and possibly sometimes in 2D) this is not even a mathematical trick, and it is possible to prove that the spin and charge components are no longer bound to each other! The system behaves as if it was composed of two separate species of particles, "spinons" (carrying the spin) and "holons" (carrying the charge). This is called "spin-charge separation", and it is a collective effect, it doesn't work in a few-body system.
This is not to say that Zhang's result has anything to do with spin-charge separation per se, just an example.
RFID tags are susceptible to everything that other electronics is - extreme high/low temperature, EM pulses, large currents ('large' in this context may in fact be rather small),cosmic rays (only an issue for very sensitive devices) etc etc.
Of course, you can try to build the circuit in such a way that it behaves well when subject to interference. Bypass capacitors and so on. Whether it is worth the expense (or even desirable) to do this for RFID tags is another story.
For example, put an RFID scanner at the door of the shop, supposedly to make sure the people walking out with a pair of jeans under their arm has actually paid for them. But an added bonus: you also know the instant someone walks into the store wearing a competitor's brand of jeans. This could easily be linked to in-store advertising (eg, anyone walking into the store wearing OTHER-brand jeans instantly gets bombarded by videos and promos for NAME-brand jeans). No civil rights abuse? Probably not, just more intrusive adversising, nothing that could not be done by a sufficiently pushy salesperson anyway.
But the problem is that if the store puts this stuff in a database, then the government has access to it (in the name of 'fighting terrorism' - if RFID tags are given the go-ahead based on the antiterrorism card, then this will probably be mandated). There is no reason for them to have this data, and it is hard to dream up a way that the current government (or even short/medium term governments) could abuse it.
But the point is that it allows very large scale and very scary abuse. For example, suppose you are at a G8 protest and the riot police charge into the crowd. You flee, only to get caught some minutes later by a cop with access to the real-time RFID database (yes, you were already listed in the database as a potential terrorist).
Unrealistic, but (1) technically possible, even in the near future, and (2) not illegal.
Once this type of surveilence is in place, copuled with strong powers to crack down on dissidents (which Ashcroft already has), there is very little protection to stop the state sliding into a 1984-type system. Yes, this is the worst-case scenario but to deny that it could happen is to ignore history.
And yes, the range of RFID tags is always going to be measured in inches or feet, rather than miles. Big technical limitation, but you can partially work around it by having more receivers.
I reckon IBM would be quite thankful for their own copyrights appearing in Linux just now. It gives them more grounds to countersue SCO ;-)
So, any bets on when the FSF get involved legally? My guess is they won't unless they absolutely have to, I think most people trust IBM to set things straight for everyone. Surprising, given IBM's history!
I think their lawyers have a sense of humor.
However, the concept of "intellectual properly", as an umbrella term for copyright, trademark, patents, and trade secrets, might be useful in some contexts, but as you admit it often leads to confused thinking, and isn't a good term to use purely on those grounds.
The concept of "intellectual properly", as an entity in its own right (ie, by taking the traditional meaning of "property" and applying it to something "intellectual" as opposed to "material"), I simply do not recognize as having any legitimacy. This is, ultimately, the biggest abuse of the term IMO.
They could sue over GPL violations, but IBM already has that in hand, the FSF would only be interested if it was their code, or maybe as a community service, if no one else had any money to pay the lawyers.
Besides, the picture on exactly what GPL violations SCO has committed is not very clear at the moment. At least wait until there is some evidence!
This means its just like a book, you have no rights to copy or make derivative works at all. None. Not unless all of the copyright holders agree to re-licence it in some way. In practice, if SCO manage to prove that some code was misappropriated into the Kernel, they are the true copyright holder of the offending code, and they choose not to licence the code under the GPL (hard to imagine, since they are already distributing said code under the GPL), then the simple answer is to remove all of the code that SCO prove copyright to. Instantly, the kernel becomes distributable again.
Of course, the kernel might not compile anymore, but fixing it is surely not too hard. The replacement code only needs to be different enough to avoid copyright problems, not patents or anything like that. A similar situation would be rewriting a chapter of a text-book.
That is in accord with copyright law. Licence agreements that put restrictions on use are completely different, and rely instead on contract law. Almost all such licences are on rather shakey ground in fact. I don't think the MS EULA has ever been tested in court, and the majority of clauses would surely be found unenforcable (well, I can't speak for the USA, but I'm fairly confident about Australia and Europe).
The GPL is in a different position, because it doesn't take away any rights at all. In the absence of GPL (ie, pure copyright law) you would have no rights to distribute, or create derivative works. The GPL allows this, but with restrictions on how you do it (ie, derivative works must also be GPL, no additional restictions beyond the GPL, etc).
Seaquest, DSV was a TV show, wasn't it?
Mind you, I don't know how the scrutineers are selected. At minimum, each party can provide scrutineers, but I'm not sure whether the general public can. The system works well enough in Australia that there has never been any suggestions of systematic problems. (Well, Queensland might be an exception!)