Besides Tivoisation and patents, there are some other good things in the GPL v3, as it is currently drafted. It will be no longer necessary to provide source code via snail mail if you distribute binaries without source. This is the 21st century, providing access to copy the Corresponding Source from a network server at no charge will be sufficient.
The GPL v2 allows that already. In section 3a, it states that to distribute you must accompany the work "with the complete corresponding machine-readable source code, which must be distributed under the terms of Sections 1 and 2 above on a medium customarily used for software interchange." (emphasis mine) Internet distribution qualifies as such a medium, so the GPLv2 in this case doesn't require much more than putting up a link to the code on a web server or ftp server and saying "The source is available on "
If that doesn't cut it, section 3b says that, you may include a written offer instead, to provide the source "for a charge no more than your cost of physically performing source distribution" without specifying what that distribution might be. Offering a written link to your website, or ftp server, or SVN repo, or whatever is also perfectly acceptable in this case.
It then continues: "If distribution of executable or object code is made by offering access to copy from a designated place, then offering equivalent access to copy the source code from the same place counts as distribution of the source code, even though third parties are not compelled to copy the source along with the object code." (emphasis, again, mine)
This is the final nail on the coffin. The nonsense that the GPL doesn't allow for providing access to the code from a server is just that. Nonsense.
This "expert" does not appear to be very well versed with Windows, and that should be pointed out, with a bright green laser pointer. He says that there is no indication of what DHCP address the computer had, but that is not entirely accurate.
Both 9x and NT-based variants keep information about DHCP address assignments in the registry, so that they can attempt to request their previous IP address after a startup. Specifically, in NT-based systems, you can look under "HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Servi ces\Tcpip\Parameters\Interfaces" to see a list of interfaces that Windows has available, and under each one of those, there exists a REG_SZ value, aptly named "DhcpIPAddress", which includes, in plain ASCII text in dotted-quad notation the last DHCP address handed to the box by the DHCP server at the IP specified by the "DhcpServer" REG_SZ value. Older entries could potentially exist under the "ControlSet001" and "ControlSet002" keys, both of which are backups.
While this method is by no means bulletproof, it could potentially disclose the last IP address the computer obtained from a particular DHCP server and that would not only be useful, but perhaps even relevant information.
This is a company whose founders contribute very little back of their wealth to charitable causes and instead choose to spend it on 747's with waterbeds and other such items.
You're a troll. But what the hell. I haven't posted in a while.
Let's say it's true that they contribute very little back. I don't see the problem; charity is, by definition, not an obligation and they should be able to enjoy the wealth they created as they see fit within the confines of the law. If they want to buy a 747 or use $50 bills as kindling for the next "Google Company BBQ' they should be able to do just that.
Bottom line... Microsoft is doing "on the fly encryption". That is what signing of kernel modules is all about.
You would be well advised to not speak about things you don't know anything about, because you will, undoubtedly, end up looking like a fool. Encryption and signing are two very, very different things, serving two very different purposes, and to call signing of kernel modules "on the fly encryption" is, well, retarted.
Also, note that the original poster was referring to encryption of user data because without that, even if the copy of Vista is deactivated, nothing stops a user from plugging the drive via USB to another machine -- whether Vista, XP or even Linux -- and pulling files out.
I'm well aware of the situation outside the United States, seeing how I was raised in a European country where the cost of gasoline is currently sky high. However, I would point out that additional reasons are probably at play there, such as taxes on horsepower, or limits tied to income that restrict the maximum displacement of a vehicle one may buy.
Prohibition was an attempt to get people to stop drinking. I'm talking about putting tax on electricity usage. I don't see the connection, except that in this case, people making their own electricity would be a good thing, and would be encouraged, and not taxed.
And your tax is an attempt to get people to be more responsible with energy usage. It will fail, just as gasoline taxes fail to force people to buy fuel efficient vehicles.
Straw man. I'm not suggesting taxing everything - just making it expensive to take part in activities that are unsustainable and damage the environment.
You're suggesting using taxes to alter the energy consumption habits of people. It won't work. It never has.
If it took 1W/h to keep the device on standby before, it will still take 1W/h to keep it on standby. You'll charge the energy storage unit (battery, capacitor, whatever) while the device is running, and when off, the power will come from the stored energy. Ignoring the losses for storage in a battery (e.g.: heat produced during the charging of batteries) how would it increase the consumption? Specifics please.
Better yet, simply educate consumers. They will prefer energy-efficient products, which will result in more energy-efficient products being introduced. Hooray for the free market!
Yeah, like the prohibition forced people to use alcohol more wisely and taxes on gasoline forced people to use bicycles. This may strike you as odd, but the solution to everything is not legislation, and it's not taxation.
I would imagine that the engineers that work for Seagate are neither incompetent nor fools, and I am fairly confident they have selected an encryption algorithm that is resistant to both different and linear cryptanalysis.
Of course, encryption is hard to get right, so it's possible -- even probable -- that attacks against the DriveTrust encryption will be discovered, but I doubt it will be because of the guts of the algorithm. The implementation is far more likely to be the weak link.
The "other parts of the world" that you specify don't include Greece. There, free speech and the freedom of the press is guaranteed by the Constitution (Section II, Article 14) with exceptions similar to those in force of the U.S.A. by Court fiat (such as yelling "fire" in a crowded theater or publicizing details detrimental to national security) and two additional restrictions against insulting the person of the President of the Republic, and the Christian faith, which is the de jure national religion.
Triangulating a mobile phone to within a couple of hundred meters (frequently less than that) does not require police on foot, with three antennas. The cellular system, in order to work and not for any sinister big-brother type reason, has to track the rough relative position of a subscriber within a cell, to account for signal delay propagation.
Just to elaborate a bit about timeslots: The GSM standards require that the phone transmit only in a defined time frame: three time slots after the phone has received the data. This gives the tower a well-defined interval during which to receive the data transmitted and to ensure that transmissions by different phones are separated by at least one guard period.
But as the distance between the tower and the cell increases, the cell phone must transmit earlier and earlier to account for the increase signal propagation delay. This process is called "adaptive frame alignment" and is determined by a parameter known as the "timing advance" parameter. The TA is dynamically updated and takes values between 0 and 63 inclusive. This parameter determines how early, in microseconds, the phone has to begin transmission, to ensure that the signal reaches the tower at the correct time, and roughly locates the phone within a specific radius around the tower.
Combining this with the fact that more than one tower usually sees the phone and the information from those other towers, the GSM system can triangulate the signal to within a couple of hundred meters, easily.
Linus is a pragmatist. He didn't write Linux for academic purpose. He wanted it to work.
I guess the 'fools' who wrote QNX were not pragmatists. They were cloistered ivory tower intellectuals, who intended QNX as an academic project. That's why they didn't start a company, and their O/S isn't widely used in a wide variety of sectors -- especially telematics.
First of all, thanks for not flipping your lid, and replying with "teh GPL!!!1!!!!!!" These issues have a way of devolving into flamewars with little actual content.
On the issue, I think this will remain very much an 'open' question until a case is brought to Court, and a decision establishing some precedent is reached.
Define "distributed." If I distribute the source code to the Linux kernel, and only a loadable module (no source) does that make my module fall under the GPL? If so, why?
Note, that there are two cases to consider. The first and trivial one, is someone wanting his module to be included in a vanilla kernel distribution. In that case, the module MUST be released under an approved license (GPL, BSD, etc) to be considered for inclusion, but that is not what we are talking about.
The second case -- whether all modules become GPL simply by virtue of linking to the kernel, which is GPL is quite another issue, and the one that we _are_ talking about. First of all, you must realize that the GPL cannot set its own scope (it cannot, for example say, that all programs stored on a computer on which a single GPL program exists now become GPL-licensed) and so nothing in the GPL can address attachment by linking.
To argue that merely linking a module to the kernel brings the module under the GPL is patently absurd. And the trivial proof is this: if merely linking created a derivative work, the derivative work itself would be eligible for copyright, and the person which linked, could claim copyright of this new 'kernel with foomatic2000' work. All that linking produces is an aggregate -- a 'combination' of two original works.
Eblen Moglen is not the one that claimed that a a piece of code A is a "derivative work" of another work B, which is written after the original code A. You did that.
So, how about you answer the question I posed, which you quite skillfully evaded: If I write a replacement for libc, and use LD_PRELOAD to have your application link against my library and not libc, does that make your application a derivative of my library or not?
You're the one that seems to have a fundamental misunderstanding of not just the legal, but the ethical principles involved in this, and you make it clear when you spout non-sense such as this:
"doing so makes them a derivitive (yes, even those drivers that predate the linux kernel)"
That has got to be the most ridiculous statement that I have ever heard. Consider this: you write an application which uses libc. I write a replacement for libc, and use LD_PRELOAD to have your application use my library instead of libc; does that, now, make your app a derivate work of my library?
I wasn't arguing against any one specific Intelligent Design application. I was proving that Intelligent Design is either self-contradictory or out of the realm of science.
And science doesn't hit an "infinite regress"? What caused the Big Bang? And what was the cause of that? And what caused that before it? Ad Nauseum... Science's dirty little secret is that they have no idea where anything came from. Anyone who tells you differently is selling something.
Whether it does or not hit an infinite regress is open to debate. But science is not afraid to say "I don't know" and to go looking for answers.
Your logic is flawed. ID says that the *physical* world could not have naturally developed. God is outside the physical world. We can only know God by what He has chosen to reveal to us, making it very difficult for us, constrained to the physical world, to put any soft of scientific test on God.
My argument makes no mention of a deity. Whether your particular flavor of deity exists and its nature is an entirely different matter -- one that I will be happy to debate with you in a more appropriate medium.
My logic is not flawed. My argument specifically addressed whether Intelligent Design is science: It is NOT. The reason for that is because ID either posits a supernatural designer and thus is outside the realm of science, or requires a natural designer that, by the premise on which ID is founded, does not require design, making the "theory" self-contradictory.
To me, trying to figure out how life originated it outside the realm of science.
Let's take a deeper look there vargasgrey, shall we?
Intelligent Design claims that complex natural life forms can only be created by something it terms a designing intelligence. OK... so, let's contemplate that for a bit.
If we allow the creating intelligence to be natural, by our original premise, it too must have a creating intelligence that created it, and so on. We're left with an infinite regress. So, how to go about breaking it?
Well, maybe we could posit a supernatural creating intelligence. But, if we take that option we instantly take Intelligent Design outside the realm of science, and thus automatically forfeit equal status to scientific theories. So, that's no good.
The other option, is to accept that intelligence can arise solely out of natural processes, which clearly contradicts the original premise of Intelligent Design, so that's out the door too.
Dang it. No matter what we do, Intelligent Design ends up being self-contradictory, or non-scientific.
Slashdot Mirror
The GPL v2 allows that already. In section 3a, it states that to distribute you must accompany the work "with the complete corresponding machine-readable source code, which must be distributed under the terms of Sections 1 and 2 above on a medium customarily used for software interchange." (emphasis mine) Internet distribution qualifies as such a medium, so the GPLv2 in this case doesn't require much more than putting up a link to the code on a web server or ftp server and saying "The source is available on "
If that doesn't cut it, section 3b says that, you may include a written offer instead, to provide the source "for a charge no more than your cost of physically performing source distribution" without specifying what that distribution might be. Offering a written link to your website, or ftp server, or SVN repo, or whatever is also perfectly acceptable in this case.
It then continues: "If distribution of executable or object code is made by offering access to copy from a designated place, then offering equivalent access to copy the source code from the same place counts as distribution of the source code, even though third parties are not compelled to copy the source along with the object code." (emphasis, again, mine)
This is the final nail on the coffin. The nonsense that the GPL doesn't allow for providing access to the code from a server is just that. Nonsense.
Both 9x and NT-based variants keep information about DHCP address assignments in the registry, so that they can attempt to request their previous IP address after a startup. Specifically, in NT-based systems, you can look under "HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Servi ces\Tcpip\Parameters\Interfaces" to see a list of interfaces that Windows has available, and under each one of those, there exists a REG_SZ value, aptly named "DhcpIPAddress", which includes, in plain ASCII text in dotted-quad notation the last DHCP address handed to the box by the DHCP server at the IP specified by the "DhcpServer" REG_SZ value. Older entries could potentially exist under the "ControlSet001" and "ControlSet002" keys, both of which are backups.
While this method is by no means bulletproof, it could potentially disclose the last IP address the computer obtained from a particular DHCP server and that would not only be useful, but perhaps even relevant information.
I wasn't aware that's the case in the United States. Hmm...
You're a troll. But what the hell. I haven't posted in a while.
Let's say it's true that they contribute very little back. I don't see the problem; charity is, by definition, not an obligation and they should be able to enjoy the wealth they created as they see fit within the confines of the law. If they want to buy a 747 or use $50 bills as kindling for the next "Google Company BBQ' they should be able to do just that.
You would be well advised to not speak about things you don't know anything about, because you will, undoubtedly, end up looking like a fool. Encryption and signing are two very, very different things, serving two very different purposes, and to call signing of kernel modules "on the fly encryption" is, well, retarted.
Also, note that the original poster was referring to encryption of user data because without that, even if the copy of Vista is deactivated, nothing stops a user from plugging the drive via USB to another machine -- whether Vista, XP or even Linux -- and pulling files out.
I'm well aware of the situation outside the United States, seeing how I was raised in a European country where the cost of gasoline is currently sky high. However, I would point out that additional reasons are probably at play there, such as taxes on horsepower, or limits tied to income that restrict the maximum displacement of a vehicle one may buy.
I know - which is why I said "ignoring those losses." The OP suggested that, excluding those losses, an energy storage system would be less efficient.
If it took 1W/h to keep the device on standby before, it will still take 1W/h to keep it on standby. You'll charge the energy storage unit (battery, capacitor, whatever) while the device is running, and when off, the power will come from the stored energy. Ignoring the losses for storage in a battery (e.g.: heat produced during the charging of batteries) how would it increase the consumption? Specifics please.
Better yet, simply educate consumers. They will prefer energy-efficient products, which will result in more energy-efficient products being introduced. Hooray for the free market!
Of course, encryption is hard to get right, so it's possible -- even probable -- that attacks against the DriveTrust encryption will be discovered, but I doubt it will be because of the guts of the algorithm. The implementation is far more likely to be the weak link.
The "other parts of the world" that you specify don't include Greece. There, free speech and the freedom of the press is guaranteed by the Constitution (Section II, Article 14) with exceptions similar to those in force of the U.S.A. by Court fiat (such as yelling "fire" in a crowded theater or publicizing details detrimental to national security) and two additional restrictions against insulting the person of the President of the Republic, and the Christian faith, which is the de jure national religion.
Triangulating a mobile phone to within a couple of hundred meters (frequently less than that) does not require police on foot, with three antennas. The cellular system, in order to work and not for any sinister big-brother type reason, has to track the rough relative position of a subscriber within a cell, to account for signal delay propagation.
Just to elaborate a bit about timeslots: The GSM standards require that the phone transmit only in a defined time frame: three time slots after the phone has received the data. This gives the tower a well-defined interval during which to receive the data transmitted and to ensure that transmissions by different phones are separated by at least one guard period.
But as the distance between the tower and the cell increases, the cell phone must transmit earlier and earlier to account for the increase signal propagation delay. This process is called "adaptive frame alignment" and is determined by a parameter known as the "timing advance" parameter. The TA is dynamically updated and takes values between 0 and 63 inclusive. This parameter determines how early, in microseconds, the phone has to begin transmission, to ensure that the signal reaches the tower at the correct time, and roughly locates the phone within a specific radius around the tower.
Combining this with the fact that more than one tower usually sees the phone and the information from those other towers, the GSM system can triangulate the signal to within a couple of hundred meters, easily.
First of all, thanks for not flipping your lid, and replying with "teh GPL!!!1!!!!!!" These issues have a way of devolving into flamewars with little actual content.
On the issue, I think this will remain very much an 'open' question until a case is brought to Court, and a decision establishing some precedent is reached.
Have a nice day.
Define "distributed." If I distribute the source code to the Linux kernel, and only a loadable module (no source) does that make my module fall under the GPL? If so, why?
Even then -- why would the kernel module have to be GPL? On what grounds?
Note, that there are two cases to consider. The first and trivial one, is someone wanting his module to be included in a vanilla kernel distribution. In that case, the module MUST be released under an approved license (GPL, BSD, etc) to be considered for inclusion, but that is not what we are talking about.
The second case -- whether all modules become GPL simply by virtue of linking to the kernel, which is GPL is quite another issue, and the one that we _are_ talking about. First of all, you must realize that the GPL cannot set its own scope (it cannot, for example say, that all programs stored on a computer on which a single GPL program exists now become GPL-licensed) and so nothing in the GPL can address attachment by linking.
To argue that merely linking a module to the kernel brings the module under the GPL is patently absurd. And the trivial proof is this: if merely linking created a derivative work, the derivative work itself would be eligible for copyright, and the person which linked, could claim copyright of this new 'kernel with foomatic2000' work. All that linking produces is an aggregate -- a 'combination' of two original works.
So, how about you answer the question I posed, which you quite skillfully evaded: If I write a replacement for libc, and use LD_PRELOAD to have your application link against my library and not libc, does that make your application a derivative of my library or not?
You're the one that seems to have a fundamental misunderstanding of not just the legal, but the ethical principles involved in this, and you make it clear when you spout non-sense such as this:
"doing so makes them a derivitive (yes, even those drivers that predate the linux kernel)"
That has got to be the most ridiculous statement that I have ever heard. Consider this: you write an application which uses libc. I write a replacement for libc, and use LD_PRELOAD to have your application use my library instead of libc; does that, now, make your app a derivate work of my library?
Please learn to read, before you post.
And science doesn't hit an "infinite regress"? What caused the Big Bang? And what was the cause of that? And what caused that before it? Ad Nauseum... Science's dirty little secret is that they have no idea where anything came from. Anyone who tells you differently is selling something. Whether it does or not hit an infinite regress is open to debate. But science is not afraid to say "I don't know" and to go looking for answers.
My argument makes no mention of a deity. Whether your particular flavor of deity exists and its nature is an entirely different matter -- one that I will be happy to debate with you in a more appropriate medium.
My logic is not flawed. My argument specifically addressed whether Intelligent Design is science: It is NOT. The reason for that is because ID either posits a supernatural designer and thus is outside the realm of science, or requires a natural designer that, by the premise on which ID is founded, does not require design, making the "theory" self-contradictory.
To me, trying to figure out how life originated it outside the realm of science.
You are, of course, entitled to your opinion.
Intelligent Design claims that complex natural life forms can only be created by something it terms a designing intelligence. OK... so, let's contemplate that for a bit.
If we allow the creating intelligence to be natural, by our original premise, it too must have a creating intelligence that created it, and so on. We're left with an infinite regress. So, how to go about breaking it?
Well, maybe we could posit a supernatural creating intelligence. But, if we take that option we instantly take Intelligent Design outside the realm of science, and thus automatically forfeit equal status to scientific theories. So, that's no good.
The other option, is to accept that intelligence can arise solely out of natural processes, which clearly contradicts the original premise of Intelligent Design, so that's out the door too.
Dang it. No matter what we do, Intelligent Design ends up being self-contradictory, or non-scientific.
So chew on that.