The choice to "license" rather than "sell" is up to the software seller.
Some give you a EULA you must click on and then claim you have agreed to a license. There are plenty of arguments that this is invalid, and in fact this is a sale and not a license, anyway.
But it is absolutely certain that any company that wants to "sell" Linux is free to do so. They decide if they want to "license" it, not the government, and are free to decide to "sell".
All this means that the distinction is meaningless.
PS: people who lease cars have signed a great deal of licensing agreements, but they aren't any more liable than a person who bought a car for an argument between GM and Ford.
If GPL'd code was found in a product like Windows, would Microsoft be forced to open source the entire thing?
No, they would only have to remove the GPL code from future versions of their product.
There may be monetary damages paid to the original copyright holder by Microsoft. But copyright law does not allow much else (despite SCO's totally bogus claims).
Though I don't think the government should require GPL, I want to say that I think Microsoft could do quite well even if they GPL'd their software. They can still contract with OEM's to install Windows and customize it and do a contract that is based on the number of units sold, like they do now. Any competitor selling "bob's customized Windows" would have a hard to impossible time competing with that. Also, more seriously, Microsoft could include some closed-source stuff like WMP or something that you are not allowed to run on non-Microsoft Windows.
Personally I feel a government mandate for open *data* standards would be best. If Microsoft writes a clever word processing algorithm, they should be able to profit from that and governments should pay Microsoft for the right to use it. However they currently don't do that, they can write crap because they rely on the lock-in of their closed formats.
Both sides in the argument (both @stake and Geer) agree that the paper was Geer's own writing and was not a "study" with any kind of funding.
Making up your own facts makes it hard for me to believe anything you say.
To be honest, I think Geer has figured out a way to get a lot of publicity, and may have already known he was going to get fired for some other reason and decided to cause a lot of trouble. It would make absolutely no sense for either Microsoft or @stake to dirty their reputations by doing this quickly. Even if you assumme 100% evil it would make more sense for them to wait a month or so or even fabricate a more definate reason to fire him.
Why don't you read the article again. It was @stake that claimed they fired him on Tuesday. Nobody (him included) seemed to know anything about it until Thursday (or at least so he claims).
That sounds very strange, though I'm unsure if an actual firing on Tuesday clears @stake much, since the paper was not written Tuesday night and they must have been aware of it earlier.
Yes, this is exactly what we do. There have been several kernel patches contributed by special effects compainies. Also read about the patches to Gimp, or about the EXR file format.
It is not clear if the schools have 100% choice in what they buy or get donated by Microsoft. There certainly is precedence for Microsoft choosing exactly what they will "donate".
Anybody who really cared about the kids getting as much software as possible would collect their refund and donate it to the schools for normal purchase of Microsoft software. If anything Microsoft does causes a different amount of software/hardware to be delivered to the school than this would then they are pulling a scam.
The admin still cant get information that is not stored on the machine, such as passwords that are stored in the domain controller. Thus doing this does have the potential of getting information they don't already have access to, so it is bad.
Ctrl+alt+del generates the *same* hardware interrupt as any other key. The difference is in what the Windows keyboard driver does with those keystrokes after it gets them.
I would say that this scheme works as I have not heard of a password-stealing program for Windows.
Though I would think a program that displays the login box (as though Ctrl+alt+del was already pushed) would fool a lot of people into typing in their password.
Hitting those keys generates the same interrupt as any other keys. If you point that interrupt at other code, ctrl+alt+del will no longer work.
Old PC BIOS's did actually check for this combination directly in the keyboard interrupt handler and reboot the machine. Thus if the interrupt was not pointed anywhere else it would always reboot the machine, even if the processor was stuck in a loop somewhere. So a crash that made ctrl+alt+del fail to work, while not impossible, was certainly rare, and often was called a "big red switch" crash since you had to turn off the power to get your machine back. All other keystrokes were just stored in a small queue by the interrupt routine and thus would not make any difference unless the processor executed instructions that read from that queue. This is probably what led you to think that ctrl+alt+del had some hardware significance.
I am reasonably certain that BIOS's after about 1986 or so were modified so this default behavior could be turned off. This was certainly a requirement as PC's started to be used in public-access devices such as info kiosks.
I agree. I certainly remember Linux GUI programs did crash all the time, leading to either a logout or to a hung X server, usually hung up so that Ctrl+Fn would not work either. To almost any user this rendered the Linux box as useless as a blue-screened Windows box and you had to reboot. And you had to wait for fsck to complete as well.
However thinking about this I realize that the most recent systems (ie less than 3 years old) have NOT shown any such crashes. It seems KDE and Gnome have improved code quality, or at least set things up so programs that screw up do not take out the window manager.
Oddly enough my newest Linux sytems *do* crash in the kernel more often, apparently due to a faulty network driver. I believe 100% of my crashes are Windows-like kernel crashes nowadays.
The Democrats are far more guilty for helping the RIAA/MPAA than the Republicans, and that is common knowledge here. I am mostly anti-Bush, but I also know that Bush is far better than Feinstein here. I don't think the politics here are as biased as you think.
The "donated" software will be valued at retail value, possibly ten or more times more than the price a school would actually pay for it. Microsoft may also "donate" software that the schools would never buy on their own. And as many have pointed out, Microsoft software donations are actually in Microsoft's benefit, since it encourages lock-in to their product.
I have received several of these for former occupants of the house I am in, but have not recieved one myself, which would indicate that these are accurately being sent only to purchasers of certain Microsoft products. Do these have anything to do with the Lindows offer? They sure do indicate that the details of the settlement have been figured out.
I worked plenty with XIntrinsics and Motif at Sun, where I worked on OLIT (the Open Look Toolkit, which was their XIntrinsics toolkit) and emulation of Motif widgets in OLIT. I also implemented several Motif widgets, including the GridBag layout I designed to get around the horrors of Xi layout (this still lives on in the Java Grid layout widget).
All I can say is that I am personally quite relieved and happy that Motif is gone.
There are probably a dozen projects called "Y", all trying to replace X.
At most only one of these can succeed in becoming the standard. That one deserves to be called "Y". For not it is probably ok for all the competetitors to call themselves "Y" in anticipation of being the winning one.
Seriously, I was wondering if there was an alternative that would prevent "embrace and extend" yet allow your software to be used by products. I don't know if this would work or not:
The license would be even more lenient than the LGPL, but more strict than public domain. The idea is simple: you can treat the code as public domain (ie link it any way you want with your code) provided you do not have to modify any of the source or header files. If you do modify the source or header files, you are required to release all those changes under this license.
The idea is that it is pretty difficult to add a new secret functionality to the kernel without modifying at least one piece of that source code.
The problem is my wording would allow somebody to insert "call_secret_stuff()" into the code and release that, without releasing the source for "call_secret_stuff()". Requirements that the result compile don't help, since they can just make a dummy version and a call that changes the pointer to point to their secret version. I don't know if there is any possible wording that would prevent people from legal embrace&extend.
I never considered the idea that the client library would be GPL. Certainly that would make this useless as it would require *all* software using the screen to be GPL.
I doubt they are that stupid, and the client library will instead be LGPL. The question is whether the server itself can/should be GPL.
Yes you were talking about exactly the same thing he was.
Full off-screen double buffering, with alpha, which allows compositing and transparency when the images are put on the screen.
It appears that in common usage "double buffering" means this. There is another technology that seems to be called "back buffering" nowadays, though many people still call it "double buffering": In back-buffering there is only an off-screen pixel for the *visible* pixels on the screen. This gives the speed and blink-less advantages of double buffering, plus the advantage that the memory usage is strictly limited to exactly the screen area, and it could be faster as clipped graphics can be skipped (contrary to the paper's assertion, clipping is often faster than drawing without clipping). The big disadvantage of back-buffering is that window pixels are necessarily opaque (since they can only belong to one window) and you still get expose events.
The paper is *definately* talking about the new form of double buffering, as used by OS/X.
Well I just created the above paragraph by typing "Paste ", then selecting it, typing Ctrl+C, and then typing Ctrl+V a number of times. This is with KMail, though I assure you it works in all other modern X11 programs.
Oh, you mean *EMACS*. Well I have bad news for you: Emacs runs on Windows! Therefore by your logic, Windows is just as inconsistent as Linux!
No, Motif and XIntrisics were crap. They were being replaced by compainies and programmers that already had paid big money to license the code (much of the basis of the modern toolkits was born then). The closed-source nature was not what made people abandon it. If people actually wanted Motif then Lestif would have been out much sooner.
Slashdot Mirror
The choice to "license" rather than "sell" is up to the software seller.
Some give you a EULA you must click on and then claim you have agreed to a license. There are plenty of arguments that this is invalid, and in fact this is a sale and not a license, anyway.
But it is absolutely certain that any company that wants to "sell" Linux is free to do so. They decide if they want to "license" it, not the government, and are free to decide to "sell".
All this means that the distinction is meaningless.
PS: people who lease cars have signed a great deal of licensing agreements, but they aren't any more liable than a person who bought a car for an argument between GM and Ford.
No, they would only have to remove the GPL code from future versions of their product.
There may be monetary damages paid to the original copyright holder by Microsoft. But copyright law does not allow much else (despite SCO's totally bogus claims).
The C libraray standard requires the digits to be contiguous, so c-'0' is defined to work.
I knew some moron would post this.
Security through obscurity is a BAD idea.
Security through obscurity is an even WORSE idea when your only defense (the obscurity) is lost, which is what has happened in this case.
Though I don't think the government should require GPL, I want to say that I think Microsoft could do quite well even if they GPL'd their software. They can still contract with OEM's to install Windows and customize it and do a contract that is based on the number of units sold, like they do now. Any competitor selling "bob's customized Windows" would have a hard to impossible time competing with that. Also, more seriously, Microsoft could include some closed-source stuff like WMP or something that you are not allowed to run on non-Microsoft Windows.
Personally I feel a government mandate for open *data* standards would be best. If Microsoft writes a clever word processing algorithm, they should be able to profit from that and governments should pay Microsoft for the right to use it. However they currently don't do that, they can write crap because they rely on the lock-in of their closed formats.
Both sides in the argument (both @stake and Geer) agree that the paper was Geer's own writing and was not a "study" with any kind of funding.
Making up your own facts makes it hard for me to believe anything you say.
To be honest, I think Geer has figured out a way to get a lot of publicity, and may have already known he was going to get fired for some other reason and decided to cause a lot of trouble. It would make absolutely no sense for either Microsoft or @stake to dirty their reputations by doing this quickly. Even if you assumme 100% evil it would make more sense for them to wait a month or so or even fabricate a more definate reason to fire him.
Why don't you read the article again. It was @stake that claimed they fired him on Tuesday. Nobody (him included) seemed to know anything about it until Thursday (or at least so he claims).
That sounds very strange, though I'm unsure if an actual firing on Tuesday clears @stake much, since the paper was not written Tuesday night and they must have been aware of it earlier.
Suprisingly little is done in photoshop. We usually use Macs for it.
Yes, this is exactly what we do. There have been several kernel patches contributed by special effects compainies. Also read about the patches to Gimp, or about the EXR file format.
It is not clear if the schools have 100% choice in what they buy or get donated by Microsoft. There certainly is precedence for Microsoft choosing exactly what they will "donate".
Anybody who really cared about the kids getting as much software as possible would collect their refund and donate it to the schools for normal purchase of Microsoft software. If anything Microsoft does causes a different amount of software/hardware to be delivered to the school than this would then they are pulling a scam.
The admin still cant get information that is not stored on the machine, such as passwords that are stored in the domain controller. Thus doing this does have the potential of getting information they don't already have access to, so it is bad.
Ctrl+alt+del generates the *same* hardware interrupt as any other key. The difference is in what the Windows keyboard driver does with those keystrokes after it gets them.
I would say that this scheme works as I have not heard of a password-stealing program for Windows.
Though I would think a program that displays the login box (as though Ctrl+alt+del was already pushed) would fool a lot of people into typing in their password.
Hitting those keys generates the same interrupt as any other keys. If you point that interrupt at other code, ctrl+alt+del will no longer work.
Old PC BIOS's did actually check for this combination directly in the keyboard interrupt handler and reboot the machine. Thus if the interrupt was not pointed anywhere else it would always reboot the machine, even if the processor was stuck in a loop somewhere. So a crash that made ctrl+alt+del fail to work, while not impossible, was certainly rare, and often was called a "big red switch" crash since you had to turn off the power to get your machine back. All other keystrokes were just stored in a small queue by the interrupt routine and thus would not make any difference unless the processor executed instructions that read from that queue. This is probably what led you to think that ctrl+alt+del had some hardware significance.
I am reasonably certain that BIOS's after about 1986 or so were modified so this default behavior could be turned off. This was certainly a requirement as PC's started to be used in public-access devices such as info kiosks.
I agree. I certainly remember Linux GUI programs did crash all the time, leading to either a logout or to a hung X server, usually hung up so that Ctrl+Fn would not work either. To almost any user this rendered the Linux box as useless as a blue-screened Windows box and you had to reboot. And you had to wait for fsck to complete as well.
However thinking about this I realize that the most recent systems (ie less than 3 years old) have NOT shown any such crashes. It seems KDE and Gnome have improved code quality, or at least set things up so programs that screw up do not take out the window manager.
Oddly enough my newest Linux sytems *do* crash in the kernel more often, apparently due to a faulty network driver. I believe 100% of my crashes are Windows-like kernel crashes nowadays.
The Democrats are far more guilty for helping the RIAA/MPAA than the Republicans, and that is common knowledge here. I am mostly anti-Bush, but I also know that Bush is far better than Feinstein here. I don't think the politics here are as biased as you think.
You have got to be fucking kidding.
Lindows is NOT suing Microsoft. Even by Microsoft's claims they are at worst fooling people into thinking they are going to get money.
The only way Lindows would be equivalent to SCO would be if they threatened all *users* of Microsoft software with a $699 lawsuit.
The "donated" software will be valued at retail value, possibly ten or more times more than the price a school would actually pay for it. Microsoft may also "donate" software that the schools would never buy on their own. And as many have pointed out, Microsoft software donations are actually in Microsoft's benefit, since it encourages lock-in to their product.
What are the "settlement agreement" letters?
I have received several of these for former occupants of the house I am in, but have not recieved one myself, which would indicate that these are accurately being sent only to purchasers of certain Microsoft products. Do these have anything to do with the Lindows offer? They sure do indicate that the details of the settlement have been figured out.
I worked plenty with XIntrinsics and Motif at Sun, where I worked on OLIT (the Open Look Toolkit, which was their XIntrinsics toolkit) and emulation of Motif widgets in OLIT. I also implemented several Motif widgets, including the GridBag layout I designed to get around the horrors of Xi layout (this still lives on in the Java Grid layout widget).
All I can say is that I am personally quite relieved and happy that Motif is gone.
There are probably a dozen projects called "Y", all trying to replace X.
At most only one of these can succeed in becoming the standard. That one deserves to be called "Y". For not it is probably ok for all the competetitors to call themselves "Y" in anticipation of being the winning one.
The license would be even more lenient than the LGPL, but more strict than public domain. The idea is simple: you can treat the code as public domain (ie link it any way you want with your code) provided you do not have to modify any of the source or header files. If you do modify the source or header files, you are required to release all those changes under this license.
The idea is that it is pretty difficult to add a new secret functionality to the kernel without modifying at least one piece of that source code.
The problem is my wording would allow somebody to insert "call_secret_stuff()" into the code and release that, without releasing the source for "call_secret_stuff()". Requirements that the result compile don't help, since they can just make a dummy version and a call that changes the pointer to point to their secret version. I don't know if there is any possible wording that would prevent people from legal embrace&extend.
I never considered the idea that the client library would be GPL. Certainly that would make this useless as it would require *all* software using the screen to be GPL.
I doubt they are that stupid, and the client library will instead be LGPL. The question is whether the server itself can/should be GPL.
Yes you were talking about exactly the same thing he was.
Full off-screen double buffering, with alpha, which allows compositing and transparency when the images are put on the screen.
It appears that in common usage "double buffering" means this. There is another technology that seems to be called "back buffering" nowadays, though many people still call it "double buffering": In back-buffering there is only an off-screen pixel for the *visible* pixels on the screen. This gives the speed and blink-less advantages of double buffering, plus the advantage that the memory usage is strictly limited to exactly the screen area, and it could be faster as clipped graphics can be skipped (contrary to the paper's assertion, clipping is often faster than drawing without clipping). The big disadvantage of back-buffering is that window pixels are necessarily opaque (since they can only belong to one window) and you still get expose events.
The paper is *definately* talking about the new form of double buffering, as used by OS/X.
Paste Paste Paste Paste Paste Paste Paste Paste Paste Paste Paste Paste Paste Paste Paste Paste Paste Paste Paste
Well I just created the above paragraph by typing "Paste ", then selecting it, typing Ctrl+C, and then typing Ctrl+V a number of times. This is with KMail, though I assure you it works in all other modern X11 programs.
Oh, you mean *EMACS*. Well I have bad news for you: Emacs runs on Windows! Therefore by your logic, Windows is just as inconsistent as Linux!
No, Motif and XIntrisics were crap. They were being replaced by compainies and programmers that already had paid big money to license the code (much of the basis of the modern toolkits was born then). The closed-source nature was not what made people abandon it. If people actually wanted Motif then Lestif would have been out much sooner.