No guarantee whatsoever that the human-readable part and the machine-readable part are identical.
As somebody else pointed out, this could be confirmed by randomly sampling the ballots and checking that the two parts matched.
Personally I think a 2-digit machine-readable number can be printed next to the name chosen. This number would also be displayed on the touch-screen and printed in the voter information booklet next to each candidate. Thus a really worried voter can check that the number is correct as well as the name. And voting officials can randomly check ballots for matches and check that the machines scanning the ballots also read the correct numbers.
I also think it would be acceptable for the touch-screens themselves to store and report their totals. This would be yet another check, and would probably be used for the official results as long as the results are undisputed. Random checks of ballots, ballot scans, and hand-counting of some precients to check for matches between the ballots and the machine reports would make it impossible or extremely risky to cheat.
I think you are seriously confused about the advantage of open-source software for this. "Grandma" does not have to understand the code one bit. What "grandma" has to know is that "smart people" from many different parties have looked at it and offered the opinion that they think it is ok. This is definately not true right now, and the only way it could be true is with an open-source system. (any NDA or otherwise closed system, no matter how many experts say they have analyzed the code, still has the possibility that every expert was paid off).
It may be possible with lots of cryptography and a central server to make a system where it is impossible to produce a mathematically-verified result unless a different registered voter pushes the actual button at a voting station, and that replacement of the software on that voting station cannot produce legal votes. If this is possible, I'm sure the cryptographic math will be enormously complex and only understandable by very, very few people. I for instance would not understand it one bit. However if it was open-source, I would be very certain that it actually works, because I know that many people who do know about it have looked at it.
Personally I suspect such a system as described above is impossible. If replacing the software on the voting station would allow cheating, the fact that it is open-source and mathematically bullet-proof is not going to help. The best solution I can think of is to have the machine print out a ballot that clearly indicates the vote, so the voter can confirm that, and deposit it in a ballot box. Then (even if the results are undisputed) some random precients are audited to check for a match between the paper ballots and thier voting machines, thus making the chance that anybody fixing the machines to lie would be caught.
Such a system has the advantage that "grandma" probably can understand it, too.
Hardline GPL zealots will of course continue to bray that "they released the product so they MUST release ALL of their sources".
I challenge you to find a real GPL zealot who says this. This is actually what is said by people attacking the GPL. Hardline GPL zealots know the limits of copyright law: the worst possible thing that can happen is a lawsuit by the original author, it is impossible to force somebody to lose their own copyright.
Also there is also a 4th possibility which I think likely:
4. Rearrange things so the iTerm portion is sufficently seperated from the rest to satisfy the GPL, and only release the code for that part. It sounds as though adding a few things to iTerm so it can listen to some outside commands would be acceptable. Perhaps the program can be used unchanged, if OSX proviedes enough ways to fake keystrokes or other events to the iTerm program.
Almost certainly the blog data will be stored in a central location, and you will probably need an MSN account to use this (though they may try to make MSN come with Windows for "free")
I'm under the impression that lying is allowed under a pure Libertarian system. In theory others will detect the lies and correct them.
To be truthful, Microsoft is starting to rely more and more on government intervention. Primarily laws against reverse engineering, corporate espinoage, copyrights, and the DMCA. I think all this would not exist under a pure Libertarian system. Thus Libertarians can make a good argument that even a monopoly like Microsoft could fall.
Personally I am not a big fan of Libertarians, but they do have consistent arguments. There are also variations on pure Libertarianism where there are some government rules, for instance copyrights would exist, or there would be laws against monopolies (which can resemble governments in their control over people's lives).
TrollTech does pretty well selling Qt. If they did not GPL the version they give away they would not be able to do that. In my opinion the GPL actually encourages selling software, at least compared to public domain.
Putting something out as public domain completely gives up any advantages you have for having the copyright. If the GPL restricts freedom in any way, it makes the "more free" version valuable and thus allows it to be sold. If you put code out under the public domain you completely lose any possible ability to sell a version.
I do agree that it is impossible to sell either GPL code itself or public domain code, and arguments about selling support or any other service apply equally to both of them. But the GPL actually allows the owner to do a few other things with the code for profit, and in fact I see it as a very capitalistic and greedy way to let people see the source code.
Personally I LGPL all my stuff, with an added statement that you are free to use it static-linked with a closed-source program provided you don't modify the LGPL code itself, and if you do modify it you only need to release the modifications, and then use it in your closed-source program anyway. I do agree that RMS's plan is to force everything to be open source, while I would like to make "embrace and extend" illegal but I don't mind somebody using my code to help make their own clever idea work and sell it. Unfortunately I see nobody else in my position, people are either RMS-like or they want to grab and extend code in any way they want.
Thats "redistribute", not "use". The GPL says NOTHING about usage of the software, anybody can use it.
What the GPL does is grant you some exceptions to normal copyright so that you can redistribute the software without compensating the original authors.
Read up on "embrace and extend". It is quite possible for a proprietary software maker to make the public domain version quite useless, thus "capturing" the software. The GPL prevents this.
That still does not explain it. Plenty of products are put on the screen without compensation from the companies. There is a big stink right now where Catapiller is suing the George of the Jungle sequel for using their machines as tools of evil.
It would seem to me that a lot of movie makers when told "I need a computer running in this scene" would grab a convienent computer, which has a 90% chance of running Windows, and put it in the scene. Rather than having a graphic designer paint a fake screen or animation and display it. As is often pointed out here, the average member of the public is unaware that there are alternatives to Windows, they consider it a natural artifact needed to make an operational computer, if they think about it as an object at all.
Yet for some reason it seems Windows is vastly unrepresented in movies, and this is a mystery.
The innovation that I give Microsoft credit for is to realize that when switching in an environment where an arbitrary subset of windows could be iconized, you could use the same action to switch to any window, rather than treating the open and iconized windows as two different sets. All X and XView environments I ever saw before that would use completely different actions to switch to an already-open window versus and iconized one. This still remains in CDE, where Alt+Tab will only switch between opened windows.
On the Mac in 1985, with Switcher, all programs other than the current one were effectively iconized (ie all windows other than the current app were invisible). And unless this program also let you switch between two or more windows in the same application using the same actions (which I doubt), it does not fall under this innovation, as it only switched between iconized windows.
Absolutely, the pop-up window displaying an icon for all the apps is a MicroSoft invention.
Using Alt+Tab to change windows existed earlier, but in all versions I ever saw it just changed between opened windows, you had to use the mouse to get at iconized windows. So I think even the idea of having a keystroke that selects a possibly iconized window is a MicroSoft invention, this is vastly more important than just a pop-up window of large icons (personally I don't understand why they did not copy the taskbar appearance with text as well).
X has lots of problems, but the network transparency is NOT one of them. This is a myth that comes up all the time, by amateurs that somehow picture it calling a central server at MIT for every graphics call.
On a mondern system (with security) there HAS to be a context switch some time between a user program producing the graphics and the system drawing on the screen. The network transparency adds zero overhead on any modern system, in fact it encourages reduction of overhead by forcing the batching of requests into single context switches. When anybody says that Windows can do each call in 1 context switch, I have to point out that X (if it was properly designed to not require so damn many syncrhonous calls) can do tens of thousands of calls in 2 context switches.
X's #1 problem is the bad graphics model which means that drawing anything more complicated that 1985 graphics (such as anti-aliased shapes) requires you to draw an image and send it, which is going to be slow even if the app could draw directly into the on-screen image buffer.
X's #2 problem, and really the cause of perceived slowness, is that seperate window manager, and people are going to have to face reality and move the window borders and resizing and all other drawing into the app's toolkit, so that synchronization between that and the rest of the app's display can be preserved. Notice that nobody complains that moving things inside the apps is slow.
Nonsense. This is exactly the same problem you would have if somebody sent you a Windows.exe that requires a.dll and they did not send you the.dll. In fact we screwed up the first release of our own Windows software for exactly the same reason, we neglected to include the.dll files the Intel compiler required.
The X problem is definately caused by your window manager.
Most likely it has a bug in how it checks for events so that it tries to read from X when there are no events there, and eventually times out. In previous versions of Linux, due to different timing, this did not happen.
It is extremely unlikely that a bug in the kernel, or even in the X server, would cause only keyboard shortcuts to the window manager to slow down so drastically.
WRONG!!!! You are free to use gnumeric any way you want. You do not need a license to do so.
What GNU does is grant you a right to violate the copyright on the source code for gnumeric. You are free to ignore this right, and most users of gnumeric do ignore it.
Having used tcsh quite a lot (on Linux and Irix and Cygwin) and having to switch to bash, I have to say that 95% of the problem is that I am very used to typing "setenv", and also a lot of "sourced" type scripts that do nothing except setenv (ie we use lots of "source setup_foobar" to get around stupid programs that have conflicting requirements for environment variables)
A big help would be a simple alias for bash that makes setenv work like tcsh. Just put this in the default login or something.
SCO does not have to GPL their code. They only need to identify and prove what portion of Linux is their copyrighted code, that was either mistakenly or fraudently inserted into Linux, and it will be removed.
SCO has NO intention of "fighting the GPL in court". The GPL is the only reason they have any rights to redistribute or work on Linux in the first place. Nobody in the world wants to "defeat" the GPL because that either means they have less rights to the code, or conversely that copyrighting anything is illegal.
It requires a tremendous amount of energy to slow down an object enough to get it to hit the sun. Anything thrown from a space elevator would just end up in a different orbit about the sun.
NOBODY thinks voters are going to exploit bugs in the terminals. All the worry is about the company itself or others with access to the secret parts of the protocol changing the results. In this case I think Open Source is vital.
My point was that a lot of people seem to think Open Source == GPL. In fact it can be totally copyrighted and patented so nobody can "steal" it. In this case 100% of the benifit of open source is that it can be examined and everybody knows how it works. The fact that people could make their own implementations is irrelevant.
I don't think there can be any acceptable solution unless a large layer of it is open-source.
Unlike your blathering, this does not mean it has to be GPL or use Linux.
Even the Microsoft Access solution those idiots at Diebold seem to like would be acceptable if the entire Microsoft Access / VB program was available and could be analyized by Access experts at Microsoft and outside Microsoft. There is not even any need for the CE or Access source code itself, as long as experts in it are quite certain that no reliance on a bug or misfeature inside access could be exploited by the code.
By "open source" I mean that anybody can look at the source. It can still be copyrighted or even patented, it does not mean GPL, and it should be easy for a commercial company to make quite a good profit by making the best version.
Unfortunatley even if the entire system was open-source I think there is going to have to be the printer and paper ballot trail. Somebody with a plan could still replace every machine with an altered version that is not actually running the peer-reviewed code, and this is the only way I can see to solve this.
The reason they do this is that otherwise an FX company could buy only a few copies of Maya and use a hundred copies of the free version for their artists to do all the animation. They then use the official copies only when needed to render or otherwise get around the watermarks.
Making the files unreadable is about the only way to prevent this, and I can't blame them for that.
Slashdot Mirror
No guarantee whatsoever that the human-readable part and the machine-readable part are identical.
As somebody else pointed out, this could be confirmed by randomly sampling the ballots and checking that the two parts matched.
Personally I think a 2-digit machine-readable number can be printed next to the name chosen. This number would also be displayed on the touch-screen and printed in the voter information booklet next to each candidate. Thus a really worried voter can check that the number is correct as well as the name. And voting officials can randomly check ballots for matches and check that the machines scanning the ballots also read the correct numbers.
I also think it would be acceptable for the touch-screens themselves to store and report their totals. This would be yet another check, and would probably be used for the official results as long as the results are undisputed. Random checks of ballots, ballot scans, and hand-counting of some precients to check for matches between the ballots and the machine reports would make it impossible or extremely risky to cheat.
I think you are seriously confused about the advantage of open-source software for this. "Grandma" does not have to understand the code one bit. What "grandma" has to know is that "smart people" from many different parties have looked at it and offered the opinion that they think it is ok. This is definately not true right now, and the only way it could be true is with an open-source system. (any NDA or otherwise closed system, no matter how many experts say they have analyzed the code, still has the possibility that every expert was paid off).
It may be possible with lots of cryptography and a central server to make a system where it is impossible to produce a mathematically-verified result unless a different registered voter pushes the actual button at a voting station, and that replacement of the software on that voting station cannot produce legal votes. If this is possible, I'm sure the cryptographic math will be enormously complex and only understandable by very, very few people. I for instance would not understand it one bit. However if it was open-source, I would be very certain that it actually works, because I know that many people who do know about it have looked at it.
Personally I suspect such a system as described above is impossible. If replacing the software on the voting station would allow cheating, the fact that it is open-source and mathematically bullet-proof is not going to help. The best solution I can think of is to have the machine print out a ballot that clearly indicates the vote, so the voter can confirm that, and deposit it in a ballot box. Then (even if the results are undisputed) some random precients are audited to check for a match between the paper ballots and thier voting machines, thus making the chance that anybody fixing the machines to lie would be caught.
Such a system has the advantage that "grandma" probably can understand it, too.
I challenge you to find a real GPL zealot who says this. This is actually what is said by people attacking the GPL. Hardline GPL zealots know the limits of copyright law: the worst possible thing that can happen is a lawsuit by the original author, it is impossible to force somebody to lose their own copyright.
Also there is also a 4th possibility which I think likely:
4. Rearrange things so the iTerm portion is sufficently seperated from the rest to satisfy the GPL, and only release the code for that part. It sounds as though adding a few things to iTerm so it can listen to some outside commands would be acceptable. Perhaps the program can be used unchanged, if OSX proviedes enough ways to fake keystrokes or other events to the iTerm program.
Almost certainly the blog data will be stored in a central location, and you will probably need an MSN account to use this (though they may try to make MSN come with Windows for "free")
I'm under the impression that lying is allowed under a pure Libertarian system. In theory others will detect the lies and correct them.
To be truthful, Microsoft is starting to rely more and more on government intervention. Primarily laws against reverse engineering, corporate espinoage, copyrights, and the DMCA. I think all this would not exist under a pure Libertarian system. Thus Libertarians can make a good argument that even a monopoly like Microsoft could fall.
Personally I am not a big fan of Libertarians, but they do have consistent arguments. There are also variations on pure Libertarianism where there are some government rules, for instance copyrights would exist, or there would be laws against monopolies (which can resemble governments in their control over people's lives).
TrollTech does pretty well selling Qt. If they did not GPL the version they give away they would not be able to do that. In my opinion the GPL actually encourages selling software, at least compared to public domain.
Putting something out as public domain completely gives up any advantages you have for having the copyright. If the GPL restricts freedom in any way, it makes the "more free" version valuable and thus allows it to be sold. If you put code out under the public domain you completely lose any possible ability to sell a version.
I do agree that it is impossible to sell either GPL code itself or public domain code, and arguments about selling support or any other service apply equally to both of them. But the GPL actually allows the owner to do a few other things with the code for profit, and in fact I see it as a very capitalistic and greedy way to let people see the source code.
Personally I LGPL all my stuff, with an added statement that you are free to use it static-linked with a closed-source program provided you don't modify the LGPL code itself, and if you do modify it you only need to release the modifications, and then use it in your closed-source program anyway. I do agree that RMS's plan is to force everything to be open source, while I would like to make "embrace and extend" illegal but I don't mind somebody using my code to help make their own clever idea work and sell it. Unfortunately I see nobody else in my position, people are either RMS-like or they want to grab and extend code in any way they want.
Thats "redistribute", not "use". The GPL says NOTHING about usage of the software, anybody can use it.
What the GPL does is grant you some exceptions to normal copyright so that you can redistribute the software without compensating the original authors.
I think you swapped "beer" and "speech" there.
Read up on "embrace and extend". It is quite possible for a proprietary software maker to make the public domain version quite useless, thus "capturing" the software. The GPL prevents this.
That still does not explain it. Plenty of products are put on the screen without compensation from the companies. There is a big stink right now where Catapiller is suing the George of the Jungle sequel for using their machines as tools of evil.
It would seem to me that a lot of movie makers when told "I need a computer running in this scene" would grab a convienent computer, which has a 90% chance of running Windows, and put it in the scene. Rather than having a graphic designer paint a fake screen or animation and display it. As is often pointed out here, the average member of the public is unaware that there are alternatives to Windows, they consider it a natural artifact needed to make an operational computer, if they think about it as an object at all.
Yet for some reason it seems Windows is vastly unrepresented in movies, and this is a mystery.
The innovation that I give Microsoft credit for is to realize that when switching in an environment where an arbitrary subset of windows could be iconized, you could use the same action to switch to any window, rather than treating the open and iconized windows as two different sets. All X and XView environments I ever saw before that would use completely different actions to switch to an already-open window versus and iconized one. This still remains in CDE, where Alt+Tab will only switch between opened windows.
On the Mac in 1985, with Switcher, all programs other than the current one were effectively iconized (ie all windows other than the current app were invisible). And unless this program also let you switch between two or more windows in the same application using the same actions (which I doubt), it does not fall under this innovation, as it only switched between iconized windows.
Absolutely, the pop-up window displaying an icon for all the apps is a MicroSoft invention.
Using Alt+Tab to change windows existed earlier, but in all versions I ever saw it just changed between opened windows, you had to use the mouse to get at iconized windows. So I think even the idea of having a keystroke that selects a possibly iconized window is a MicroSoft invention, this is vastly more important than just a pop-up window of large icons (personally I don't understand why they did not copy the taskbar appearance with text as well).
X has lots of problems, but the network transparency is NOT one of them. This is a myth that comes up all the time, by amateurs that somehow picture it calling a central server at MIT for every graphics call.
On a mondern system (with security) there HAS to be a context switch some time between a user program producing the graphics and the system drawing on the screen. The network transparency adds zero overhead on any modern system, in fact it encourages reduction of overhead by forcing the batching of requests into single context switches. When anybody says that Windows can do each call in 1 context switch, I have to point out that X (if it was properly designed to not require so damn many syncrhonous calls) can do tens of thousands of calls in 2 context switches.
X's #1 problem is the bad graphics model which means that drawing anything more complicated that 1985 graphics (such as anti-aliased shapes) requires you to draw an image and send it, which is going to be slow even if the app could draw directly into the on-screen image buffer.
X's #2 problem, and really the cause of perceived slowness, is that seperate window manager, and people are going to have to face reality and move the window borders and resizing and all other drawing into the app's toolkit, so that synchronization between that and the rest of the app's display can be preserved. Notice that nobody complains that moving things inside the apps is slow.
Nonsense. This is exactly the same problem you would have if somebody sent you a Windows .exe that requires a .dll and they did not send you the .dll. In fact we screwed up the first release of our own Windows software for exactly the same reason, we neglected to include the .dll files the Intel compiler required.
The X problem is definately caused by your window manager.
Most likely it has a bug in how it checks for events so that it tries to read from X when there are no events there, and eventually times out. In previous versions of Linux, due to different timing, this did not happen.
It is extremely unlikely that a bug in the kernel, or even in the X server, would cause only keyboard shortcuts to the window manager to slow down so drastically.
Set the environment varialbe LD_LIBRARY_PATH to the directory the library is in.
/foo/bar"
In bash this is "export LD_LIBRARY_PATH
./worm_draw: error while loading shared libraries: libcxa.so.3: cannot open shared object file: No such file or directory
WRONG!!!! You are free to use gnumeric any way you want. You do not need a license to do so.
What GNU does is grant you a right to violate the copyright on the source code for gnumeric. You are free to ignore this right, and most users of gnumeric do ignore it.
Having used tcsh quite a lot (on Linux and Irix and Cygwin) and having to switch to bash, I have to say that 95% of the problem is that I am very used to typing "setenv", and also a lot of "sourced" type scripts that do nothing except setenv (ie we use lots of "source setup_foobar" to get around stupid programs that have conflicting requirements for environment variables)
A big help would be a simple alias for bash that makes setenv work like tcsh. Just put this in the default login or something.
SCO does not have to GPL their code. They only need to identify and prove what portion of Linux is their copyrighted code, that was either mistakenly or fraudently inserted into Linux, and it will be removed.
SCO has NO intention of "fighting the GPL in court". The GPL is the only reason they have any rights to redistribute or work on Linux in the first place. Nobody in the world wants to "defeat" the GPL because that either means they have less rights to the code, or conversely that copyrighting anything is illegal.
It requires a tremendous amount of energy to slow down an object enough to get it to hit the sun. Anything thrown from a space elevator would just end up in a different orbit about the sun.
It works fine in Konqueror.
NOBODY thinks voters are going to exploit bugs in the terminals. All the worry is about the company itself or others with access to the secret parts of the protocol changing the results. In this case I think Open Source is vital.
My point was that a lot of people seem to think Open Source == GPL. In fact it can be totally copyrighted and patented so nobody can "steal" it. In this case 100% of the benifit of open source is that it can be examined and everybody knows how it works. The fact that people could make their own implementations is irrelevant.
I don't think there can be any acceptable solution unless a large layer of it is open-source.
Unlike your blathering, this does not mean it has to be GPL or use Linux.
Even the Microsoft Access solution those idiots at Diebold seem to like would be acceptable if the entire Microsoft Access / VB program was available and could be analyized by Access experts at Microsoft and outside Microsoft. There is not even any need for the CE or Access source code itself, as long as experts in it are quite certain that no reliance on a bug or misfeature inside access could be exploited by the code.
By "open source" I mean that anybody can look at the source. It can still be copyrighted or even patented, it does not mean GPL, and it should be easy for a commercial company to make quite a good profit by making the best version.
Unfortunatley even if the entire system was open-source I think there is going to have to be the printer and paper ballot trail. Somebody with a plan could still replace every machine with an altered version that is not actually running the peer-reviewed code, and this is the only way I can see to solve this.
The reason they do this is that otherwise an FX company could buy only a few copies of Maya and use a hundred copies of the free version for their artists to do all the animation. They then use the official copies only when needed to render or otherwise get around the watermarks.
Making the files unreadable is about the only way to prevent this, and I can't blame them for that.