There's another, simpler explanation for Kaplan's ruling. The law firm he worked for (while he was working there and after he became a judge) worked for Toshiba and Time Warner on antitrust matters related to DVDs. I'm sure he didn't want to embarass his old colleagues by ruling against the legal scheme they created (with CSS licensing and the DMCA), whether or not he was part of creating it.
A DMCA-copycat is moving through the political system in Australia. I can't find the article I read about it, so I don't know how far it is, but 2600.au will probably be in trouble soon enough.
Just for the fun of it, I tried to find out who had voted for the DMCA, so I could make an extra effort not to vote for them. But they did it with a voice vote - there is no record. How reassuring.
The DMCA was passed by voice vote in the House and by Unanimous Consent in the Senate. As I understand from watching CSPAN, both of these mean that no one objected strongly enough to passage to insist on a record of the vote (only a single objection is necessary to force a recorded vote). So the answer to the question of who voted for the DMCA: they all did.
This is only to be expected. The movie industry spent its lobbying money lavishly to buy the politicians. As we have just seen, since the politicians appoint and confirm the judges, this means they are also slowly buying the judiciary. We might win in the courts this time, but unless something changes, we will lose in the future.
I personally think this dual licensing is unfortunate. Every piece of code in Mozilla
is under a free-software license. I do not not think it should be necessary for Mozilla to be dual-licensed for it to be legally combined with GPLed software. The GPL exists to promote free software. Free software does not exist to promote the GPL. For GPLed code to not cooperate with proprietary code is expected, for GPLed code to not cooperate with other free software is sad. When the GPL makes it harder for free-software projects to cooperate (like Mozilla and Kaffe), the GPL hurts the cause of free software and so it should be changed to solve the problem.
That the GPL made this mistake is understandable. It was created at a time when free software was being invented, and it was unclear whether there would be reasons for important non-GPL free-software projects. Now we know that many free-software projects, corporations and other organizations prefer to release code under free-software licenses tailored to their needs. I think the GPL needs to be changed to permit linking with software under any free-software license as long as the entire derived work is under free software licenses (no using other licenses as "shields" to get GPLed code into a proprietary application). Expressing this requirement is a license is difficult, and that is why I didn't want to see this dual-licensing. Mozilla is one of the most prominent non-GPL codebases and provided an important incentive to solve the problem: getting GPLed software to "play with" other free software without letting it "play with" proprietary software. "Keeping free software free" is an important goal of the GPL, but it is not the only goal of the GPL, and the GPL should be changed to acknowledge that.
I think that the free software community is weakened because the GPL is actively hostile to non-GPL approaches to free software. I think this discourages experimentation with free-software development and business models and therefore makes it harder for free software to compete with proprietary software. While some of this cost is visible (licensing flamewars, for example) I think the vast majority of it is invisible. How do you count projects not started, and experiments that were never made? Many people here think Galeon is a great project. Would it have been created sooner if the GPL didn't discourage working with non-GPL free software? I think one of the strengths of the free-software community is how easy it is for developers across the world to cooperate with each other. We should be working on enhancing this strength, not standing by and letting the GPL dilute it.
The issue is not that CSS was weak, the issue is that they implemented CSS as a software program
on a general-purpose computer. Given that, it was
only a matter of time until CSS was reverse-engineered. If CSS had been better, the algorithm would have fallen, but only the player
keys of software players would have fallen (the ones buried it hardware would likely not have been
found). They could have revoked those keys and gone on pressing new disks that couldn't be accessed with the revoked keys, but they wanted
to beef up their installed base with software players so they couldn't be safe.
Why not? There's wheat and there's chaff in
the comments, but as long as the CTO sorts
through them, where's the problem? There's
no sin in asking for help. In fact, I would
say that one of the hallmarks of any good
executive is asking for help when you need
it (or even when you don't if, from your
employer's point of view, getting help is
more efficient than doing it yourself).
The banner ad revenue scenario suggested in the articles is only partly true. While it is true that lots of banner ads will be loaded, and ABC will be paid for them, over the long term articles like this are counterproductive. When people are angry at an article, I think they're in no mood to even look at a banner ad, let alone click on it. This means ABC's click-through rates (which are what advertisers really want to pay for) will plummet, so advertisers will be convinced banner ads are ineffective and pay less in the future. This may be something ABC actually wants, since they're a television network, but I don't think the scenario is as simple as it is portrayed in the article.
To my mind, trade secret law is the least objectionable of all IP-related law. The main thrust of trade secret law is to prevent people from "misappropriating" trade secrets. What this generally means is that if you agreed to keep someone else's economically significant secret, you can't turn around later and disclose it. Also, if you break into their factory/computers/etc. to get the trade secret, you can't use it. BUT, if you reverse-engineer the trade secret, there is nothing the holder of the trade secret can do about it. Now I admit that companies often try to abuse trade secret law (just look at Microsoft and Kerberos), but there are far more ways to route around trade secrets than copyrights and patents, and that makes them much less dangerous. And if companies couldn't sign agreements to give others access to their trade secrets, they would be much less efficient. Whenever a company hired a contracter, or worked with another company, or lots of other things, they would have to twist themselves into pretzels to hide their secrets.
Now that I've seen Ingo's answers, I think that
the question about versions of TUX for other OSes
turned out to be phrased poorly. The questions I think were intended were:
1) Is the TUX interface (now revealed as the TUX
system call) something that can be used and
implemented by non-GPLed software? The use case
has been answered affirmatively and I can't see any reason why the answer would be different for the implementation case, but it hasn't been addressed yet.
2) If other OSes decide to implement the TUX
system call (or another TUX-like interface) are
you willing to work with other OSes to keep the
interfaces compatible (to ease the headaches for
software running on top of these interfaces like Apache)? Again, I don't see why not, but it hasn't been addressed.
You are deliberately misrepresenting Open Source arguments. One reason software to be free is to avoid _pointless_ duplication of effort (i.e. when a developer wants something _exactly_ like some other piece of software or something with some general functionality whose details don't matter they can use what already exists rather than writing a new version from scratch). Another reason software should be free is to allow for diversity, so that people can take projects in interesting directions that their creators may not have intended or noticed. These arguments are complementary, not opposed. One way I would expect a GPLed StarOffice to help AbiWord is by allowing StarOffice functionality to be "ported" to AbiWord (or for functionality that the projects philosophically agree on to be abstracted out as libraries that both projects (and others) can use).
Slashdot Mirror
There's another, simpler explanation for Kaplan's ruling. The law firm he worked for (while he was working there and after he became a judge) worked for Toshiba and Time Warner on antitrust matters related to DVDs. I'm sure he didn't want to embarass his old colleagues by ruling against the legal scheme they created (with CSS licensing and the DMCA), whether or not he was part of creating it.
A DMCA-copycat is moving through the political system in Australia. I can't find the article I read about it, so I don't know how far it is, but 2600.au will probably be in trouble soon enough.
Just for the fun of it, I tried to find out who had voted for the DMCA, so I could make an extra effort not to vote for them. But they did it with a voice vote - there is no record. How reassuring.
The DMCA was passed by voice vote in the House and by Unanimous Consent in the Senate. As I understand from watching CSPAN, both of these mean that no one objected strongly enough to passage to insist on a record of the vote (only a single objection is necessary to force a recorded vote). So the answer to the question of who voted for the DMCA: they all did.
This is only to be expected. The movie industry spent its lobbying money lavishly to buy the politicians. As we have just seen, since the politicians appoint and confirm the judges, this means they are also slowly buying the judiciary. We might win in the courts this time, but unless something changes, we will lose in the future.
I personally think this dual licensing is unfortunate. Every piece of code in Mozilla is under a free-software license. I do not not think it should be necessary for Mozilla to be dual-licensed for it to be legally combined with GPLed software. The GPL exists to promote free software. Free software does not exist to promote the GPL. For GPLed code to not cooperate with proprietary code is expected, for GPLed code to not cooperate with other free software is sad. When the GPL makes it harder for free-software projects to cooperate (like Mozilla and Kaffe), the GPL hurts the cause of free software and so it should be changed to solve the problem.
That the GPL made this mistake is understandable. It was created at a time when free software was being invented, and it was unclear whether there would be reasons for important non-GPL free-software projects. Now we know that many free-software projects, corporations and other organizations prefer to release code under free-software licenses tailored to their needs. I think the GPL needs to be changed to permit linking with software under any free-software license as long as the entire derived work is under free software licenses (no using other licenses as "shields" to get GPLed code into a proprietary application). Expressing this requirement is a license is difficult, and that is why I didn't want to see this dual-licensing. Mozilla is one of the most prominent non-GPL codebases and provided an important incentive to solve the problem: getting GPLed software to "play with" other free software without letting it "play with" proprietary software. "Keeping free software free" is an important goal of the GPL, but it is not the only goal of the GPL, and the GPL should be changed to acknowledge that.
I think that the free software community is weakened because the GPL is actively hostile to non-GPL approaches to free software. I think this discourages experimentation with free-software development and business models and therefore makes it harder for free software to compete with proprietary software. While some of this cost is visible (licensing flamewars, for example) I think the vast majority of it is invisible. How do you count projects not started, and experiments that were never made? Many people here think Galeon is a great project. Would it have been created sooner if the GPL didn't discourage working with non-GPL free software? I think one of the strengths of the free-software community is how easy it is for developers across the world to cooperate with each other. We should be working on enhancing this strength, not standing by and letting the GPL dilute it.
The issue is not that CSS was weak, the issue is that they implemented CSS as a software program on a general-purpose computer. Given that, it was only a matter of time until CSS was reverse-engineered. If CSS had been better, the algorithm would have fallen, but only the player keys of software players would have fallen (the ones buried it hardware would likely not have been found). They could have revoked those keys and gone on pressing new disks that couldn't be accessed with the revoked keys, but they wanted to beef up their installed base with software players so they couldn't be safe.
Why not? There's wheat and there's chaff in the comments, but as long as the CTO sorts through them, where's the problem? There's no sin in asking for help. In fact, I would say that one of the hallmarks of any good executive is asking for help when you need it (or even when you don't if, from your employer's point of view, getting help is more efficient than doing it yourself).
The banner ad revenue scenario suggested in the articles is only partly true. While it is true that lots of banner ads will be loaded, and ABC will be paid for them, over the long term articles like this are counterproductive. When people are angry at an article, I think they're in no mood to even look at a banner ad, let alone click on it. This means ABC's click-through rates (which are what advertisers really want to pay for) will plummet, so advertisers will be convinced banner ads are ineffective and pay less in the future. This may be something ABC actually wants, since they're a television network, but I don't think the scenario is as simple as it is portrayed in the article.
To my mind, trade secret law is the least objectionable of all IP-related law. The main thrust of trade secret law is to prevent people from "misappropriating" trade secrets. What this generally means is that if you agreed to keep someone else's economically significant secret, you can't turn around later and disclose it. Also, if you break into their factory/computers/etc. to get the trade secret, you can't use it. BUT, if you reverse-engineer the trade secret, there is nothing the holder of the trade secret can do about it. Now I admit that companies often try to abuse trade secret law (just look at Microsoft and Kerberos), but there are far more ways to route around trade secrets than copyrights and patents, and that makes them much less dangerous. And if companies couldn't sign agreements to give others access to their trade secrets, they would be much less efficient. Whenever a company hired a contracter, or worked with another company, or lots of other things, they would have to twist themselves into pretzels to hide their secrets.
Now that I've seen Ingo's answers, I think that the question about versions of TUX for other OSes turned out to be phrased poorly. The questions I think were intended were:
1) Is the TUX interface (now revealed as the TUX system call) something that can be used and implemented by non-GPLed software? The use case has been answered affirmatively and I can't see any reason why the answer would be different for the implementation case, but it hasn't been addressed yet.
2) If other OSes decide to implement the TUX system call (or another TUX-like interface) are you willing to work with other OSes to keep the interfaces compatible (to ease the headaches for software running on top of these interfaces like Apache)? Again, I don't see why not, but it hasn't been addressed.
You are deliberately misrepresenting Open Source arguments. One reason software to be free is to avoid _pointless_ duplication of effort (i.e. when a developer wants something _exactly_ like some other piece of software or something with some general functionality whose details don't matter they can use what already exists rather than writing a new version from scratch). Another reason software should be free is to allow for diversity, so that people can take projects in interesting directions that their creators may not have intended or noticed. These arguments are complementary, not opposed. One way I would expect a GPLed StarOffice to help AbiWord is by allowing StarOffice functionality to be "ported" to AbiWord (or for functionality that the projects philosophically agree on to be abstracted out as libraries that both projects (and others) can use).