"This is part of an implied social contract that Novell violated, i.e., they acted unethically"
I think passing an ethical judgement on the basis of "an implied social contract" is a bit shaky.
"I will admit that taking a large chunk of BSD code, adding a couple of lines, and then GPLing it would seem unethical"
I don't see how the number of lines or percentage of lines affects the ethical evaluation in any way. I'm not saying that GPLing BSD code is unethical, I'm just saying that it's unethical to the same degree that merely following the "letter of the law" with respect to the GPL would be unethical. As I implied in another post, I think that some within the "free" software movement think that their goals transcend other people's ethics.
"Which is the initial "if it's legal, it's okay" argument."
I think it's a different argument, but if you want to take the 10,000 foot level view, your argument really boils down to "Do what is advantageous to use regardless of what we actually agreed on".
"Containing some code is generally seen as okay. Taking a BSD library and merely adding a couple lines of code and releasing it under GPL is generally seen as not okay. That would be an example of the generally accepted ethics of the OS/FS community."
Yes, but it's the ethics of the code "borrower" that matters to you when the borrowing is being done by the OS/FS community, but when it's been done by somebody else it's not their ethics that matter in that case. Again, it isn't a consistent position, it's just a matter of claiming that OS/FS point of view is all that really matters.
You're really suggesting that the DOJ just ignore the law because you don't like MS. If you don't think the patent system isn't fair (probably most technical people would agree) then lobby to have the law changed. In the meantime you can't prosecute a company for applying for and defending the patents that have been granted to them (even under the infinite "immaginary rules for monolpolies" that so many Slashdotters believe in for MS).
The existence of a community doesn't imply any particular level or type of ethics. There are plenty of communities that we might consider unethical.
Keep in mind that distributing GPL'd software requires an agreement to follow the terms. Agreeing to terms is not the same as agreeing with the principles they are based on. If you ever have rented an apartement, you have probably agreed to a rental agreement, but it's not likely that every provision of the license you agree with in principle.
It's likely that the GNU libraries contain some BSD'd code. This is perfectly legal since the BSD license allows the software to be extended without restriction. Does the fact the FSF doesn't follow the spirit of unrestricted contributions that the BSD license is based on mean that they are unethical or are merely following "letter of the law"?
You're right, there's always a difficulty with creating "universal" ethics. On the other hand, anyone who knows of Novell's history shouldn't expect them to be more "ethical" than any other company.
The problem with the "free" software movement (in the context of ideals) is that it isn't purely a moral one. It chose to use ethically-neutral legal tools to further it's agenda and just as it didn't need any ethical approval from society to create the GPL it shouldn't expect ethical behavior (as it uniquely defines it) from society beyond the "letter of the law".
I don't think they want to do that. They want to be able to say to their customers "if you buy Linux from us and it turns out that there's content that violates one of MS's patents, MS won't sue you. If you buy from somebody else like Red Hat, you're on your own". It's basically an insurance policy for Novell's customers.
But it's not clear that MS is acting unethically in this case. The "ethics of Open Source and Free Software" to the extent that they are explicitly defined at all, are just the opinions of a group of people. They decided to forego their IP rights somewhat and created a license to control access to their created works. If that license is flawed and doesn't fully represent their intent, is it unethical for others to take the license at face value? Is there really any individual that has been elected by all the F/OSS contributors to speak on their behalf on ehtical matters anyway?
I don't know how many free software projects would be able to do this.
In the case of Linux, this "modified license" approach would require Linux to be rewritten from scratch. The GPL won't allow you to make a derivative work without licensing it under the same version of the GPL that Linux already uses. Ironically, the same legal constraint that keeps closed source vendors from making proprietary extentions to Linux also prevents the use of more restrictive licenses.
In the business world you have to differentiate your product to survive. That in my view is the fundamental incompatability between a for-profit company and GPL'd software. That's not to say that there aren't a lot of companies using F/OSS software, but many of them (if not most) are avoiding triggering the GPL and are just using GPL'd code as if it were licensed under BSD and are not contributing anything back.
The bottom line is that you should not assume that licensing your code under the GPL or contributing to a GPL'd project will keep you from being ripped-off. It won't.
If you think sex without hookers is free you must not have much experience with women. But seriously, for a software product to be adopoted widely, it must not require service agreements for the vast majority of its users.
Technically, you are correct, but he was talking about Novell making deals on behalf of other open source vendors, not the customers. Whatever legal basis MS has today to sue Samba for a patent violation, it will be no greater if this deal goes through.
I think an agreement like this and Red Hat's recent problems indicate that the business model that has been touted for open source isn't sustainable. Ultimately people won't pay you money for something that they get for free elsewhere. If, in fact, Linux becomes more mainstream, all distros will be easier to use and greater customer experience will greatly reduce the need for the kind of hand-holding that Red Hat provides.
"The issue of delegates I agree with to some degree, as it is nice syntactic sugar, but one that is, again, easily done equivalently well through the use of listener interfaces. Admittedly, it uses more lines of code, but it also guarantees that people actually impliment the interface, as opposed to simply subscribing to events haphazardly and hoping they got the right ones."
It's pretty clear that using interfaces for events is a major kludge. If you have to handle the same event from different objects than you have to have an adapter.
Your comment about "hoping they got the right ones" doesn't make any sense to me. If you don't know what events you want to hook up to you're going to have a problem that neither Java nor C# can prevent.
I think for your argument to explain why Java apps are slow, the vast majority of Java programmers would have to be unskilled, not just "a number" of them. I think you're grasping at straws to defend Java performance.
In case you somehow have assumed that I'm just a typical consumer, I actually attended the Windows NT launch video conference. So, yes, NT was sold to people very much like ME although I didn't actually program for it until 1996.
As a professional software programmer for over 20 years, I'd advise anyone who wishes to still be programming professionally in their 50's to spend a summer selling used cars. Being able to sell yourself will be much more important in the later years of your career than your technical knowledge will be.
"And the 2600 was definitely an instance of nothing but good timing"
As a former 2600 game programmer I disagree. It was the brilliant low-cost, deeply flexible design of the 2600 that kept it dominant when there were plenty of competitors around.
"You're trying to confuse necessary and sufficient in an attempt to disprove that "being DOS" was necessary."
I'm not confusing anything. You said that Windows was popular because it was an extension of MS-DOS. That's simply not true for the reasons I stated.
"MS-DOS existed under the covers in consumer Windows until it was finally superceded by Windows NT in the consumer space."
"At a certain level, anything before XP was still just MS-DOS."
I think you're a bit confused. Apparently you're having trouble deciding if you think NT was the first Windows not based on MS-DOS or whether it was XP (the correct answer is NT).
"In 1995, Microsoft was still selling MS-DOS to people like YOU."
Oh, the shame. In 1995 was I supposed to buy a Mac, or a Sun workstation? Or was I supposed to be using Linux 1.0 without KDE or gnome or any applications beyond utilities like cat and ls?
I think this Java language vs. Swing is just splitting hairs. People don't choose a language in isolation, the bundled libraries are part of the package. In any case, my point is the same. If it's harder to make Swing applications perform well, than that is a negative aspect of choosing the Java platform.
Yes, that was MS's version of the Java language that could have made Java popular on the Windows desktop if Sun hadn't sued them. Sun decided that pure Java was more important than success on the Windows desktop so I guess they should be happy that things turned out the way they wanted.
These days MS isn't interested in Java and Windows developers are even less interested.
Slashdot Mirror
"This is part of an implied social contract that Novell violated, i.e., they acted unethically"
I think passing an ethical judgement on the basis of "an implied social contract" is a bit shaky.
"I will admit that taking a large chunk of BSD code, adding a couple of lines, and then GPLing it would seem unethical"
I don't see how the number of lines or percentage of lines affects the ethical evaluation in any way. I'm not saying that GPLing BSD code is unethical, I'm just saying that it's unethical to the same degree that merely following the "letter of the law" with respect to the GPL would be unethical. As I implied in another post, I think that some within the "free" software movement think that their goals transcend other people's ethics.
"Which is the initial "if it's legal, it's okay" argument."
I think it's a different argument, but if you want to take the 10,000 foot level view, your argument really boils down to "Do what is advantageous to use regardless of what we actually agreed on".
"Containing some code is generally seen as okay. Taking a BSD library and merely adding a couple lines of code and releasing it under GPL is generally seen as not okay. That would be an example of the generally accepted ethics of the OS/FS community."
Yes, but it's the ethics of the code "borrower" that matters to you when the borrowing is being done by the OS/FS community, but when it's been done by somebody else it's not their ethics that matter in that case. Again, it isn't a consistent position, it's just a matter of claiming that OS/FS point of view is all that really matters.
You're really suggesting that the DOJ just ignore the law because you don't like MS. If you don't think the patent system isn't fair (probably most technical people would agree) then lobby to have the law changed. In the meantime you can't prosecute a company for applying for and defending the patents that have been granted to them (even under the infinite "immaginary rules for monolpolies" that so many Slashdotters believe in for MS).
The existence of a community doesn't imply any particular level or type of ethics. There are plenty of communities that we might consider unethical.
Keep in mind that distributing GPL'd software requires an agreement to follow the terms. Agreeing to terms is not the same as agreeing with the principles they are based on. If you ever have rented an apartement, you have probably agreed to a rental agreement, but it's not likely that every provision of the license you agree with in principle.
It's likely that the GNU libraries contain some BSD'd code. This is perfectly legal since the BSD license allows the software to be extended without restriction. Does the fact the FSF doesn't follow the spirit of unrestricted contributions that the BSD license is based on mean that they are unethical or are merely following "letter of the law"?
You're right, there's always a difficulty with creating "universal" ethics. On the other hand, anyone who knows of Novell's history shouldn't expect them to be more "ethical" than any other company.
The problem with the "free" software movement (in the context of ideals) is that it isn't purely a moral one. It chose to use ethically-neutral legal tools to further it's agenda and just as it didn't need any ethical approval from society to create the GPL it shouldn't expect ethical behavior (as it uniquely defines it) from society beyond the "letter of the law".
I don't think they want to do that. They want to be able to say to their customers "if you buy Linux from us and it turns out that there's content that violates one of MS's patents, MS won't sue you. If you buy from somebody else like Red Hat, you're on your own". It's basically an insurance policy for Novell's customers.
But it's not clear that MS is acting unethically in this case. The "ethics of Open Source and Free Software" to the extent that they are explicitly defined at all, are just the opinions of a group of people. They decided to forego their IP rights somewhat and created a license to control access to their created works. If that license is flawed and doesn't fully represent their intent, is it unethical for others to take the license at face value? Is there really any individual that has been elected by all the F/OSS contributors to speak on their behalf on ehtical matters anyway?
The agreement only protects Novell's customers against patent suits, not Novell.
I don't know how many free software projects would be able to do this.
In the case of Linux, this "modified license" approach would require Linux to be rewritten from scratch. The GPL won't allow you to make a derivative work without licensing it under the same version of the GPL that Linux already uses. Ironically, the same legal constraint that keeps closed source vendors from making proprietary extentions to Linux also prevents the use of more restrictive licenses.
In the business world you have to differentiate your product to survive. That in my view is the fundamental incompatability between a for-profit company and GPL'd software. That's not to say that there aren't a lot of companies using F/OSS software, but many of them (if not most) are avoiding triggering the GPL and are just using GPL'd code as if it were licensed under BSD and are not contributing anything back.
The bottom line is that you should not assume that licensing your code under the GPL or contributing to a GPL'd project will keep you from being ripped-off. It won't.
If you think sex without hookers is free you must not have much experience with women. But seriously, for a software product to be adopoted widely, it must not require service agreements for the vast majority of its users.
Technically, you are correct, but he was talking about Novell making deals on behalf of other open source vendors, not the customers. Whatever legal basis MS has today to sue Samba for a patent violation, it will be no greater if this deal goes through.
I think an agreement like this and Red Hat's recent problems indicate that the business model that has been touted for open source isn't sustainable. Ultimately people won't pay you money for something that they get for free elsewhere. If, in fact, Linux becomes more mainstream, all distros will be easier to use and greater customer experience will greatly reduce the need for the kind of hand-holding that Red Hat provides.
I think you're confused. Reverse engineering offers no protection against patent violations. Perhaps you are thinking of trade secrets.
The terms of the GPL might prevent Java from being released under the GPL, but obviously the GPL has no power to override patents.
"The issue of delegates I agree with to some degree, as it is nice syntactic sugar, but one that is, again, easily done equivalently well through the use of listener interfaces. Admittedly, it uses more lines of code, but it also guarantees that people actually impliment the interface, as opposed to simply subscribing to events haphazardly and hoping they got the right ones."
It's pretty clear that using interfaces for events is a major kludge. If you have to handle the same event from different objects than you have to have an adapter.
Your comment about "hoping they got the right ones" doesn't make any sense to me. If you don't know what events you want to hook up to you're going to have a problem that neither Java nor C# can prevent.
Sorry, but their are plenty on non-linux POS terminals out there.
Dana Blankenhorn is delusional if he thinks Linux has already won.
I think for your argument to explain why Java apps are slow, the vast majority of Java programmers would have to be unskilled, not just "a number" of them. I think you're grasping at straws to defend Java performance.
In case you somehow have assumed that I'm just a typical consumer, I actually attended the Windows NT launch video conference. So, yes, NT was sold to people very much like ME although I didn't actually program for it until 1996.
As a professional software programmer for over 20 years, I'd advise anyone who wishes to still be programming professionally in their 50's to spend a summer selling used cars. Being able to sell yourself will be much more important in the later years of your career than your technical knowledge will be.
"And the 2600 was definitely an instance of nothing but good timing"
As a former 2600 game programmer I disagree. It was the brilliant low-cost, deeply flexible design of the 2600 that kept it dominant when there were plenty of competitors around.
"You're trying to confuse necessary and sufficient in an attempt to disprove that "being DOS" was necessary."
I'm not confusing anything. You said that Windows was popular because it was an extension of MS-DOS. That's simply not true for the reasons I stated.
"MS-DOS existed under the covers in consumer Windows until it was finally superceded by Windows NT in the consumer space."
"At a certain level, anything before XP was still just MS-DOS."
I think you're a bit confused. Apparently you're having trouble deciding if you think NT was the first Windows not based on MS-DOS or whether it was XP (the correct answer is NT).
"In 1995, Microsoft was still selling MS-DOS to people like YOU."
Oh, the shame. In 1995 was I supposed to buy a Mac, or a Sun workstation? Or was I supposed to be using Linux 1.0 without KDE or gnome or any applications beyond utilities like cat and ls?
I think this Java language vs. Swing is just splitting hairs. People don't choose a language in isolation, the bundled libraries are part of the package. In any case, my point is the same. If it's harder to make Swing applications perform well, than that is a negative aspect of choosing the Java platform.
Yes, that was MS's version of the Java language that could have made Java popular on the Windows desktop if Sun hadn't sued them. Sun decided that pure Java was more important than success on the Windows desktop so I guess they should be happy that things turned out the way they wanted.
These days MS isn't interested in Java and Windows developers are even less interested.