On Mon, Dec 19, 2005 at 08:32:18AM -0500, Olin Lathrop wrote: > Peter wrote: > > Malicious is any pigopolist who takes PD code, changes the attributions > > and deletes the copyrights, burns it into a Flash and ships 100,000 > > routers to all over the world based on that. > > Changing attributions and deleting copyrights is just plain wrong or > illegal. Otherwise, what's wrong with this? In other words, if someone > took public domain code, kept references to the author in the source and > shipped 100,000 copies as binary (assuming the copyright, if any, allowed > this), this is not malicious but a smart way to make a product. Without a doubt. That's why the GPL specifically prohibits this activity without the release (or the offer to release) the source along with it. > I really object to you calling this "malicious" or "stealing". It's malicious or stealing when it's done to software whose license prohibits this. As you state... > It's only that if the copyright doesn't allow it. Right. > I'm not suggesting people ignore > the GPL copyright, only pointing out that the GPL is not a good license if > your aim is to have more and better software out there (as the proponents of > the GPL usually claim). I agree with out on more software. I think that better is at least debatable. The premise is that since software development and updates are no longer a single source process, that end users and other developers have the ability to get the software that suits their needs. I'm not sure that I'm buying that argument. But I do think it has some logic to it. > > This does not mean that you relinquish your rights on the code, you > > still own it, > > It's getting really confusing who is "you". The author of the code. > > > This prevents you from stealing code from others, and others from > > stealing your code. And that's the point. > > Let's assume that everyone is following the law. There is then no stealing. > The question is what should others be *allowed* to do with the code. That's the whole point of licensing. > Labeling a certain type of use as "stealing" or "malicious" is just rude and > inflammatory. I believe that the label was applied to those who was not following the law. Therefore your statement is invalid if everyone is following the law. You'll get this discussion from GPL advocates specifically because there are so many developers trying to figure out ways to circumvent the license. So let's table that and move back to your assumption. Then the question becomes: What are the pros and cons to both developer/resellers and end users to have source code available? I'll be the first to admit that there are pros and cons on both sides. Most Open Source folks cannot see the downsides. > > > You have to understand that people who have made their code available > > under the GPL did not relinquish any rights on it. They still own it. > > They did not give up *anything* by doing this. Not true. They have given up single source rights. Once GPLed anyone can use, modify, and redistribute the code. Every copy and every modification dilutes the value of the code to the original author. That's one of the cons to the question above. Now this may not be too damaging to the author. Maybe they don't have the ability to fill a distribution channel. Or as Olin has pointed out, maybe they do not have a profit motive. But they definitely give up a lot by GPLing their code. > > In a way, putting code > > under the GPL is like a patent disclosure. It's got full blueprints, > > and anyone can see them and use them, but they cannot build something > > based on it except by also publishing the full blueprint. That > > implicitly exposes them to being sued by you should they try to make it > > proprietary thereafter, and gives you the right to sell it for as much > > money as they would (suppose). By the nature of a free market, if there are lots of potential vendors, then the price of the goods drops. A single source software provider can sell their wares for more than one of a multitude of vendors. The GPL ensures that there can be a multitude of vendors. > Yes, I think we all know this. Again you are completely missing my point. > > It is these incentives which cause the pigopolists > > to gnash their teeth at GPL. It's a device that discourages illicit use > > of intellectual property. > > No it doesn't. Anyone willing to use intellectual property illicitly is > ignoring the license by definition, and therefore it's probably irrelevant > what it says. Agreed. It's a device the protects the original author from being taken when illicit activity happens. It doesn't prevent the illicit activity from happening. > The GPL prohibits a certain type of use. All I'm saying is that this > prohibition is not beneficial to some of the stated aims of the proponents > of the GPL. Some of the proponents. > > Imho, you really do not understand what 'derivative works' means. You > > have a program that consists of code you contribute (A) and a body of > > code you extend or link against (B), which is GPLd. A+B is a derivative > > work. A is released by you (suppose) under GPL (you have to, by tying > > it with B, which is GPLd). I think we're all clear on that. > > Then someone comes and writes C and makes the work being A+B+C. He > > cannot but publish the source (i.e. including A+B+C) or he is in breach > > of GPL. Just to be picky. The writer of C is only obligated to give the source of A+B+C to anyone who receives the work A+B+C. If they give no one the work, then they are not obligated to give it to anyone. Also they do not need to make the source public, only give a copy to whomever receives A+B+C. Finally they cannot prevent the end user receiving A+B+C from publicly publishing or redistributing A+B+C (source or executable) if that receiver so desires. > > At the same time you can find someone who wants to use your A > > in a product. You then use A or an improved/adapted A1 and sell/license > > it to him, by granting him a non-exclusive license on A or A1. You do > > *not* need to publish A1 for this. That is correct for the original author of A. However none of A, B, or C can do this with a combined work unless all authors contributing agree to relicense. So if B adds something really good to A, A cannot proprietize it without working out something with B. > > The only time when you must publish > > A1 is, if, and only if, you have to use B and/or C together with A1 to > > make it work for your client. Bingo. > > Then it is up to the client to find the > > developers of B and C and ask for licenses, so he can make the product > > using A+B+C and not need to publish the source (after aquiring licenses A1+B+C > > from you and the developers of B and C). But you own A and A1 all the > > time, and you can even revoke the GPL on A if you are really mad at the > > world. Yes and No. Anyone who received A under the GPL retains their GPL rights. Anyone who comes after may not be able to get it from the author of A. However, since GPLed copies of A exist and those license holders have the right to redistribute, most likely someone can ask one of those license holders for a copy. > That's a good explanation of how the GPL works and its infectious > propagation (I agree that "viral" is a bad term because it also implies > direct harm to the host. "Infectious" is probably not much better in > connotation, but at least more technically accurate.). > > The problem with the GPL is that in this case I make A+B and have to publish > the source to A, as you described above. This is a serious disinsentive to > use B in the first place. Publishing A has a certain cost, which makes the > GPLed software more costly to use than more openly released software. Olin, I need some clarification of the above. A is GPL. You create and add B to A producing A+B. Am I on point? If so then your obligation is to give the source to B to anyone that you give A+B. So do you mean that publishing B has a certain cost? That's certainly true. But Gerhard's point is still made. In this instance you contact the author of A and ask for a separate license for A. The typical battle line is the developer who wants to get A for free and make A+B propietary. > > Again, I have no problem with people putting whatever restrictions on their > code they want. After all, I'm not forced to use it. What bothers me is > people using rather restrictive licenses like the GPL, then claiming they > are doing it for the common good and to promote more and better software. More I don't agree with. I think the jury is still out whether as a class that Open Source software is a better process. BAJ -- http://www.piclist.com PIC/SX FAQ & list archive View/change your membership options at http://mailman.mit.edu/mailman/listinfo/piclist