I think there are some problems with GPL Java work, although in practice you get the same thing that happened with KDE - the problems belong to the copyright holders, who aren't about to sue themselves. They could, however, sue you for distributing their own software, which is what used to sour Debian on KDE.
No, don't refrain from using it. If that sort of licensing bothers you, just refrain from putting your own work under the same license. I generally would do BSD-license work if paid to do so, GPL and LGPL otherwise.
Do you really think that the X11 authors, who put that "anything goes" license on their software, would want you not to use it if you don't like "anything goes" on your own code? No, sorry. It happens that Jim Gettys, one of the two creators of X, works on GNOME now.
The TCP/IP in BSD is derived from one written by BBN under an ARPA contract. Of course this was long before MS got involved, and was explicitly written as part of the creation of the Internet.
Many proprietary TCP/IP implementations come from the BSD one, I don't know the provenance of the Microsoft one.
Here is my code, which I offer to share with you and everyone. If you would like to share code with me, share with me and everyone any derivatives that you make of my code. That is sharing like-for-like, and only fair. If you can not share your derivative of my work, I'm sorry, but I am only interested in sharing and do not wish to give you my code with no strings attached. Do not make derivative works of it, but you may still run it, and redistribute it, as it is.
On second reading, perhaps I have mis-understood your argument?
What do you think the difference is between reusable modularized components and objects that are themselves black boxes? Do you say "monolithic" because the internal objects are all stuck together, or because the objects do not expose their own internals?
Actually, people who apply the GPL to their code are not giving it away. They are sharing. And when people share, both of them have to consider the responsibility that is a consequence of sharing.
In contrast, code under the MIT license (which is the BSD license without the advertising clause) is a gift. There is no expectation that the recepient will reciprocate.
DMCA attempts to create a class of programs that can not legally be free software. This seems to be succeeding so far and is incredibly bad news for us.
UCITA does not allow all warranties to be waived when software is sold. Sometimes, free software is sold, and the seller has a right of recourse back to the author if he is sued for damages. This could make it economicaly unfeasable for authors to distribute free software.
These problems are a lot bigger than linking and the ASP problem. I don't have much to say about how those two laws effect those specific problems, sorry.
What if libc is Free Software? I think this might have been the case on the NeXT, but I have no evidence at hand. Certainly we could make GNU LIBC run on Solaris, at which point probably the boundary between free and proprietary would be at the system-call interface.
I think it's a moot point because Sun is perfectly capable of distributing the desktop on its own CD that is a separate package but is generally added to an order.
I think you have it backwards. But you don't have to take my word on that. Ask the folks who make Qt at Troll Tech, who have applied the GPL to their library and continue to charge a royalty for use of the library in proprietary programs. They won't tell you that the GPL is spite against people who try to earn a living from software. They'll tell you it protects their profits!
RMS says that the GPL is a straight copyright permission. I know of one other attorney who disagrees, and points to language in the GPL that says "you indicate your acceptance" to make his point. What are you accepting if it's not a contract?
Most other Free Software licenses do not shy from use of contract law. It may be necessary for the GPL to incorporate more contract law than it does today.
In a time when we should start having modularized interchangable components instead of monolithic black
boxes, the FSF has instead downplayed the use of the LGPL.
I think there's something wrong with your paradigm.
The object-oriented revolution has mostly not happened. Part of the problem is that black boxes don't work for software as well as they work in electronics. One can have interoperable components that are entirely proprietary, anyway, if black box interfaces are all you want. Where Free Software excels is the ability to snip code elsewhere than at the object boundaries. This is not the object-oriented model - we're getting behind the interfaces and messing with actual code:-) . Perhaps that's one reason for our success. And IMO, that's a good reason to do something that encourages people to make more software free, so that you can have more reuse, which is what the GPL does.
Now you see why this is being done very slowly and carefully. RMS wants to close out the loophole to provide freedom. Thus, look to him to find a solution that yields the maximal freedom.
I don't think it would require a click-through license. Most of the Open Source licenses are contracts, not copyright permissions, and do not require a click-through. I think that the act of modifying the program is enough to indicate your acceptance. If you haven't modified it, we wouldn't get much from your distributing it anyway.
What I worry is that if it is possible for individuals to take advantage of our code and not have to share their own
contributions back, it may stifle the free environment we are attempting to establish. I suspect that if sharing was
required, then assuming all commercial participants follow this rule the competition would be fair. However if it
is possible to hold parts of the game system proprietary, then it will be seen (rightly or wrongly) as a competitive
advantage to do so, and thus in spite of all the problems we all KNOW are intrinsic in hiding source code, they
will take this route instead.
You have elucidated the reason that I support the GPL. It provides a fair quid-pro-quo for all developers, commercial or volunteer. That's why it's the most popular Free Software license - the people who actually write the code want the quid-pro-quo.
Does the need for hiding source code for purposes of ensuring mystery pose a
legitimate exception to the free software / open source principles?
I don't think so. If I understand you correctly, the same people who would look at it are those who cheat at solitare. Is it then important to keep these secrets? Or is it "spoilers" you are worried about. We seem to be able to handle spoilers with nettiquette rather than licensing.
Since the GPL appears to allow keeping code secret either by not
distributing the binaries, or by dynamically linking it, does this mean that it is morally okay to do so?
Is it OK with every developer of the program? If so, then yes. Otherwise, no. They created the work, they should collectively get to say how it's used.
The approach we at WorldForge have been toying with is providing for "softcode" additions to the game,
which are stored in a database and kept segregated from the primary game code, rather than encouraging use
of the "ASP loophole" or dynamic linking loophole as workarounds.
I'm still not clear why you'd have to encourage any loopholes. I think, however, that an ASP taking advantage of your code and not returning anything could destroy that quid-pro-quo and thus demotivate your developers.
Yes, I told the Telkel folks that the LGPL would be a better idea for now.
The lack of vision was clear during the "One Click " patent war. This doesn't prevent any software from doing
this, only from other businesses from doing this. No one from free/open software groups said anything about
this. at all. very disturbing.
You may have a misconception about patents. If it's a business-system patent, you have to license it even if you are a non-profit organization or an individual using the principle. I can't think of anyone else who would run a program:-)
2) There are alternatives for many GPL'd software, and I hope that GPL doesn't become more of an affront to
businesses using them.
Take gtk+ for example. Do you think that sun/hp would consider taking gtk+/gnome on, if all the libraries
were GPL'd?
GTK+ was explictly made LGPL so that proprietary programs could use it. Qt is GPL but a commercial license is available. Both seem to be acceptable to business. In the case of Qt, GPL actually makes it possible for Troll Tech to make money from proprietary software. This is hardly anti-business.
Does anyone besides RMS see more LGPL software as a threat?
It's a threat in the case of software patents, as are most licenses other than the GPL. LGPL code can be linked to code that you would not have a right to use without a patent license, and the patent can be used to keep the patented principle from being implemented in LGPL code. In general, software patents are bad for Free Software. Other than that, it's not a threat, but one should consider when to use it. In general, I put my work under the GPL because I want modifications back.
3) "linking" - yes. This needs to be cleared up, but so does distribution. If ASP are considered distributing,
then wouldn't intra-organization? More attacks on business
It was an ASP business that asked me to work on adding ASP-model source distribution to a Free Software license. That would allow them to make their software Free while staying in business. If they can't get the modifications back from their competitors, they won't make it Free at all. This is just one way in which the GPL is pro business.
4) Legality of GPL. No one mentions that GPL doesn't make my changes GPL, it just allows me to continue
using the GPL software (ie: the "viral" nature is in the LICENSE, not in the copyright). You cannot copyright
someone elses works, only refuse usage of the GPL software unless certain conditions are met (ie: patches
under GPL/LGPL too).
I think people do make that point. And it's why I think the GPL doesn't force anyone to free their software. They have to choose to derive from your work, knowing that free software is one of the concequences of that choice.
Is it not possible to write a GPL program with non GPL software (ie:
motif, MFC, etc)?
Yes, you can include non-free components that are normaly distributed with the OS or compiler, as long as the resulting work is not distributed together with the OS and compiler.
I cannot
risk using GPL software that I might have to disclose our changes.
OK. But under those terms, I don't want you as the developer of a derived work of my GPL-ed software. And the license lets me enforce that. You do have other choices, and of course I have the right to say "no".
It would be a mistake for FSF to get into judging the usefulness of your modifications. You won't see that happen.
Someone above mentioned the case where gcc
is modified to output their internal representation of the parsing structure.
The GPL covers this:
the output of the Program is covered only if its contents constitute a work based on the Program (independent of having been made by running the program).
But you don't have a right to posess a copy of the program at all without a license. In order to get that right you accept a license that is a contract that can restrict rights you would otherwise have had.
Licenses are contracts and can restrict rights you would otherwise have. In this case, you would accept certain restrictions on either public performance or use for the right to posess a copy of the program.
We want to be careful about this, because we have cloned a lot of people's proprietary APIs claiming that it was fair use on our part. A DTD is a work, and is digital data, and you can restrict derivative works, but if someone creates a work-alike DTD using none of your bytes, you would probably not be able to restrict it.
Probably the best effort in this direction so far is Sun's Industry Standards Software License, otherwise known as the SISSL, pronounced "sizzle" and not to be confused with the non-Open-Source SCSL. It allows proprietary derivative works as long as you publish an open reference platform that implements your changes. You might consider applying that to your DTDs, but I doubt it will protect you in all cases.
Slashdot Mirror
Thanks
Bruce
I think we will see Java under GPL eventually.
Thanks
Bruce
Do you really think that the X11 authors, who put that "anything goes" license on their software, would want you not to use it if you don't like "anything goes" on your own code? No, sorry. It happens that Jim Gettys, one of the two creators of X, works on GNOME now.
Thanks
Bruce
Thanks
Bruce
Many proprietary TCP/IP implementations come from the BSD one, I don't know the provenance of the Microsoft one.
Thanks
Bruce
Here is my code, which I offer to share with you and everyone. If you would like to share code with me, share with me and everyone any derivatives that you make of my code. That is sharing like-for-like, and only fair. If you can not share your derivative of my work, I'm sorry, but I am only interested in sharing and do not wish to give you my code with no strings attached. Do not make derivative works of it, but you may still run it, and redistribute it, as it is.
Thanks
Bruce
I'd like to see logical argument on this point.
Thanks
Bruce
What do you think the difference is between reusable modularized components and objects that are themselves black boxes? Do you say "monolithic" because the internal objects are all stuck together, or because the objects do not expose their own internals?
Thanks
Bruce
In contrast, code under the MIT license (which is the BSD license without the advertising clause) is a gift. There is no expectation that the recepient will reciprocate.
Thanks
Bruce
UCITA does not allow all warranties to be waived when software is sold. Sometimes, free software is sold, and the seller has a right of recourse back to the author if he is sued for damages. This could make it economicaly unfeasable for authors to distribute free software.
These problems are a lot bigger than linking and the ASP problem. I don't have much to say about how those two laws effect those specific problems, sorry.
Thanks
Bruce
Thanks
Bruce
I think it's a moot point because Sun is perfectly capable of distributing the desktop on its own CD that is a separate package but is generally added to an order.
Thanks
Bruce
Bruce
Most other Free Software licenses do not shy from use of contract law. It may be necessary for the GPL to incorporate more contract law than it does today.
Bruce
I think there's something wrong with your paradigm.
The object-oriented revolution has mostly not happened. Part of the problem is that black boxes don't work for software as well as they work in electronics. One can have interoperable components that are entirely proprietary, anyway, if black box interfaces are all you want. Where Free Software excels is the ability to snip code elsewhere than at the object boundaries. This is not the object-oriented model - we're getting behind the interfaces and messing with actual code :-) . Perhaps that's one reason for our success. And IMO, that's a good reason to do something that encourages people to make more software free, so that you can have more reuse, which is what the GPL does.
By the way, yes, I'm an O-O programmer by choice.
Thanks
Bruce
Thanks
Bruce
Thanks
Bruce
Thanks
Bruce
You have elucidated the reason that I support the GPL. It provides a fair quid-pro-quo for all developers, commercial or volunteer. That's why it's the most popular Free Software license - the people who actually write the code want the quid-pro-quo.
Does the need for hiding source code for purposes of ensuring mystery pose a legitimate exception to the free software / open source principles?
I don't think so. If I understand you correctly, the same people who would look at it are those who cheat at solitare. Is it then important to keep these secrets? Or is it "spoilers" you are worried about. We seem to be able to handle spoilers with nettiquette rather than licensing.
Since the GPL appears to allow keeping code secret either by not distributing the binaries, or by dynamically linking it, does this mean that it is morally okay to do so?
Is it OK with every developer of the program? If so, then yes. Otherwise, no. They created the work, they should collectively get to say how it's used.
The approach we at WorldForge have been toying with is providing for "softcode" additions to the game, which are stored in a database and kept segregated from the primary game code, rather than encouraging use of the "ASP loophole" or dynamic linking loophole as workarounds.
I'm still not clear why you'd have to encourage any loopholes. I think, however, that an ASP taking advantage of your code and not returning anything could destroy that quid-pro-quo and thus demotivate your developers.
Thanks
Bruce
The lack of vision was clear during the "One Click " patent war. This doesn't prevent any software from doing this, only from other businesses from doing this. No one from free/open software groups said anything about this. at all. very disturbing.
You may have a misconception about patents. If it's a business-system patent, you have to license it even if you are a non-profit organization or an individual using the principle. I can't think of anyone else who would run a program :-)
2) There are alternatives for many GPL'd software, and I hope that GPL doesn't become more of an affront to businesses using them. Take gtk+ for example. Do you think that sun/hp would consider taking gtk+/gnome on, if all the libraries were GPL'd?
GTK+ was explictly made LGPL so that proprietary programs could use it. Qt is GPL but a commercial license is available. Both seem to be acceptable to business. In the case of Qt, GPL actually makes it possible for Troll Tech to make money from proprietary software. This is hardly anti-business.
Does anyone besides RMS see more LGPL software as a threat?
It's a threat in the case of software patents, as are most licenses other than the GPL. LGPL code can be linked to code that you would not have a right to use without a patent license, and the patent can be used to keep the patented principle from being implemented in LGPL code. In general, software patents are bad for Free Software. Other than that, it's not a threat, but one should consider when to use it. In general, I put my work under the GPL because I want modifications back.
3) "linking" - yes. This needs to be cleared up, but so does distribution. If ASP are considered distributing, then wouldn't intra-organization? More attacks on business
It was an ASP business that asked me to work on adding ASP-model source distribution to a Free Software license. That would allow them to make their software Free while staying in business. If they can't get the modifications back from their competitors, they won't make it Free at all. This is just one way in which the GPL is pro business.
4) Legality of GPL. No one mentions that GPL doesn't make my changes GPL, it just allows me to continue using the GPL software (ie: the "viral" nature is in the LICENSE, not in the copyright). You cannot copyright someone elses works, only refuse usage of the GPL software unless certain conditions are met (ie: patches under GPL/LGPL too).
I think people do make that point. And it's why I think the GPL doesn't force anyone to free their software. They have to choose to derive from your work, knowing that free software is one of the concequences of that choice.
Is it not possible to write a GPL program with non GPL software (ie: motif, MFC, etc)?
Yes, you can include non-free components that are normaly distributed with the OS or compiler, as long as the resulting work is not distributed together with the OS and compiler.
I cannot risk using GPL software that I might have to disclose our changes.
OK. But under those terms, I don't want you as the developer of a derived work of my GPL-ed software. And the license lets me enforce that. You do have other choices, and of course I have the right to say "no".
Thanks
Bruce
Someone above mentioned the case where gcc is modified to output their internal representation of the parsing structure.
The GPL covers this:
the output of the Program is covered only if its contents constitute a work based on the Program (independent of having been made by running the program).
Thanks
Bruce
Thanks
Bruce
Thanks
Bruce
Bruce
Probably the best effort in this direction so far is Sun's Industry Standards Software License, otherwise known as the SISSL, pronounced "sizzle" and not to be confused with the non-Open-Source SCSL. It allows proprietary derivative works as long as you publish an open reference platform that implements your changes. You might consider applying that to your DTDs, but I doubt it will protect you in all cases.
Thanks
Bruce