Re: [License-discuss] System 76's BeanBooks Public License v1.0
On Wed, 18 Sep 2013, John Cowan wrote: Sec. 4.3 strikes me as actually conceptually somewhat interesting, inasmuch as many commercial lawyers have argued that this type of clause is often implicit in software that contains a protect trademark embedded in the software and not removed by a downstream licensee. In this case, however, Section 4.2 prevents you from removing the protected trademark. Taking the two clauses together, you are effectively prevented from making commercial use of the software without paying for the trademark license, which obviously contravenes clauses 6 and 7 of the OSD. So this license is on its face not Open Source. It says: 4.3 - Commercial distribution of the Software requires a trademark license agreement and you may be required to pay. Using the Software within a corporation or entity is not considered commercial distribution. This license does not grant You rights to use any party's name, logo, or trademarks, except solely as necessary to comply with Section 4.2. Wouldn't the except solely as necessary to comply with section 4.2 clause make it okay? Section 4.2 prevents you from removing the protected trademark, and section 4.3 allows you to use the trademark under those circumstances. ___ License-discuss mailing list License-discuss@opensource.org http://projects.opensource.org/cgi-bin/mailman/listinfo/license-discuss
Re: [License-discuss] License which requires watermarking? (Attribution Provision)
On Wed, 2 Jan 2013, Lawrence Rosen wrote: Regardless of whether a library is licensed under the GPL or the LGPL, a licensee will have to disclose *source code* of the library and *source code* of derivative works of the library. If you agree with the FSF's position on what a derivative work is, a work that links to a LGPL library is a derivative work but you are not required to release source for it. ___ License-discuss mailing list License-discuss@opensource.org http://projects.opensource.org/cgi-bin/mailman/listinfo/license-discuss
Re: [License-discuss] License which requires watermarking? (Attribution Provision)
On Tue, 1 Jan 2013, Lawrence Rosen wrote: Some people use ordinary GPL on libraries with the intent of crippling competing commercial reuse (since any competitors have to release their source and competitors wouldn't want to do that). Really? That's not wise. How would the choice of license affect the *legal* determination of whether the resulting work is or is not a derivative work for which source code must be disclosed? The choice of license affects whether source code must be disclosed at all. If the library was under the LGPL, the competitor would not have to provide source. ___ License-discuss mailing list License-discuss@opensource.org http://projects.opensource.org/cgi-bin/mailman/listinfo/license-discuss
Re: [License-discuss] Permissive but anti-patent license
On Wed, 2 Jan 2013, John Funnell wrote: I do believe a clause can apply retroactively. If not, I could download the Linux kernel with good intentions, wait a week and then turn evil and, with that copy, violate GPL exactly as I pleased. The requirements of the GPL apply when making copies. If you turned evil a week later, that wouldn't matter, since the GPL wouldn't actually require that you do anything, as long as you don't start making copies after you turn evil. It wouldn't apply retroactively--the copies that you already made are still good. ___ License-discuss mailing list License-discuss@opensource.org http://projects.opensource.org/cgi-bin/mailman/listinfo/license-discuss
Re: [License-discuss] License which requires watermarking? (Attribution Provision)
On Mon, 31 Dec 2012, Rick Moen wrote: I conclude that, in general, the overwhelming majority of such entrepreneurs are thus seeking the crippling of competing commercial reuse -- not just attribution. So, OSI should give them the bum's rush. Some people use ordinary GPL on libraries with the intent of crippling competing commercial reuse (since any competitors have to release their source and competitors wouldn't want to do that). Is the GPL also considered unfree when applied to libraries? ___ License-discuss mailing list License-discuss@opensource.org http://projects.opensource.org/cgi-bin/mailman/listinfo/license-discuss
Re: [License-discuss] Permissive but anti-patent license
On Sun, 30 Dec 2012, John Funnell wrote: When a user makes their first copy, their acceptance of the anti-patent license is also an agreement to the restrictions on usage. If a US person violates this by entering into an agreement for a limited-scope patent license, they violate their original copyright license and are thus not allowed to copy, own or use the software. But can it retroactively invalidate an existing copy? If I write a license saying someone can only copy the software if they are a vegetarian, and they copy the software and start eating meat later on, they still were a vegetarian at the moment they copied the software. You won't be able to use the copy restriction like a use restriction because the copy restriction only applies to conditions that are true when the copy was made. ___ License-discuss mailing list License-discuss@opensource.org http://projects.opensource.org/cgi-bin/mailman/listinfo/license-discuss
Re: OSL 2.0 and linking of libraries
On Fri, 2 Apr 2004, Peter Prohaska wrote: On 1) This way, it doesn't matter what derived work is anymore because we just define it. That should reduce the FAQ size. But isn't derived work a legal term? Open source licenses can put limits on derived works since if you reject the license you have no rights to make derived works at all. If you define derived work to be something broader than what the law says, this is no longer true and the license doesn't function as intended. -- license-discuss archive is at http://crynwr.com/cgi-bin/ezmlm-cgi?3
Re: Silly question: are usage restrictions covered by the OSD?
On Thu, 16 Oct 2003, Chris F Clark wrote: Still, I am interested in other peoples impressions of this argument. The reason being, I am considering drafting a license which makes approximately that distinction. It is a license that is viral like the GPL except that it defines its point of requiring open sourcing of the resulting works the point of derivation rather than the point of redistribution. That is, one must release an open source copy of the derived work when one creates such a derived work, not only when one distributes such a derived work. (There are many details to work out, which is why I have not submitted it for review.) So when someone's editing a file each time they write out the file they have to send the result to other people? Or does this have to be done each time the person inserts a character, even if he doesn't write it out? -- license-discuss archive is at http://crynwr.com/cgi-bin/ezmlm-cgi?3
Re: Berkeley DB License
On Fri, 1 Nov 2002, Bennett Todd wrote: Does the new frontmatter at URL:http://www.sleepycat.com/licensing.html where they define redistribute to include use in multiple different physical locations, even within the same organization affect their license's Open Source Definition compliance? I don't know about this specific case, but in general, it obviously has to. What if someone had a license which defined redistribute to be use without paying the author a thousand dollars? -- license-discuss archive is at http://crynwr.com/cgi-bin/ezmlm-cgi?3
Re: OSL Version 1.0 dated 8/2/2002
On Fri, 2 Aug 2002, Lawrence E. Rosen wrote: I decided to borrow some of their wording and added an external deployment provision to the OSL. Here is the clause I added: ... As an express condition for the grants of license hereunder, You agree that any External Deployment by You shall be deemed a distribution and shall be licensed to all under the terms of this License, as prescribed in section 1(c) herein. Someone could refuse to accept your license (thus also refusing to accept your definition of distribution). If he refuses to accept your license, he could then Externally Deploy the code without being bound by the license. This provision accomplishes little. The GPL and similar licenses can put conditions on distribution because if you don't accept the license you have no right to distribution at all. Redefining distribution doesn't work--someone who doesn't accept the license only lacks the right to distribution as originally defined by the law, not as redefined in the license. -- license-discuss archive is at http://crynwr.com/cgi-bin/ezmlm-cgi?3
RE: OSL Version 1.0 dated 8/2/2002
On Fri, 2 Aug 2002, Lawrence E. Rosen wrote: Someone could refuse to accept your license (thus also refusing to accept your definition of distribution). If he refuses to accept your license, he could then Externally Deploy the code without being bound by the license. This provision accomplishes little. Are you suggesting that the following provision in the OSL won't work to prevent the copying that is essential for use of a software program? 8. Acceptance and Termination. Nothing else but this License (or another written agreement between Licensor and You) grants You permission to create Derivative Works based upon the Original Work, and any attempt to do so except under the terms of this License (or another written agreement between Licensor and You) is expressly prohibited by U.S. copyright law, the equivalent laws of other countries, and by international treaty. Therefore, by exercising any of the rights granted to You in Section 1 herein, You indicate Your acceptance of this License and all of its terms and conditions. This license shall terminate immediately upon Your failure to honor the proviso in Section 1(c) herein. Yes, I'm suggesting exactly that. Copying of the program into memory in order to run it is permitted by law. The user may therefore do so without accepting the license. The clause above says you can't make derivative works without accepting the license, not that you can't make copies. (And even if you argue that copies are derivative works--IANAL, but I don't know if you could get away with that--that would simply mean that that clause is wrong. Section 117 of the US copyright law lets you copy the program into memory in order to use it. If you're going to consider that a derivative work, then yes, something else *does* let you make derivative works.) -- license-discuss archive is at http://crynwr.com/cgi-bin/ezmlm-cgi?3
RE: OSL Version 1.0 dated 8/2/2002
On Fri, 2 Aug 2002, Lawrence E. Rosen wrote: Section 117 of the Copyright Act allows the copying of a computer program as an essential step in the utilization of the computer program. That doesn't mean that one cannot contract that right away. I would not suggest that any open source license do so. But an open source license can -- unless I'm misreading the OSD -- provide that anyone who makes such copies and distributes them provide access to the source code. And such a contract is NOT against the Copyright Act. Yes... but nobody has to accept the contract, either. There's a clause in the GPL which may sound like that, but it's poorly worded. The GPL says that by distributing the program, you've accepted the license. What that *really* means is something like by distributing the program, you've either accepted the license, or broken copyright law. It isn't a generalized by doing X, you've accepted the license where X can be anything at all. X can only be a limited range of things--it has to be something that's against the law if done with no license. Copying the program into memory and 'distributing' the program over the web aren't among such things. As I suggested earlier, I honestly don't much care whether someone makes a copy of an open source program. What many open source software owners care about is when they create derivative works, use those derivative works to make money by selling the *use* of the program (perhaps by providing an Internet service), and then not return the source code so that everyone else can benefit from those modifications. Legally, I can prevent that result through contract, and section 117 of the Copyright Act doesn't prevent me from doing so. If you try that, people will just refuse to accept the contract and do it anyway. Since those activities are (unlike similar GPL-violating activities) legal without a contract, that will work for them. -- license-discuss archive is at http://crynwr.com/cgi-bin/ezmlm-cgi?3
Re: discuss: WGPL (WebGPL)
Does this license make it illegal to use an ad-filtering proxy on the page without accepting the license? After all, using an ad-filtering proxy copies and modifies the page, and it's not clear that this is 'running the Web'. What about putting the page on a site like Geocities which automatically modifies the code? Geocities ad-popup code is not GPL, after all. What exactly is a web page? More specifically, are framed content, inlined images, etc. considered part of the web page? -- license-discuss archive is at http://crynwr.com/cgi-bin/ezmlm-cgi?3
RE: OSD modification regarding what license can require of user
On 18 Mar 2002, Ean Schuessler wrote: What if you simply added a requirement that: http://[service host name]:80/gnu-sources Must always either supply the sources or a redirect to the sources? This rule could even apply for internal distribution (ie. services only available to AOL users). That would seem to take care of the problem without placing potentially unfulfillable constraints on the user interface. It assumes http. Imagine if someone had put a requirement in ten years ago which mandated that the sources be available by a Gopher link. -- license-discuss archive is at http://crynwr.com/cgi-bin/ezmlm-cgi?3
RE: Static v. Dynamic Linking -- redux
(Huge To: line deleted. Feel free to repost if you want.) On Fri, 15 Mar 2002, Lawrence E. Rosen wrote: That said, and after some reflection, I would now argue that the trivial conversion of a static link into a dynamic link or API interface, simply to get around the provisions of the GPL, won't work. The court would hear from a parade of experts who would explain that the intent was clearly to circumvent the license. Why should that matter? Here, intent to circumvent the license just means intent to follow the letter of the license while not following it's spirit. Why would such an activity be prohibited? -- license-discuss archive is at http://crynwr.com/cgi-bin/ezmlm-cgi?3
Re: Open source shareware?
On Thu, 8 Nov 2001, Karsten M. Self wrote: Clause 1: The license shall not require a royalty or other fee for such sale The terms for payment are interpreted by me to be sale +time, which in the general case reduces to a required fee for sale or transfer. Even in the general case, where the time period can be 0, it's not a required fee for sale or transfer. You can sell or transfer it to anyone you want without a fee--the person you transfer it to just can't use it. -- license-discuss archive is at http://crynwr.com/cgi-bin/ezmlm-cgi?3
Re: Intel's proposed BSD + Patent License
Intel hereby grants Recipient and Licensees a non-exclusive, worldwide, royalty-free patent license under Licensed Patents to make, use, sell, offer to sell, import and otherwise transfer the Software, if any, in source code and object code form. This license shall include changes to the Software that are error corrections or other minor changes to the Software that do not add functionality or features when the Software is incorporated in any version of a operating system that has been distributed under the GNU General Public License 2.0 or later. This is odd. What is the software when linked into an operating system? Does the software include anything linked with it, and therefore, would a change to the operating system that adds features be considered a change to the software? If yes, then the software becomes pretty much useless--you can link it into an operating system only if you don't change the operating system. If no, than anyone can change the software any way they want anyway, merely by claiming the changes are changes to the operating system. Also, wouldn't it violate the fields of endeavor clause and specific to a product clause to only allow the software to be linked with operating systems? (I can see a way around this: claim that of course the user is allowed to link it with anything, it's just that it would violate patents, but it's not the license which says you're prohibited from violating patents, just an external law. The problems with this workaround are left as an exercise to the reader.) -- license-discuss archive is at http://crynwr.com/cgi-bin/ezmlm-cgi?3
RE: Is inherited class a derivative work?
On Sun, 21 Oct 2001, Michael Beck wrote: When you derive a class, you are creating a copy of the original class. When you make changes to the new class, you are creating a derivative work, the same way as you would do it by making changes to a copy of book, copy of a picture, copy of a house design, or a copy of a chip design. You don't change the original, but you still are creating a derivative work. When you derive a class, you're creating a copy of the original class *on your machine*. That doesn't mean that if you write code that derives a class, and distribute the code, you're distributing copies of the original, modified or not. It means that you're distributing instructions which tell other people how to modify copies they already have. You can't copy a house design, change it, and distribute it. But you *can* distribute a set of instructions which says take this house design, reduce the size of room 3A by 5 feet in the east-west direction, and change the pipes to... Someone can get their own copy of the house design and apply those instructions. Telling someone how to make a derivative work isn't the same as distributing one. -- license-discuss archive is at http://crynwr.com/cgi-bin/ezmlm-cgi?3
RE: Is inherited class a derivative work?
On Tue, 16 Oct 2001, Lawrence E. Rosen wrote: While the FSF *may* be correct, I would expect a more thorough analysis of the situation from them before I accept their conclusion. In particular, how does inheritance differ in a substantive and legally significant way from traditional subroutine linkage which, as many of us believe, does *not* create a derivative work at least the context of dynamic linking? Well, the FSF believes that that does too, so I presume they don't see a difference... -- license-discuss archive is at http://crynwr.com/cgi-bin/ezmlm-cgi?3
Re: licenses for RPGs
On Wed, 21 Mar 2001, Ryan S. Dancey wrote: The OGL framework, when applied to the System Reference Document, provides a way to make DD compatible content that is far, far more extensive than the basic rights you might have as they relate to the public domain status of the game rules of DD. And there's no grey area. Both conditions which make it possible to bring to market a commercial product without having to provide for a substantial threat of litigation. That's where the extortion comes in. "If you submit to our conditions, you can do something which is legal anyway, but which we'll otherwise sue you into bankruptcy for". And it's furthermore quite silly to point at the former TSR (now Wizards of the Coast) business and say that the climate of litigation is fostered by one company. Every commerical hobby game publisher has taken the exact same position for 25 years - that the mere game rule content in an RPG is the least part of the copyrighted work of an RPG, and that derivative works based on such a product are infringing. Ryan, the problem is that TSR has bizarre ideas of what constitutes a derivative work, as well as bizarre ideas that the company owns all derivative works, which together amounts to a threat, not only to sue, but to take other people's works without compensation.
Re: licenses for RPGs
The Open Gaming License isn't quite what it seems. The difference between that license and the GPL is that without the GPL, you can't distribute copies at all, and the GPL gives you the right to distribute copies under some conditions--that is, it adds rights. The Open Gaming license is closer to extortion. The key is that game rules are not copyrightable, and normally you'd have complete rights to create D20-compatible material. However, TSR threatens to sue people who do that. Since nobody can afford the cost of a lawsuit from TSR, TSR can prevent perfectly legal activity this way. What the Open Gaming License says "if you adhere to these conditions, we promise not to sue you for something that you really have the right to do anyway". Without the license you have *more* rights--just less freedom from being bankrupted by frivolous lawsuits.
Re: What is Copyleft?
On Fri, 23 Feb 2001, Rod Dixon, J.D., LL.M. wrote: Interesting point. In the ordinary course of programming, I suspect there would be no derivative work created, hence the GPL should provide no obstacle for distributing the program as open source. As you mentioned, there could be a counter-example such as the development of a text editor for Windows that not only makes calls to print drivers that MS has thrown in its OS, but also to DLLs written by printer manufacturers. Add to that fact that the windows text editor reads into those drivers additional True Type fonts (independently copyrightable) and you my have a genuine derivative work question. Something similar first turned up long ago when GNU software first was ported to PCs. Even regular compilers link programs to libraries that come with the compiler, and there was concern that the GPL did not allow this since those libraries didn't come with the OS, but with the compiler. The "solution" was for RMS to state that standard compiler libraries count as operating system components even if they don't really come with the operating system. I still find this unconvincing; but I would expect that a similar rationalization would work with printer drivers. (One of the biggest reasons this is unconvincing is that it doesn't bind anyone else. RMS can only redefine OS components on GPL software that is his, not on GPL software from other people.)
Re: Open Source *Game* Programming License
On Thu, 18 Jan 2001, Bryan George wrote: Oh, you could make a new world file, but by that reasoning, couldn't I make a program that needs readline, distribute it without readline, and tell the user to make a new readline function? World files are at least as important to a game like Quake as readline is to a program that uses it. Yes - likewise, a Website is useless without its content, a CD is useless without songs, a cell phone is useless without a service contract. Shall I go on? Service contracts aren't code, or even copyrighted material. As for the other two examples... Web servers and CD players are *typically* used by separately getting the server/player, and the content. Saying "we won't supply any web pages, get them yourself" is a reasonable thing to do because the web server isn't tied to any particular page; users can and do use many different web pages with it. I can go get a different CD for a GPL'ed CD player. I can't get a different set of Quake data for use with my GPL'ed Quake engine (conversions typically require owning the original game). Compare this to the readline situation: as long as the readline library is the only library available for use with the program, the GPL on one requires that the GPL be on the other one too. (Disclaimer: I personally reject the readline reasoning, but my rationale for that rejection does not apply to Quake.)
Re: Open Source *Game* Programming?
On Wed, 17 Jan 2001, Ben Tilly wrote: IANAL but I think the general reaction would be that the graphics are part of the overall work and said game company would then be obliged to also give away the graphics, which you would then have access to. Doom and Quake have been released as GPL. Graphics and data files have not. By this reasoning, anyone except the creators of Doom or Quake could not distribute those programs at all, since the graphics are necessary to use the program and they are not open source.
Re: LGPL clarification
On 1 Nov 2000, Ian Lance Taylor wrote: The LGPL puts restrictions on P when it is linked with L. But so what? That linking will only happen on the end user system. The typical effect is that the end user is not permitted to distribute the executable now found in memory, because it is impossible to satisfy both the conditions of the vendor of P and the conditions of the LGPL. But the LGPL puts no restrictions on the distribution of P, which is what the proprietary user cares about. That is not, however, what RMS believes. If there is only one shared library that exists, he considers P to be derivative of it even before it is linked; and this triggers all licensing conditions on L even if P is not distributed with L. Remember readline?
Re: Plan 9 license
On Sun, 3 Sep 2000, Mark Wells wrote: Here's a simple test to determine if something has been stolen: does the original owner still have it? Doesn't work. "Because my work is copied and the coies are widely spread, I do not have the potential market that I did before. That market has been stolen from me."
Re: Using GPL'd software in BSD-licensed app
On Wed, 10 May 2000, David Johnson wrote: If the GPLd code is compiled along with the rest of your stuff, then the rest of your stuff needs to be GPL as well. If it can be compiled into a separate library, then it could be used *provided* that there is a functionally equivalent alternative available elsewhere in existance. It should be noted that this claim is highly controversial. I, for one, don't buy it. It would mean that, for instance, if you make a program that only runs under Microsoft Windows and dynamically links Windows libraries, it is a derivative of Windows and Microsoft can legally prevent you from selling it. (I wrote to RMS. He confirmed this, though he did suggest that Microsoft probably doesn't do it because they want people to buy Windows.)
Re: Wired Article on the GPL
On Thu, 30 Mar 2000, Chip Salzenberg wrote: A license that isn't a contract (a bare permission) can be freely revoked by the licensor, as in an invitation to enter onto land: if the landowner changes his mind, the licensee instantly becomes a trespasser. I never thought I'd say this, but: 'Only UCITA can save us now.' IANAL, but there's always the possibility of a court ruling that recipients of GPL software are providing consideration, so it's a contract. If you distribute a binary, the consideration might be that you distribute the source. Putting one's own code under GPL if linked with GPL might also be consideration--I don't think consideration has to directly benefit the original copyright owner to count as consideration. The FSF really should get involved in this, because the implications of this question are obviously pretty big.
Re: How To Break The GPL
On Sat, 4 Mar 2000, David Johnson wrote: But what does "direct functionality" mean in terms of licensing? I can see functionality being added to a GPL application in ways that both would and would not violate the GPL. If I wrote a new plugin for Gimp, it would add functionality, but would only have runtime linkage. But putting the exact some code within the body of the Gimp source code cause it to come under the purview of the GPL. According to RMS, plugins are *also* derivative works, so both your examples would come under the GPL. (Which produces the odd result that it is legal to write a GPL plugin for Internet Explorer but not for Netscape 4, since Internet Explorer comes under the system component exception.)
Re: How To Break The GPL
This basically sounds like "user does the link". The FSF takes the position that if you distribute software that can only be run by linking it with something GPLed, your software is a derivative work of the GPLed software even if you don't include any parts of it. So by these standards, Alice would be distributing a derivative work, so she would be the one violating the GPL. Of course this position has some other unpleasant consequences too; for instance, a program designed only to run with Microsoft Windows DLLs is a derivative work of Windows, which means that Microsoft can deny you the legal right to write Windows programs. I wrote to RMS asking him about just this scenario, and his reply was basically that this is correct, but that it is not in the interests of the makers of proprietary operating systems to do that. (I didn't buy that--it wouldn't be in Microsoft's interest to ban *all* Windows software, but it would be in their interest to ban, say, Word Perfect or Netscape.)
Re: How To Break The GPL
On Fri, 3 Mar 2000, John Cowan wrote: The FSF takes the position that if you distribute software that can only be run by linking it with something GPLed, your software is a derivative work of the GPLed software even if you don't include any parts of it. What if there were a non-GPL and a GPL implementation written to the same interface definition? The FSF's position, as far as I know, is that if there exist two implementations, and one of them is not GPL, then the work can't be said to be derivative of the GPL version. It is only derivative if the GPL one is the only possible one to link with. GNU readline is probably the most infamous case of the FSF taking this position.
Re: Apple Public Source License - is it OSI certified?
Personally I object to the APSL because of Apple's right to withdraw it if there is a claim of infringement. The problem is that this means that the janitor's mother-in-law could say "I think the whole thing infringes" and the entire source code would be withdrawn; the license doesn't require that the claim be one which is made in court. (And the clause requiring Apple to un-withdraw the rights if they win in court doesn't apply if it never _goes_ to court.) But I'm a layman with no legal training whatsoever.
Re: Can Java code EVER be GPLd, at all?
On Sat, 13 Nov 1999, Justin Wells wrote: But new technologies like perl, Java, CORBA, etc., are changing this and I would like to think about what kind of license could be used that would be fair and would still impose a "copyleft" of some sort on these new technologies in order to promote the development of future free software (let's leave the philosophical debate about whether I would succeed, whether this is right, etc., for another day--let's just talk about the technical and OSD compliance issues today.) I had an idea a while back. I make no claim that doing this is a good idea even if it works (it was originally a reductio ad absurdum, not a serious idea), and anyway I'm not a lawyer. The GPL only applies when you distribute copies of a GPL program or derived works of it. It operates by the fact that under copyright law, the default is not to allow people to copy at all, so that if someone doesn't agree to the GPL, they can't distribute. Modifying the GPL to restrict use instead of distribution would not work, because the default is that you _can_ use a program that you have a legitimate copy of, so someone who doesn't agree to such a modified GPL would still be able to use it. My idea was that if you want to restrict use, you put in a clause saying that people may only distribute to other people who agree to the usage restriction. You have no right to distribute to someone who doesn't agree. That is, if you're distributing to someone who doesn't agree, you're distributing illegal copies. Since there is no right to use illegal copies, anyone who doesn't agree has no right to use the program. So either the recipient agrees (and has the right to use the program only in accordance with the restrictions) or disagrees, and received a pirated copy, and has no right to use it at all. Notable flaws of this scheme include the fact that since the distributor has to get his recipients to agree to a contract, it would be impossible to distribute the software in situations like stores or ftp sites. This might be fixable if the author agrees in advance not to sue such distributors, but I'm not sure; again, I'm not a lawyer.
Re: SOS license
On Wed, 10 Nov 1999, Brian Behlendorf wrote: No. You can charge for the patch, you can charge for the act of giving the patch to someone, but you can't charge for the patched version. Hmm. What are the implications for this compared to the FSF's stand on "user does the link"?
Re: Does a GPL API infect its apps?
On Thu, 21 Oct 1999, Ross N. Williams wrote: I just want to point out that the use of the virus metaphor to describe the legalities of association in relation to the GPL is inappropriate because the "infection" does not proceed beyond one level. GNU GPL V2: 2. These requirements apply to the modified work as a whole. If identifiable sections of that work are not derived from the Program, and can be reasonably considered independent and separate works in themselves, then this License, and its terms, do not apply to those sections when you distribute them as separate works. But when you distribute the same sections as part of a whole which is a work based on the Program, the distribution of the whole must be on the terms of this License, whose permissions for other licensees extend to the entire whole, and thus to each and every part regardless of who wrote it. That seems to keep it from proceeding beyond one level, but really, it doesn't. In fact, it does nothing at all. Suppose I add a function to a GPL program. Someone else wants to use the function--only the part I added, nothing from the original GPL program--with their own code. The GPL says that my added function must be GPL when it is distributed as part of a GPL program. Once it becomes GPL, it stays GPL--the GPL doesn't contain any clause saying that the function once again becomes non-GPL after it is removed from the program. This causes the virus effect, since the other person who is using my function must place their code under GPL. They would have to go directly to me in order to get a non-GPL copy of the function, even though that copy is byte-for-byte the same as the copy they snarfed from the program. Here's a hypothetical version which does not have this problem: But when you distribute the same sections as part of a whole which is a work based on the Program, the distribution of the whole must be on the terms of this License or a license which grants permissions which are identical to this License except for the possible addition of extra permissions on the sections which are identifiable works as described above.
Re: Does a GPL API infect its apps?
Now that I think about it, there's an even weirder result. Suppose I add a function to a GPL program, and I want to dual license the function so that it can be used with the GPL program but can also be used in way X (where X is not permitted by GPL. For instance, "may be used as a Netscape plugin".) I add the function to the GPL program, and then I add a copyright statement saying that the function in question is dual-licensed under GPL+X. Now we get into semantics. If dual-licensed means "is under a single license GPL+X whose terms mean that you can use it in either of two ways", then I cannot do this, because functions added to the GPL'ed program must be GPL'ed. The single license GPL+X is not GPL, so it is not allowed. If dual-licensed means "under a single license which may be one of (GPL, X), selected by the recipient", then every time the recipient selects X, he's received a GPL program with an added function licensed under X--this makes me a GPL violator depending on the whim of the recipient. The upshot of this? If I want to send someone a GPLed program with a function of mine, and I want the function to have additional permissions, I have to *send them two copies of the function*. This is bizarre. (I can't even relicense the copy they already have--that would retroactively create one of the two above cases and thus retroactively make me a GPL violator.)
Re: Does a GPL API infect its apps?
On Thu, 21 Oct 1999, John Cowan wrote: Here's a hypothetical version which does not have this problem: But when you distribute the same sections as part of a whole which is a work based on the Program, the distribution of the whole must be on the terms of this License or a license which grants permissions which are identical to this License except for the possible addition of extra permissions on the sections which are identifiable works as described above. This is already possible, since a distribution of the PART is not a distribution of the WHOLE. I don't think so. Call the GPL program A and the added part B. The GPL on A only says what you can do with A+B. It does not say what you can do with B alone. That *looks* like it's only restricting distribution of the whole and not the part. But one of the requirements of the GPL on A is that you also GPL A+B. The GPL on A+B restricts what you can do with derivatives of A+B (unlike the first GPL on A, which only restricts what you can do with derivatives of A). B (when obtained by cutting it from A+B) is a derivative of A+B, and therefore B is restricted by GPL even though it contains no A code. That is, the fact that you cannot take the section B out of the program and use it under another license is a second-order effect. The initial GPL on A does not prohibit this action, but it requires that you GPL A+B, and GPLing A+B in turn causes the prohibition. If B is dual-licensed, that dual license got lost at the step where you distributed A+B. You must distribute A+B under GPL-only, which makes that particular copy of B GPL-only too. Any dual-licensed copies of B must be obtained separately.
Re: Does a GPL API infect its apps?
On Thu, 21 Oct 1999, John Cowan wrote: If B is dual-licensed, that dual license got lost at the step where you distributed A+B. You must distribute A+B under GPL-only, which makes that particular copy of B GPL-only too. Any dual-licensed copies of B must be obtained separately. I think so. But this is easily cured by distributing both A+B and B. Yes, it is. But doesn't it seem a bit strange?
Re: Does a GPL API infect its apps?
On Fri, 22 Oct 1999, Ross N. Williams wrote: Thus, my understanding is that if A+B is GPLed, then you could still extract B and use it in a way that violates the GPL, so long as you can provide a complete genealogy for B under another licence or get all of the contributors to B to re-issue under another licence. If it is possible to provide a complete genealogy for B under another license, then you haven't extracted B. You've obtained a separate copy of B. The copy of B that is in A+B cannot have a genealogy under another license unless the _entire_ combination A+B has a genealogy under another license. Furthermore, my understanding is that if a program were released under GPL and modified by 1000 authors, then even ten years later, the entire mess could be taken commercial so long as all of the 1000 authors agreed to licence their contributions to the commercial venture. If the authors are contacted and they all agree to use a new license, they certainly can, of course. But this does require contacting the author--the author can't just say "you can license this under GPL or some other license" and expect people to be able to cut it out of the GPL program and use it under the other license. He has to include an extra copy or he has to be contacted at the moment of cutting.
Re: Free World Licence and clauses 5,6,8 9 of OSS definition.
On Tue, 19 Oct 1999, Ross N. Williams wrote: The FWL does not discriminate in relation to a PARTICULAR software distribution. Instead, it discriminates in relation to a CLASS of distributions - the free ones. ... clause 8 was designed to prevent OS vendors calling "OSS" software that could only be legally run on their platform. If it means what you say it means, what's to prevent someone from defining a class consisting of "Windows 95 and all successor operating systems"? That's certainly a class as opposed to a particular operating system. "With respect to a particular software distribution" means that if you need to know what the distribution is to know whether you're allowed to use the software, it's failed the test.
Re: Free World Licence and clauses 5,6,8 9 of OSS definition.
On Tue, 19 Oct 1999, Ross N. Williams wrote: The FWL does not discriminate in relation to a PARTICULAR software distribution. Instead, it discriminates in relation to a CLASS of distributions - the free ones. If it means what you say it means, what's to prevent someone from defining a class consisting of "Windows 95 and all successor operating systems"? But your example is a class defined by an identity, not a property. I'm not sure what you mean by an identity. If you mean that it defines its class only by listing all its members, then no, it doesn't. It doesn't even mention Windows 98, yet Windows 98 would be included in it. And honestly, if _you_ saw the phrase "particular", would you really think it meant "being a member of a class defined by an identity, but not a class defined by a property"? That's just too long and convoluted a definition to have plausibly been intended all along.
Re: Simple Public License, draft
On Thu, 2 Sep 1999, Justin Wells wrote: This "or any later version" has always bugged me. What if the FSF loses a lawsuit against Evil Corp., and in order to pay a judgement, all of the FSF's intellectual property rights are awarded to Evil Corp.; which then releases a version of the GPL which allows Evil Corp. to make proprietary use of any GPL'd software. ... That's why I don't do it. If someone can convince me that this would never happen, I don't mind adding the "or any later version", but in the past, for this reason, I have always specified an exact version of the GPL for my software. What happens if later on there's a GPL version 3, and someone wants to mix code from a GPL program which uses version 3, and yours?
Re: Put it in laymen's terms
On 1 Aug 1999 [EMAIL PROTECTED] wrote: However, you can also take Linus' note as an interpretation of the scope of the GPL and not an exception at all. If you accept that another person can reinterpret phrases like "dervived work", and you also accept that this reinterpretation can apply to code written by other people, then the GPL is for all practical purposes nonexistent. If someone wants to violate it, they just have to redefine "redistribution", "derived work", etc. so that their violation is not really a violation. It doesn't make sense that one person can reinterpret what a phrase in someone else's license means.
Re: A new open source license
Perhaps this is a bit of an odd objection, since I'm objecting to something taken verbatim from the GPL, but I've argued about this clause in the GPL before. On 11 May 1999, Russell Nelson wrote: You are not required to accept this license, since you have not signed it. However, nothing else grants you permission to modify or distribute any code, library, or any derivative works thereof. These actions are prohibited by law if you do not accept this license. The logical consequence of this is not "somebody who distributes has accepted the license". It is "somebody who distributes either has accepted the license, or is illegally pirating the Code". This is not trivial, since a distributor might decide that they are better off being a pirate than having accepted the license; for instance, consider what happened to MOSIX, where if the author accepted the GPL he might be violating Israeli export restrictions. I would imagine that violating those would lead to punishment rather more severe than the punishment for pirating software. Therefore, by using, modifying or distributing any code, library (or any work based on the code), you indicate your acceptance of the License set forth below, and all its terms and conditions for copying, distributing or modifying such Code, library, or works based on them. The clause following the "Therefore" doesn't logically follow from the clauses that it is supposed to logically follow from. Someone who copies the code doesn't _automatically_ accept the conditions. If they distribute but don't accept the conditions, then they're a software pirate and you can sue them... but you can't act as if the license has been accepted and can't, for instance, copy the modifications.