Re: Linux and GPLv2
Raul Miller [EMAIL PROTECTED] writes: If you can find us a country whose laws make this illegal, this issue would be worth discussing. On Fri, Apr 01, 2005 at 06:15:34PM +0200, Måns Rullgård wrote: You are obviously convinced that using a command line interface can't be protected by copyright. Right, in the sense that copyright is about tangible forms of creative expression, and it's not about functional mechanisms such as interfaces. Mechanisms like header files, for instance. So, for example, this lack of copyright protection for command line interfaces doesn't mean that I can put some copyrighted story on the command line and make its copyright protection go away. If writing that story on the command line is required for using the program, I can think of three possibilities: 1) the author of the program owns the copyright to the story, and lets you use the story in this way, 2) the author of the program owns the copyright to the story, doesn't let you use it, and effectively prevents you from using the program at all, which would seem quite silly, and 3) the author of the program doesn't own the copyright to the story, and has no possibility of allowing you to use it, which is even sillier. Why, then, are you so persistent in insisting that other interfaces somehow are awarded such protection? That wasn't what I was trying to say. I was trying to say that just because something is a relevant interface in case A doesn't mean that that kind of interface is relevant in case B. Who gets to decide what is relevant, and when? -- Måns Rullgård [EMAIL PROTECTED] -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Linux and GPLv2
On Thu, Mar 31, 2005 at 09:17:51PM +0200, Måns Rullgård wrote: Thanks for mentioning command lines. Running a program from the command line, usually involves passing it options. These options are (obviously) copies of strings from the actual program. Can this copying be a copyright violation? IMHO, it is no different, in principle, from using function names declared in a header file. If you can find us a country whose laws make this illegal, this issue would be worth discussing. If the laws of that country were available in english, we could probably do justice to that (hypothetical) discussion. If there are any such countries, and they make a practice of enforcing such laws (rather than just having them on the books to confuse novices), we should probably put together a page warning people to about using computers in that country (or those countries). -- Raul -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Linux and GPLv2
Raul Miller [EMAIL PROTECTED] writes: On Thu, Mar 31, 2005 at 09:17:51PM +0200, Måns Rullgård wrote: Thanks for mentioning command lines. Running a program from the command line, usually involves passing it options. These options are (obviously) copies of strings from the actual program. Can this copying be a copyright violation? IMHO, it is no different, in principle, from using function names declared in a header file. If you can find us a country whose laws make this illegal, this issue would be worth discussing. If the laws of that country were available in english, we could probably do justice to that (hypothetical) discussion. If there are any such countries, and they make a practice of enforcing such laws (rather than just having them on the books to confuse novices), we should probably put together a page warning people to about using computers in that country (or those countries). You are obviously convinced that using a command line interface can't be protected by copyright. Why, then, are you so persistent in insisting that other interfaces somehow are awarded such protection? -- Måns Rullgård [EMAIL PROTECTED] -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Linux and GPLv2
On Fri, 01 Apr 2005, Måns Rullgård wrote: You are obviously convinced that using a command line interface can't be protected by copyright. Why, then, are you so persistent in insisting that other interfaces somehow are awarded such protection? Whether or not a specific interface is covered by copyright is necessarily jurisdictionally dependent. A conservative tack is to assume that if there's any creative component at all, then there is a possibility of copyright. [Even that may not go far enough, as some things that are devoid of creativity may have the protection of copyright in specific localities, cf. the database directive.] If you wish to say that there is no copyright protection for a specific instance in a specific jurisdiction, that may indeed be the case,[1] but it's quite irresponsible to claim that it is so for all jurisdictions. Don Armstrong 1: If it is so, I'd strongly suggest finding relevant case law or talking to a lawyer before using this to take actions which would be infringing if a copyright actually did exist. -- Quite the contrary; they *love* collateral damage. If they can make you miserable enough, maybe you'll stop using email entirely. Once enough people do that, then there'll be no legitimate reason left for anyone to run an SMTP server, and the spam problem will be solved. -- Craig Dickson in [EMAIL PROTECTED] http://www.donarmstrong.com http://rzlab.ucr.edu
Re: Linux and GPLv2
Those .h files were held to be not protected by copyright because no viable alternatives were available to interface with the system. It's hard to see how this reasoning would apply in a context where there is some viable alternative available to interface with the system. On Wed, Mar 30, 2005 at 04:15:57AM +0200, Måns Rullgård wrote: Alternative to what? There can be no alternative to the full set of interfaces to the system. Are you trying to argue, that several interfaces exist, use of each one is protected due to the existence of the others? For example: gcc provides a command line interface as an alternative to rebuilding gcc every time you need to compile a program. Suppose there is only one interface, such that it, per your reasoning, is not protected. Now add another. Does this addition suddenly make the first interface protected? What if they were created in the opposite order? That all depends on the design of the program in question, how it's documented, how it's licensed, and so on... -- Raul -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Linux and GPLv2
Raul Miller [EMAIL PROTECTED] writes: Those .h files were held to be not protected by copyright because no viable alternatives were available to interface with the system. It's hard to see how this reasoning would apply in a context where there is some viable alternative available to interface with the system. On Wed, Mar 30, 2005 at 04:15:57AM +0200, Måns Rullgård wrote: Alternative to what? There can be no alternative to the full set of interfaces to the system. Are you trying to argue, that several interfaces exist, use of each one is protected due to the existence of the others? For example: gcc provides a command line interface as an alternative to rebuilding gcc every time you need to compile a program. Thanks for mentioning command lines. Running a program from the command line, usually involves passing it options. These options are (obviously) copies of strings from the actual program. Can this copying be a copyright violation? IMHO, it is no different, in principle, from using function names declared in a header file. -- Måns Rullgård [EMAIL PROTECTED] -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Linux and GPLv2
On Wednesday 30 March 2005 03:53, Raul Miller wrote: Those .h files were held to be not protected by copyright because no viable alternatives were available to interface with the system. It's hard to see how this reasoning would apply in a context where there is some viable alternative available to interface with the system. I don't know the details of the case at hand, but I remember the discussion around errno.h from the TSG fallout: The basic reasoning there was, that if one wants to implement a C stdlib an a unix-like system, a optimal errno.h would always look similar to that from ancient BSD (which most modern errno.h derive from). Since there is no way to be unixish/compatible without defining the various E* to these values, having a errno.h file with the same values is not infringing. This, I believe, can be extended to all forms of compatability: If a header file with certain contents is needed to use the interface of a library, it is no copyright infringement. To be on the safe side, this has to be interpreted very strict: Non-trivial comments and inline functions are probably not covered. Regards, David -- - hallo... wie gehts heute? - *hust* gut *rotz* *keuch* - gott sei dank kommunizieren wir über ein septisches medium ;) -- Matthias Leeb, Uni f. angewandte Kunst, 2005-02-15
Re: Linux and GPLv2#
In message [EMAIL PROTECTED], Glenn Maynard [EMAIL PROTECTED] writes On Mon, Mar 28, 2005 at 01:30:16PM -0500, Raul Miller wrote: Andrew seems to avoid Red Herring arguments more than I. I asked for the rationale behind his calling fair use a perversion, and he refused to supply one. It's that simple. (The only reason I can find is that he had no reason; he just attacks things--people and laws alike--out of sheer habit, not for any particular reason.) I thought he gave a simple answer - if he tried (or I tried) to claim fair use in court, the judge would simply reply what's that? and then convict us of copyright theft. Cheers, Wol -- Anthony W. Youngman - wol at thewolery dot demon dot co dot uk HEX wondered how much he should tell the Wizards. He felt it would not be a good idea to burden them with too much input. Hex always thought of his reports as Lies-to-People. The Science of Discworld : (c) Terry Pratchett 1999 -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Linux and GPLv2#
On Mon, Mar 28, 2005 at 01:43:53PM -0500, Glenn Maynard wrote: On Mon, Mar 28, 2005 at 01:30:16PM -0500, Raul Miller wrote: Andrew seems to avoid Red Herring arguments more than I. I asked for the rationale behind his calling fair use a perversion, and he refused to supply one. It's that simple. Maybe he was using the word in a facetious/stylistic manner, rather than a literal one, and didn't feel it needed justification. --Adam -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Linux and GPLv2#
On Tue, Mar 29, 2005 at 09:34:42PM +0100, Anthony W. Youngman wrote: In message [EMAIL PROTECTED], Glenn Maynard [EMAIL PROTECTED] writes On Mon, Mar 28, 2005 at 01:30:16PM -0500, Raul Miller wrote: Andrew seems to avoid Red Herring arguments more than I. I asked for the rationale behind his calling fair use a perversion, and he refused to supply one. It's that simple. (The only reason I can find is that he had no reason; he just attacks things--people and laws alike--out of sheer habit, not for any particular reason.) I thought he gave a simple answer - if he tried (or I tried) to claim fair use in court, the judge would simply reply what's that? and then convict us of copyright theft. I didn't ask why fair use isn't useful for Debian. I knew the answer to that long before this thread. I asked him why he considers fair use itself to be a bad thing. -- Glenn Maynard -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Linux and GPLv2#
http://dilbert.com/comics/dilbert/archive/dilbert-20050324.html I am continually entertained by the way that Adams manages to be right all the time. -- .''`. ** Debian GNU/Linux ** | Andrew Suffield : :' : http://www.debian.org/ | `. `' | `- -- | signature.asc Description: Digital signature
Re: Linux and GPLv2
On Mon, Mar 28, 2005 at 11:25:39AM -0300, Humberto Massa wrote: My claim was: *Basically*, bits in .h files are not copyrightable. Which I now solemnly amend to The kind of bits you normally (99% of the times) find in .h files in c-language based projects, and often (50% of the times) find in .h files in c++ based projects, are those defining interfaces, deeming them uncopyrightable by current USofAn and Brazilian law. Better? Raul Miller wrote: However, for U.S. law, this isn't necessarily the case. On Mon, Mar 28, 2005 at 04:14:47PM -0300, Humberto Massa wrote: I was referring to the fact that there is some case law in the USofA that deemed interface definitions, as present normally in .h files, uncopyrightable. HTH Those .h files were held to be not protected by copyright because no viable alternatives were available to interface with the system. It's hard to see how this reasoning would apply in a context where there is some viable alternative available to interface with the system. -- Raul -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Linux and GPLv2
Raul Miller [EMAIL PROTECTED] writes: On Mon, Mar 28, 2005 at 11:25:39AM -0300, Humberto Massa wrote: My claim was: *Basically*, bits in .h files are not copyrightable. Which I now solemnly amend to The kind of bits you normally (99% of the times) find in .h files in c-language based projects, and often (50% of the times) find in .h files in c++ based projects, are those defining interfaces, deeming them uncopyrightable by current USofAn and Brazilian law. Better? Raul Miller wrote: However, for U.S. law, this isn't necessarily the case. On Mon, Mar 28, 2005 at 04:14:47PM -0300, Humberto Massa wrote: I was referring to the fact that there is some case law in the USofA that deemed interface definitions, as present normally in .h files, uncopyrightable. HTH Those .h files were held to be not protected by copyright because no viable alternatives were available to interface with the system. It's hard to see how this reasoning would apply in a context where there is some viable alternative available to interface with the system. Alternative to what? There can be no alternative to the full set of interfaces to the system. Are you trying to argue, that several interfaces exist, use of each one is protected due to the existence of the others? Suppose there is only one interface, such that it, per your reasoning, is not protected. Now add another. Does this addition suddenly make the first interface protected? What if they were created in the opposite order? -- Måns Rullgård [EMAIL PROTECTED] -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Linux and GPLv2
Glenn Maynard [EMAIL PROTECTED] writes: On Tue, Mar 29, 2005 at 08:53:52PM -0500, Raul Miller wrote: On Mon, Mar 28, 2005 at 11:25:39AM -0300, Humberto Massa wrote: My claim was: *Basically*, bits in .h files are not copyrightable. Which I now solemnly amend to The kind of bits you normally (99% of the times) find in .h files in c-language based projects, and often (50% of the times) find in .h files in c++ based projects, are those defining interfaces, deeming them uncopyrightable by current USofAn and Brazilian law. Better? Raul Miller wrote: However, for U.S. law, this isn't necessarily the case. On Mon, Mar 28, 2005 at 04:14:47PM -0300, Humberto Massa wrote: I was referring to the fact that there is some case law in the USofA that deemed interface definitions, as present normally in .h files, uncopyrightable. HTH Those .h files were held to be not protected by copyright because no viable alternatives were available to interface with the system. I'd question whether that'd apply to a *free* system, anyway. I havn't looked at these cases (since I don't know which they are), but I recall a case that sounds just like it: an author of a work created (under contract) for a movie claimed that no license to actually use that material was granted, but as the paid-for work was useless without a license to use it, a license was implied. That doesn't seem relevant where the work is being given out entirely for free; the creator has no obligation to anyone else to grant a license to make the library's release useful. (For a commercial SDK, this would seem to apply to header files.) So now the degree of protection by copyright depends on how much you charge for it? What if someone gets paid to develop open source? -- Måns Rullgård [EMAIL PROTECTED] -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Linux and GPLv2
On Wed, Mar 30, 2005 at 05:41:10AM +0200, Måns Rullgård wrote: I'd question whether that'd apply to a *free* system, anyway. I havn't looked at these cases (since I don't know which they are), but I recall a case that sounds just like it: an author of a work created (under contract) for a movie claimed that no license to actually use that material was granted, but as the paid-for work was useless without a license to use it, a license was implied. That doesn't seem relevant where the work is being given out entirely for free; the creator has no obligation to anyone else to grant a license to make the library's release useful. (For a commercial SDK, this would seem to apply to header files.) So now the degree of protection by copyright depends on how much you charge for it? What if someone gets paid to develop open source? Then, I'd imagine--which is the best I can do; this is over my layman's head--that the person that paid him would receive such an implicit license for the header files. That is, if I pay you to write a library for use in my work, you can't say after the fact: here's the library, you can use it all you want, but you can't copy the material in the headers, since the work I paid for only has value if I can actually distribute programs that use it. (Just as special effects created for a movie only have value if the movie producer can actually put it in the movie.) I don't see that this affects open source as such: it's between the creator of the work and the person that paid him to do so. The license other parties receive is unrelated, and the creator can--other contracts so on permitting, of course--license or not license to other people however he pleases. (Maybe I'm miles off; I'm open to informed corrections.) -- Glenn Maynard -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Linux and GPLv2
Glenn Maynard wrote: On Tue, Mar 29, 2005 at 08:53:52PM -0500, Raul Miller wrote: On Mon, Mar 28, 2005 at 11:25:39AM -0300, Humberto Massa wrote: My claim was: *Basically*, bits in .h files are not copyrightable. Which I now solemnly amend to The kind of bits you normally (99% of the times) find in .h files in c-language based projects, and often (50% of the times) find in .h files in c++ based projects, are those defining interfaces, deeming them uncopyrightable by current USofAn and Brazilian law. Better? Raul Miller wrote: However, for U.S. law, this isn't necessarily the case. On Mon, Mar 28, 2005 at 04:14:47PM -0300, Humberto Massa wrote: I was referring to the fact that there is some case law in the USofA that deemed interface definitions, as present normally in .h files, uncopyrightable. HTH Those .h files were held to be not protected by copyright because no viable alternatives were available to interface with the system. I'd question whether that'd apply to a *free* system, anyway. I havn't looked at these cases (since I don't know which they are), but I recall a case that sounds just like it: an author of a work created (under contract) for a movie claimed that no license to actually use that material was granted, but as the paid-for work was useless without a license to use it, a license was implied. That doesn't seem relevant where the work is being given out entirely for free; the creator has no obligation to anyone else to grant a license to make the library's release useful. (For a commercial SDK, this would seem to apply to header files.) Glenn, If memory serves the two landmark US interface cases were 1) Lotus where a third party had written a commercial macro package that used 1-2-3's Macro interface, and 2) Lexmark or HP where a third party had written software that integrated with their printer control system. Alas, I am a long way from my files so all or some of the above may be correct. Dave -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Linux and GPLv2
Andrew Suffield wrote: Fair use is an American perversion. It does not exist in most of the rest of the world in anything like the same form. Anything that relies on the American notion of fair use is non-free, because in the UK that means Non-commercial use only. Just to be clear: fair use means in USC 17 and in Brazilian Author's rights act, that copying any copyrighted work for: academic quotation/commentary/study, parody and some others (I'll look it up later) is granted, and without copyright protection. First sale doctrine means that if you have some license, you can pass it along to another party (losing your license, of course). -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Linux and GPLv2
Henning Makholm wrote: Your claim was: Bits in .h files are not copyrightable. Troll editing. My claim was: *Basically*, bits in .h files are not copyrightable. Which I now solemnly amend to The kind of bits you normally (99% of the times) find in .h files in c-language based projects, and often (50% of the times) find in .h files in c++ based projects, are those defining interfaces, deeming them uncopyrightable by current USofAn and Brazilian law. Better? It's the fact that you do not transform it, so you do not create a derivative work. Derivative work, anthology: Does the difference matter? To copy either you need the permission of the original author. Yes, but (1) the GPL exempts anthology works, in the mere aggregation clause; and (2) the rules are slightly different on the distribution of anthology works. So if I buy a book that has been produced in an offset press, you assert that the book is not *per se* copyrightable. Hence I can freely create as many reprints as I like and sell them? You are disappointing me. By now, I would have expected you to understand: The copyright does not apply to the printed book /per-se/... it's applied to the intellectual content (the original creation of spirit). Imagine I make a program that prints random words. The result is not copyrightable, even if it makes any sense at all. But you came up with: That has nothing to do with whether an offset press has been produced to print the words. The *words* are copyrighted, not the book. That is the part I wanted you to understand: No, it's a processed work. Which is still coverved by the copyright of the original work. You are missing the difference: Not derived. Never. To derive you need inteligence (in Brazilian letter-of-law, spirit). Says who? Again, the offset press is not intelligent, but the books that it spits out are still covered by the author's copyright. Not the book. The words. Ok, so you are saying that: oh, well, there are some words in the .o file that came from the .h file. And I am counterargumenting: were those words the interface definition? If so, then the law explicitly exempted those from copyright protection. Why do you distinguish between those to cases? Repeating: (1) they are distinguished by GPL and (2) they are distinguished by copyright law. The rules for anthology works and derivative works are different. How so? Your examples seem to try to explain how you choose to apply different words to slightly different situations, but the end result about which rights you need to have to copy the result is the same. Let's see if I can get to the right point here: 1. The GPL has two different disposition about derivative works versus anthology works. At first (clause 0) it tries do redefine what it will call works based on the Program, and does a lousy job at that, because it ends with two different definitions (square parentheses are mine: The Program, below, refers to any such program or work, and a work based on the Program means either the Program or any derivative work under copyright law [definition A]: that is to say, a work containing the Program or a portion of it [redefinition B], either verbatim or with modifications and/or translated into another language. Ok, everybody can claim almost any one of the two definitions as the right one, ie, that one that will apply to the scope of the GPL, except that in clause 2, paragraph 3, the GPL separates what would cover anthology works: mere aggregation of another work not based on the Program with the Program (or with a work based on the Program) on a volume of a storage or distribution medium does not bring the other work under the scope of this License. Notice that the FSF used the work based on the Program again. It's pretty clear to me that many if not all Brazilian judges facing this would consider these two dispositions as separating derivative works from aggregate works. Now, if the library license was a proprietary license, that did not even permit aggregate works, I would give in, but in the case of a GPL'd library, what you *do* have when you #include a file is a reference to an interface, and in the case where copyrightable bits end up in the final .o or a.out/elf file, they are redistributable anyway. Massa -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Linux and GPLv2
Henning Makholm wrote: If it is wrong to begin with, then it is also basically wrong. It seems that my bad English got me: I used the word basically in the same sense that its cognate can be used in Portuguese, in the basic/normal cases I apologize for the confusion. Which I now solemnly amend to The kind of bits you normally (99% of the times) find in .h files in c-language based projects, and often (50% of the times) find in .h files in c++ based projects, are those defining interfaces, deeming them uncopyrightable by current USofAn and Brazilian law. Better? Well, yes. And which conclusion do you draw from that observation? It is my most humble, but firm opinion: 1. That a plugin (written in C or C++) that includes an GPL'd plugin-interfaces.h file, and uses the interfaces described there for its normal chores, is not a derivative work of the GPL'd I can load plugins program. 2. That said plugin, when in its compiled form, if it contains at all any inlined functions or macros from the GPL'd plugin-interfaces.h file, is merely a volume of storage or distribution in accordance to the disposition on the GPL, section 2, paragraph 3. 2.a. That is, as long as it does not mess with implementation details of the plugin-loading mechanism, attaining to the use of published interfaces -- if it doesn't, pieces of the implementation will leak past the Filtration phase and it can be deemed a derivative work on the plugin loader. 3. That the plugin's licensing is completely independent of the I can load plugins program's one, and that it can be licensed as the author sees fit. Which, of course, includes the QPL. 4. That, reinforcing, no dispostion in the GPL would stop the user from linking dynamically at run-time said plugin, to make possible its use. HTH, Massa -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Linux and GPLv2
On Mon, Mar 28, 2005 at 11:25:39AM -0300, Humberto Massa wrote: Troll editing. My claim was: *Basically*, bits in .h files are not copyrightable. Which I now solemnly amend to The kind of bits you normally (99% of the times) find in .h files in c-language based projects, and often (50% of the times) find in .h files in c++ based projects, are those defining interfaces, deeming them uncopyrightable by current USofAn and Brazilian law. Better? I don't know about Brazilian law. However, for U.S. law, this isn't necessarily the case. A key feature of U.S. copyright law is that it's creative expression which is being protected -- not form or function. In other words, if the software in question provides some other interface then U.S. law overriding copyright for interface purposes probably wouldn't apply. [A strong legal case could be made here, though I don't think anyone has actually taken this particular issue to court.] On the other hand, in a case where this mattered, the .h files would not probably not be considered in isolation. You could say that copyright law does not factor issues on functional boundaries, except as a notational convenience. Instead, in the context of copyright, issues are factored on creative boundaries. -- Raul -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Linux and GPLv2#
On Sat, Mar 26, 2005 at 11:16:29AM -0500, Raul Miller wrote: This seems to imply that there's some single perspective which can be easily described to clarify this issue. On Sat, Mar 26, 2005 at 12:08:08PM -0500, Glenn Maynard wrote: I was asking for Andrew's perspective, not The Perspective. Now you seem to be implying that Andrew has only one perspective on this issue. I don't even feel strongly about this, and wasn't particularly expecting to argue about the answer; I was just curious why he dislikes fair use so much as to use such a strongly negative word, and found it strange that a simple why do you think that? inquiry resulted in evasive maneuvers. Andrew seems to avoid Red Herring arguments more than I. -- Raul -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Linux and GPLv2
Scripsit Humberto Massa [EMAIL PROTECTED] 2. That said plugin, when in its compiled form, if it contains at all any inlined functions or macros from the GPL'd plugin-interfaces.h file, is merely a volume of storage or distribution in accordance to the disposition on the GPL, section 2, paragraph 3. You have a really strange and untenable understanding of merely, then. 3. That the plugin's licensing is completely independent of the I can load plugins program's one, and that it can be licensed as the author sees fit. Yes, but not of the license for the inlined functions that get compiled in, providing that those are sufficiently nontrivial to enjoy copyright protection at all. -- Henning Makholm It will be useful even at this early stage to review briefly the main features of the universe as they are known today. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Linux and GPLv2#
On Mon, Mar 28, 2005 at 01:30:16PM -0500, Raul Miller wrote: Andrew seems to avoid Red Herring arguments more than I. I asked for the rationale behind his calling fair use a perversion, and he refused to supply one. It's that simple. (The only reason I can find is that he had no reason; he just attacks things--people and laws alike--out of sheer habit, not for any particular reason.) It's fairly disappointing that a simple request for rationale--an honest attempt to understand someone's opinion--results in such a waste of time as this thread, and no rationale. -- Glenn Maynard -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Linux and GPLv2
Raul Miller wrote: On Mon, Mar 28, 2005 at 11:25:39AM -0300, Humberto Massa wrote: Troll editing. My claim was: *Basically*, bits in .h files are not copyrightable. Which I now solemnly amend to The kind of bits you normally (99% of the times) find in .h files in c-language based projects, and often (50% of the times) find in .h files in c++ based projects, are those defining interfaces, deeming them uncopyrightable by current USofAn and Brazilian law. Better? I don't know about Brazilian law. However, for U.S. law, this isn't necessarily the case. I was referring to the fact that there is some case law in the USofA that deemed interface definitions, as present normally in .h files, uncopyrightable. HTH -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Linux and GPLv2
Henning Makholm wrote: Scripsit Humberto Massa [EMAIL PROTECTED] 2. That said plugin, when in its compiled form, if it contains at all any inlined functions or macros from the GPL'd plugin-interfaces.h file, is merely a volume of storage or distribution in accordance to the disposition on the GPL, section 2, paragraph 3. You have a really strange and untenable understanding of merely, then. No, but I have already explained why I firmly believe that the merely aggregates language in GPL#2§3 refers to all aggregate works, as opposed to the works based on the Program from GPL#0. 3. That the plugin's licensing is completely independent of the I can load plugins program's one, and that it can be licensed as the author sees fit. Yes, but not of the license for the inlined functions that get compiled in, providing that those are sufficiently nontrivial to enjoy copyright protection at all. We'll have to agree in disagreeing :-) Friendly -- /Really/, Massa -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Linux and GPLv2
Scripsit Humberto Massa [EMAIL PROTECTED] No, but I have already explained why I firmly believe that the merely aggregates language in GPL#2§3 refers to all aggregate works, as opposed to the works based on the Program from GPL#0. If you believe that a compiled program that contains code from both sources combined into a monolithic executable object is merely an aggregation, then I can only conclude that you have no idea whatsoever what the word merely means. -- Henning Makholm Al lykken er i ét ord: Overvægtig!
Re: Linux and GPLv2
Scripsit Anthony W. Youngman [EMAIL PROTECTED] Even when the work is not copyrightable? (eg header files :-) It is false that header files are not copyrightable. -- Henning Makholm What has it got in its pocketses? -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Linux and GPLv2#
On Fri, Mar 25, 2005 at 09:55:40AM +, Andrew Suffield wrote: 'Worse' is purely a matter of perspective. There's irony here... On Fri, Mar 25, 2005 at 05:31:27AM -0500, Glenn Maynard wrote: No, there isn't. It's very simple. You called it a perversion, which means you think it's worse. I asked why you think it's a perversion. (In other words, what is your perspective that makes you consider is worse?) You won't answer. This seems to imply that there's some single perspective which can be easily described to clarify this issue. Personally, while I wouldn't describe things using the terms Andrew has, I think I see something similar to his point. For example, one perspective is: computer programs, in the general case, shouldn't be regulated by copyright law. Note that this particular flavor of perspective is not claiming that computer programs should not be regulated (though it does fit with that kind of argument). It could just as easily be a part of an argument for other kinds of regulations (perhaps far more restrictive than anything we've seen to date). That said, it's probably worth noting that until the mid-1970s, computer programs were thought to be not copyrightable. And, because of the economics of the time this wasn't a big deal for a lot of people. I think one of the reasons IBM has taken to open sourced software so well is that their business model was developed under those conditions -- they actually started losing ground, financially, when they switched to an ultra-restrictive license model. [But they were in some sense forced to do so, because if something isn't copyright protected it can be incorporated in something which is copyright protected by someone else, which has unpleasant business connotations.] Anyways, the original quip should probably just be taken as a statement that Andrew is dissatisfied with the system of copyright laws. We don't have to turn this into some kind of forum for discussing potential alternative systems. -- Raul -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Linux and GPLv2#
On Sat, Mar 26, 2005 at 11:16:29AM -0500, Raul Miller wrote: On Fri, Mar 25, 2005 at 09:55:40AM +, Andrew Suffield wrote: 'Worse' is purely a matter of perspective. There's irony here... On Fri, Mar 25, 2005 at 05:31:27AM -0500, Glenn Maynard wrote: No, there isn't. It's very simple. You called it a perversion, which means you think it's worse. I asked why you think it's a perversion. (In other words, what is your perspective that makes you consider is worse?) You won't answer. This seems to imply that there's some single perspective which can be easily described to clarify this issue. I was asking for Andrew's perspective, not The Perspective. I don't even feel strongly about this, and wasn't particularly expecting to argue about the answer; I was just curious why he dislikes fair use so much as to use such a strongly negative word, and found it strange that a simple why do you think that? inquiry resulted in evasive maneuvers. -- Glenn Maynard -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Linux and GPLv2
On Wed, 23 Mar 2005 10:49:34 +0100 (MET) Gerardo Ballabio wrote: From: Francesco Poli [EMAIL PROTECTED] [...] I don't think it's forbidding to remove the code: it's merely forbidding to drop a feature. You could reimplement it in a better (or even worse) way, but you must support it. Then if my reimplementation has a bug that prevents it from working properly, I may be accused of infringement? I don't know (IANAL). Maybe if it can be proved that your reimplementation is intentionally buggy. But I repeat: I don't really know... Anyway I agree it's non-free. Then you may be surprised to hear that the GPLv2 does have a don't remove this feature clause: Well, I was aware of that clause (even though maybe I was not thinking about it, while discussing the get the source through HTTP feature and its corresponding do not drop this feature clause...). c) If the modified program normally reads commands interactively when run, you must cause it, when started running for such interactive use in the most ordinary way, to print or display an announcement including an appropriate copyright notice and a notice that there is no warranty (or else, saying that you provide a warranty) and that users may redistribute the program under these conditions, and telling the user how to view a copy of this License. (Exception: if the Program itself is interactive but does not normally print such an announcement, your work based on the Program is not required to print an announcement.) This clause is known to be controversial. I would be happier if it were not present at all in the GNU GPL... I'm not sure, but my mind seems to recall to have read a statement from the FSF itself recommending against abusing this clause: but, after searching for such a statement for a while, I failed to find it and gave up... :( However it's some orders of magnitude less demanding than the do not drop the get-the-source-through-HTTP feature clause, I would say. Printing one or two statements to stdout or in a splash screen is really really easier than implementing (or keeping) HTTP support and the capability to send the program's own source. I think that clause 2c of GPLv2 is really borderline with respect to DFSG-freeness, but judging whether it's on one side or on the other one is not easy. On the one hand, it resembles to an acceptable requirement to include copyright notices and warranty disclaimers. On the other hand, it's a functional constraint, though a very little one... I'd even read this as saying if the original program isn't interactive, but the modified one is, you must _add_ that feature. I had never noticed that, I must admit! :-( But reading and re-reading the clause seems to bring me to same conclusion... And this is a little worrysome... -- :-( This Universe is buggy! Where's the Creator's BTS? ;-) .. Francesco Poli GnuPG Key ID = DD6DFCF4 Key fingerprint = C979 F34B 27CE 5CD8 DC12 31B5 78F4 279B DD6D FCF4 pgpAFMBskfg4c.pgp Description: PGP signature
Re: Linux and GPLv2#
On Fri, Mar 25, 2005 at 09:55:40AM +, Andrew Suffield wrote: Fair use is an American perversion. It does not exist in most of the rest of the world in anything like the same form. Anything that relies on the American notion of fair use is non-free, because in the UK that means Non-commercial use only. Out of curiosity, what is it about fair use that makes you call it a perversion? In opposition to the norm. That's the real definition, once you discard the arbitrary labels of 'right' and 'wrong'. Something which is merely in opposition to the norm is unusual. A perversion is something which is both unusual and worse. Fair use may be unusual, but I don't really understand how it's worse. 'Worse' is purely a matter of perspective. There's irony here... No, there isn't. It's very simple. You called it a perversion, which means you think it's worse. I asked why you think it's a perversion. (In other words, what is your perspective that makes you consider is worse?) You won't answer. -- Glenn Maynard -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Linux and GPLv2
On Thu, Mar 24, 2005 at 07:45:24AM +, Andrew Suffield wrote: Fair use is an American perversion. It does not exist in most of the rest of the world in anything like the same form. Anything that relies on the American notion of fair use is non-free, because in the UK that means Non-commercial use only. Out of curiosity, what is it about fair use that makes you call it a perversion? It may not be useful to Debian's purposes, but I have a hard time looking down on things that weaken the strength of IP laws, even if only by a little. (Perhaps it causes confusion about what's free enough and all that, but that's the licensor's fault, not the law's.) -- Glenn Maynard -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Linux and GPLv2
Andrew Suffield [EMAIL PROTECTED] writes: On Wed, Mar 23, 2005 at 08:10:57PM +0100, M?ns Rullg?rd wrote: Henning Makholm [EMAIL PROTECTED] writes: Concluding: when you write a .c file, it is or not a derivative work on another original work independently of what the compiler and linker will do in the future. I repeat: No, but the resulting .o file may be derived from another work that the compiler also read while producing it. The object file may contain bits from header files, or whatever, but this has no bearing on the distributability of it. Nonsense. Literal copying is always copyright infringement. Unless you had permission to make copies, which the GPL explicit grants you. We were talking about GPL'd stuff here, right? They only found their way there as the result of implementation details. Under your rather strange theory, copying a file can never be copyright infringement, because the way cp moves the bits around is just an 'implementation detail'. So presumably you don't think copyright infringement using a computer is possible. You are obviously deliberately misinterpreting what I said. -- Måns Rullgård [EMAIL PROTECTED] -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Linux and GPLv2
On Thu, Mar 24, 2005 at 10:55:40AM +0100, Måns Rullgård wrote: Andrew Suffield [EMAIL PROTECTED] writes: On Wed, Mar 23, 2005 at 08:10:57PM +0100, M?ns Rullg?rd wrote: Henning Makholm [EMAIL PROTECTED] writes: Concluding: when you write a .c file, it is or not a derivative work on another original work independently of what the compiler and linker will do in the future. I repeat: No, but the resulting .o file may be derived from another work that the compiler also read while producing it. The object file may contain bits from header files, or whatever, but this has no bearing on the distributability of it. Nonsense. Literal copying is always copyright infringement. Unless you had permission to make copies, which the GPL explicit grants you. We were talking about GPL'd stuff here, right? No, all of the above was spoken in very broad terms, not specifically about the GPL. You said The object file may contain bits from header files, or whatever, but this has no bearing on the distributability of it, which is false: if you create an object file with substantial bits from my header file, and I grant you no permission to redistribute the header file, the object file is undistributable as a direct result of containing bits from my header files. -- Glenn Maynard
Re: Linux and GPLv2
Scripsit Jeremy Hankins [EMAIL PROTECTED] Henning Makholm [EMAIL PROTECTED] writes: Scripsit Humberto Massa [EMAIL PROTECTED] Trying to explain more: my myfile.c is not a derivative work on errno.h, No, but myfile.o may be. (I feel like I'm repeating myself here). My understanding is that, in practice, myfile.c could infringe as well, if the only reasonable way to use it is by creating a work that is derivative of errno.h Some people say that. I do not agree with them. -- Henning MakholmI, madam, am the Archchancellor! And I happen to run this University! -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Linux and GPLv2
Scripsit Humberto Massa [EMAIL PROTECTED] Henning Makholm wrote: Snip explanation that does not do anything the idea that bits are treated differently by copyright just becuase they are in a file called .h. Repeating: bits that are in files called .h are not copied in your work, They may be, as I have explained. The compiled executable may or may not be an anthology work containing the contents of the .h. Which means that they are copied there. It's not the magical fact that the file is ending in .h (or .inc). Your claim was: Bits in .h files are not copyrightable. It's the fact that you do not transform it, so you do not create a derivative work. Derivative work, anthology: Does the difference matter? To copy either you need the permission of the original author. So what? They are being reproduced, and the mechanicalness of the reproduction does not prevent copyright from applying to the result. But the relationship of the result WRT the individual copyrights is different than many people (including who wrote the damned do not link paragraph in the GPL) expect. The relationship is: If you copy the result you need to have the permission of the author (or fall within whatever specific exemptions your jurisdiction happens to offer). Nothing that comes out of an automated process is *per-se* copyrightable. Do you still think this applies if the automated process is an offset press? Yes. So if I buy a book that has been produced in an offset press, you assert that the book is not *per se* copyrightable. Hence I can freely create as many reprints as I like and sell them? The copyright does not apply to the printed book /per-se/... it's applied to the intellectual content (the original creation of spirit). Imagine I make a program that prints random words. The result is not copyrightable, even if it makes any sense at all. That has nothing to do with whether an offset press has been produced to print the words. No, it's a processed work. Which is still coverved by the copyright of the original work. What you are calling a processed work is just another face of the same work; Whatever you like. The outcome is still that you need the original author's permission to copy it. I repeat: No, but the resulting .o file may be derived from another work that the compiler also read while producing it. Not derived. Never. To derive you need inteligence (in Brazilian letter-of-law, spirit). Says who? Again, the offset press is not intelligent, but the books that it spits out are still covered by the author's copyright. No, but myfile.o may be. (I feel like I'm repeating myself here). An anthology work, maybe, a derivative work, never. Why do you distinguish between those to cases? The rules for anthology works and derivative works are different. How so? Your examples seem to try to explain how you choose to apply different words to slightly different situations, but the end result about which rights you need to have to copy the result is the same. -- Henning Makholm However, the fact that the utterance by Epimenides of that false sentence could imply the existence of some Cretan who is not a liar is rather unsettling. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Linux and GPLv2
On Thu, Mar 24, 2005 at 03:38:19PM +, Andrew Suffield wrote: On Thu, Mar 24, 2005 at 03:10:41AM -0500, Glenn Maynard wrote: On Thu, Mar 24, 2005 at 07:45:24AM +, Andrew Suffield wrote: Fair use is an American perversion. It does not exist in most of the rest of the world in anything like the same form. Anything that relies on the American notion of fair use is non-free, because in the UK that means Non-commercial use only. Out of curiosity, what is it about fair use that makes you call it a perversion? In opposition to the norm. That's the real definition, once you discard the arbitrary labels of 'right' and 'wrong'. Something which is merely in opposition to the norm is unusual. A perversion is something which is both unusual and worse. Fair use may be unusual, but I don't really understand how it's worse. -- Glenn Maynard -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Re: Linux and GPLv2
On Wed, Mar 23, 2005 at 10:34:28AM +0100, Gerardo Ballabio wrote: So, the FTP plugin's license must be _the GPL_, and it must not depend on GPL-incompatible code. No. It must be GPL-compatible, but it need not be the GPL. For example, you can reuse a third party's MIT-licensed code in your code, and that software does not suddenly receive the restrictions of the GPL--in fact, if you havn't modified the code enough to have a copyright claim in it (which you often don't have to do, with well-written code), you don't have any grounds to be placing the GPL's restrictions on that code at all. -- Glenn Maynard -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Linux and GPLv2
Matthew Palmer wrote: That said, it looks questionable whether the FTP plugin should reallybe considered a derivative of the plugin loader. If the latter has a documented API and the former only communicates with it through that API, I'd probably say no. Even more so if that plugin could conceivably work with another, non-GPL'd plugin loader. It's a tricky issue. Even if the plugin does only communicate via the published interface, it is entirely possible that the plugin includes copyrighted elements from the plugin loader code itself. It'd have to be decided on a case-by-case basis. Basically, .h bits are *not* copyrightable. Which other elements of the plugin loader may be _included_ in the plugin? - Matt Massa -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Linux and GPLv2
Matthew Palmer wrote: On Wed, Mar 23, 2005 at 10:22:12AM -0300, Humberto Massa wrote: Matthew Palmer wrote: That said, it looks questionable whether the FTP plugin should reallybe considered a derivative of the plugin loader. If the latter has a documented API and the former only communicates with it through that API, I'd probably say no. Even more so if that plugin could conceivably work with another, non-GPL'd plugin loader. It's a tricky issue. Even if the plugin does only communicate via the published interface, it is entirely possible that the plugin includes copyrighted elements from the plugin loader code itself. It'd have to be decided on a case-by-case basis. Basically, .h bits are *not* copyrightable. Under what theory do you come to that conclusion? Note that a .h file can contain more than function prototypes, and function prototypes don't have to be in a .h file. Whoa, slow down, cowboy! Re-read what I have written up there: .h _bits_ are not copyrightable. Now take a deep breath. The thing is: it is considered by USofA and other countries case law that the bits that are at compile/link time from a .h file (as you mentioned down here, as macros and inline functions) are not really being included in the work, but are in reality being referenced on it. I will try to ilustrate it (any coincidence on names of real people is what it seems to be): // errno.h: // (C) LT, 1991 #define ENOENT 23 extern int errno; /* implementation detail: never use this array; its name can change at any time */ extern char **__err_msgs; #define perror(s) (fprintf(stderr,%d:%s:%s\n,errno,__err_msgs[errno])) // myfile.c: // (C) Massa, 2005 if( !(f = fopen(arqname.txt, r)) ) perror(The file arqname.txt must exist!); Is myfile.c a derivative work on errno.h? The answer is NO. In the Abstraction, Filtration, Comparison process, bits that come from a .h by way of its interface (as opposed to by way of its implementation) are filtered OUT in the filtration phase. This basically translates as following: to the extent that you don't mess with the inner workings of a .h file (eg, in the above example the __err_msgs array), independently if it has macros/inline functions in it, myfile.c (that is, in the ultimate analisys, the copyrighted work) does not include, properly, errno.h, merely reference it. IOW: myfile.c is not a derivative work on errno.h. Now, look: // myerrno.h: // (C) LT, 1991 // (C) modifications Massa, 2005 #define ENOENT 24 /* the stupid other guy used the wrong number */ extern int errno; extern int perror(const char *s); /* let's move this to the lib */ THAT is Obviously a derivative work of errno.h. Got the difference? This example have something that is a *transformation* -- novel, intellectually-worked -- on the original work. When you abstract out what errno.h does, filter the non-copyrightable parts (if any) and compare, myerrno.h is clearly a derivative work. Even if kernel developers did not know it at the time, this is the real power behind EXPORT_SYMBOL_GPL vs EXPORT_SYMBOL: the latter mark things in the kernel as being part of the API/ABI and the former, as being part of the implementation, *effectively* regulating what is messing with the kernel inner workings enough to be considered a derivative work (and hence, to be mandatorily GPL-compatible-licensed). Which other elements of the plugin loader may be _included_ in the plugin? Macros and inline functions spring immediately to mind, although I don't think inlines normally cross library boundaries. My linker fu is rusty. Any others? - Matt Massa -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Linux and GPLv2
Scripsit Humberto Massa [EMAIL PROTECTED] Matthew Palmer wrote: Basically, .h bits are *not* copyrightable. Under what theory do you come to that conclusion? Note that a .h file can contain more than function prototypes, and function prototypes don't have to be in a .h file. Whoa, slow down, cowboy! Re-read what I have written up there: .h _bits_ are not copyrightable. Now take a deep breath. Deep breath taken. I still want to know why you think bits are treated differently by copyright just because they happen to be in a file whose name ends in .h. Well, of course bits in general are not copyrightable. The digits 0 and 1 are everybody's property. A particular sequence of bits can, however, form a copy of a copyrightable work. The thing is: it is considered by USofA and other countries case law that the bits that are at compile/link time from a .h file (as you mentioned down here, as macros and inline functions) are not really being included in the work, but are in reality being referenced on it. Inline functions are certainly being included in the machine code that comes out of the compiler, at least if they are called by the rest of the compilation unit. extern char **__err_msgs; #define perror(s) (fprintf(stderr,%d:%s:%s\n,errno,__err_msgs[errno])) Is myfile.c a derivative work on errno.h? The answer is NO. Of course. But myfile.o might have been if perror() were complex enough to leave any room for expressive choice. In the Abstraction, Filtration, Comparison process, bits that come from a .h by way of its interface (as opposed to by way of its implementation) are filtered OUT in the filtration phase. The compiler does not even know which bits in it input come from .h file and which come from a .c file. It has no means of filtering on them specifically. (Well, excluding #line markers, but they should *not* influence the machine code being generated). -- Henning Makholm En tapper tinsoldat. En dame i spagat. Du er en lykkelig mand ... -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Linux and GPLv2
On Wed, Mar 23, 2005 at 10:56:49AM -0300, Humberto Massa wrote: Whoa, slow down, cowboy! Re-read what I have written up there: .h _bits_ are not copyrightable. Now take a deep breath. The thing is: it is considered by USofA and other countries case law that the bits that are at compile/link time from a .h file (as you mentioned down here, as macros and inline functions) are not really being included in the work, but are in reality being referenced on it. Even assuming that this considered has some legal basis, this rule utterly misses the point. You have a decent heuristic there, but it's just a heuristic -- it doesn't mean anything legally. In the U.S., derivative works don't need to include a literal copy of any of the original to be derivative works. All they need to do is include more creative content than fair use. See circular 14 (http://www.copyright.gov/circs/circ14.pdf) for more detail on what is and is not a derivative work. In particular, note that it uses the phrase based on 11 times. See circular 21 (http://www.copyright.gov/circs/circ21.pdf) for more detail on fair use. Outside the free software community, copyright infringment cases frequently involve works with many superficial differences. Just because we're usually dealing with relatively unambiguous questions doesn't mean that's always the case for copyright law. Which, perhaps, is one of the reasons proving financial harm is one of the key issues in most copyright infringement cases. As another heuristic, that's when use isn't fair use. -- Raul -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Linux and GPLv2
Raul Miller wrote: Even assuming that this considered has some legal basis, this rule utterly misses the point. You have a decent heuristic there, but it's just a heuristic -- it doesn't mean anything legally. Yes and no. Every legal issue is judged (by an attorney, before be definitively judged by a Judge of Law) with basis on heuristics. In Brasil, the most important heuristic is the letter of law and its possible hermeneutics (case law has a very small weight here, compared to the USofA, for instance). In the USofA, the most important heuristic is the former intepretation given to some law (case law). In the U.S., derivative works don't need to include a literal copy of any of the original to be derivative works. All they need to do is include more creative content than fair use. But it is well-established in the US that using/replicating/implementing APIs and ABIs are fair use. And that is the point I am discussing here. See circular 14 (http://www.copyright.gov/circs/circ14.pdf) for more detail on what is and is not a derivative work. In particular, note that it uses the phrase based on 11 times. See circular 21 (http://www.copyright.gov/circs/circ21.pdf) for more detail on fair use. Again, using an API to make a program is not making a program based on another work -- and this, too, is well established in USofAn case law. Outside the free software community, copyright infringment cases frequently involve works with many superficial differences. Just because we're usually dealing with relatively unambiguous questions doesn't mean that's always the case for copyright law. Which, perhaps, is one of the reasons proving financial harm is one of the key issues in most copyright infringement cases. As another heuristic, that's when use isn't fair use. Agreed. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Linux and GPLv2
Scripsit Humberto Massa [EMAIL PROTECTED] Henning Makholm wrote: Deep breath taken. I still want to know why you think bits are treated differently by copyright just because they happen to be in a file whose name ends in .h. Well, this is kind of easy to explain. Snip explanation that does not do anything the idea that bits are treated differently by copyright just becuase they are in a file called .h. Inline functions are certainly being included in the machine code that comes out of the compiler, at least if they are called by the rest of the compilation unit. They are not being included by the author, in a creative -- intelectually-novel -- process; they are being included by the compiler/linker in an automated/automatable process. So what? They are being reproduced, and the mechanicalness of the reproduction does not prevent copyright from applying to the result. Nothing that comes out of an automated process is *per-se* copyrightable. Do you still think this applies if the automated process is an offset press? Obviously, something that comes out of an automated process and is equivalent (repeatable result of processing of) a copyrightable work is, to the eyes of the law, the unprocessed work itself No, it's a processed work. Which is still coverved by the copyright of the original work. The compiler does not even know which bits in it input come from .h file and which come from a .c file. It has no means of filtering on them specifically. (Well, excluding #line markers, but they should *not* influence the machine code being generated). Looking from a legal standpoint, the compiler and linker are tools like the binding machine in a book factory: they just transform and stitch together things (imagine the manufacturing of an anthology book) but they realize no transformation on the work itself. They do not affect copyrights. Exactly. The copyright of the original function definition is *not* affected by its having been placed in a .h file somewhen in the process. Concluding: when you write a .c file, it is or not a derivative work on another original work independently of what the compiler and linker will do in the future. I repeat: No, but the resulting .o file may be derived from another work that the compiler also read while producing it. The output of a compiler/linker is related to the source code as the ready-to-be-sold-in-the-bookstore book is related to the original, typewritten, version of one of the novelettes inside the book. Yes. And if I want to reprint the entire book it may well be that I need the permission not only of the author byt also of the guy who wrote the preface. Trying to explain more: my myfile.c is not a derivative work on errno.h, No, but myfile.o may be. (I feel like I'm repeating myself here). Now, just to make my self more clear, when you do a derivative work, then you are transforming something -- akin to Peter Jackson transforming the works of Tolkien. You had something -- a series of books -- and you got other thing in the other extremity -- a series of screenplays, plus casting instructions, storyboards, etc. The second series of stuff (derivative) came out of the spirit of Peter Jackson, but resulted of transformation (novel, intellectual) of the works of Tolkien. I do not see how that has anything to do with the supposed magical copyright-evading capabilities of filenames that end in .h. -- Henning Makholm No one seems to know what distinguishes a bell from a whistle. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Linux and GPLv2
Henning Makholm [EMAIL PROTECTED] writes: Concluding: when you write a .c file, it is or not a derivative work on another original work independently of what the compiler and linker will do in the future. I repeat: No, but the resulting .o file may be derived from another work that the compiler also read while producing it. The object file may contain bits from header files, or whatever, but this has no bearing on the distributability of it. They only found their way there as the result of implementation details. Allowing the a particular method of implementation to stretch the reach of copyright in such a way would be absurd, and this is what fair use is about. -- Måns Rullgård [EMAIL PROTECTED] -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Linux and GPLv2
Jeremy Hankins wrote: Trying to explain more: my myfile.c is not a derivative work on errno.h, No, but myfile.o may be. (I feel like I'm repeating myself here). My understanding is that, in practice, myfile.c could infringe as well, if the only reasonable way to use it is by creating a work that is derivative of errno.h (if, e.g., errno.h contains something that is significant and used in myfile.o). There is a lot of case law that says otherwise: using an API does not contaminate the copyrights of the user. Massa -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Linux and GPLv2
On Wed, Mar 23, 2005 at 06:41:58PM +0100, Måns Rullgård wrote: Inline functions are certainly being included in the machine code that comes out of the compiler, at least if they are called by the rest of the compilation unit. Irrelevant. If someone chooses to implement a particular interface using an inline function, that should not impact the derivedness of code using the interface. Obviously it doesn't impact whether *source code* using that interface is a derivative work, but it certainly affects whether *binary code* is. The resulting binary from compiling against a header containing inline code can contain substantial pieces of that code--almost verbatim, if it's inline assembly--and that obviously makes the binary connected to the source. (Whether it's a derived work or a combined work or an aggregate work or something else, I'm not sure--it's clearly not a creative transformation--but the result is certainly affected by the license of that code.) extern char **__err_msgs; #define perror(s) (fprintf(stderr,%d:%s:%s\n,errno,__err_msgs[errno])) Is myfile.c a derivative work on errno.h? The answer is NO. Of course. But myfile.o might have been if perror() were complex enough to leave any room for expressive choice. Again, irrelevant. If your implementation puts things in macros, that's your problem. Uh, what? If my implementation puts things in macros, and you distribute my implementation as part of your binaries as a result, that's *your* problem. I don't even know what you're trying to say here--you put your copyrighted code in a header and I copied it into my object file--that's your problem, not mine! doesn't make any sense at all. -- Glenn Maynard -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Linux and GPLv2
Glenn Maynard [EMAIL PROTECTED] writes: extern char **__err_msgs; #define perror(s) (fprintf(stderr,%d:%s:%s\n,errno,__err_msgs[errno])) Is myfile.c a derivative work on errno.h? The answer is NO. Of course. But myfile.o might have been if perror() were complex enough to leave any room for expressive choice. Again, irrelevant. If your implementation puts things in macros, that's your problem. Uh, what? If my implementation puts things in macros, and you distribute my implementation as part of your binaries as a result, that's *your* problem. I don't even know what you're trying to say here--you put your copyrighted code in a header and I copied it into my object file--that's your problem, not mine! doesn't make any sense at all. The only reasonable way to use your library (which for this discussion shall be assumed to have been legally obtained), is to compile programs using its header files, and link these programs against it. What did you expect me to do with those headers? Frame them and hang them on the wall? -- Måns Rullgård [EMAIL PROTECTED] -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Linux and GPLv2
On Wed, Mar 23, 2005 at 04:19:31PM -0300, Humberto Massa wrote: Henning Makholm wrote: Snip explanation that does not do anything the idea that bits are treated differently by copyright just becuase they are in a file called .h. Repeating: bits that are in files called .h are not copied in your work, I think you need to stop referring to .h files. There's no general rule that can be applied to files based purely on their extension. I think what you meant to say is something like files referenced by a .c file by means of #include are only mechanically copied into your work, no creative transformation takes place and therefore no derivation takes place. Would that be accurate? That having been said, your example earlier of errno.h and the internal __err_msgs array could be an interesting example. If I reference that __err_msgs array in my own code, does that then pull in errno.h in a deeper fashion, such that I have then created a derived work of errno.h? Concluding: when you write a .c file, it is or not a derivative work on another original work independently of what the compiler and linker will do in the future. I repeat: No, but the resulting .o file may be derived from another work that the compiler also read while producing it. Not derived. Never. To derive you need inteligence (in Brazilian letter-of-law, spirit). A compiler does not have it. Neither does a linker. Only a person does. So whether or not a source file is derived from a file it includes by reference is determined before you ever wave a compiler near it? Seems sensible enough, if a little tricky to determine (I guess that's why we have courts for this, to stuff it up *properly*). I do not see how that has anything to do with the supposed magical copyright-evading capabilities of filenames that end in .h. This is an artifact of your imagination... I said only that, in general, the *usual* .h does not contain copyrightable bits. And I suspect you What you actually said initially was Basically, .h bits are *not* copyrightable. Nothing in there about in general or *usual*, except what might be implied by Basically. I'm happy to believe that you merely misexpressed yourself, but on the face of it, what you wrote implies that if you put your novel in a file called foo.h, it's not copyrightable. Remember, all we have here is written word. Cultural or linguistic shorthand rarely works well on a mailing list. Something like the common contents of a .h file, being prototypes, symbol definitions, and trivial macros, are not copyrightable would have probably had the same meaning (for you) as what you did write, and would have saved all of this subsequent confusion for the rest of us. - Matt signature.asc Description: Digital signature
Re: Linux and GPLv2
On Wed, Mar 23, 2005 at 11:45:45PM +0100, Måns Rullgård wrote: Glenn Maynard [EMAIL PROTECTED] writes: extern char **__err_msgs; #define perror(s) (fprintf(stderr,%d:%s:%s\n,errno,__err_msgs[errno])) Is myfile.c a derivative work on errno.h? The answer is NO. Of course. But myfile.o might have been if perror() were complex enough to leave any room for expressive choice. Again, irrelevant. If your implementation puts things in macros, that's your problem. Uh, what? If my implementation puts things in macros, and you distribute my implementation as part of your binaries as a result, that's *your* problem. I don't even know what you're trying to say here--you put your copyrighted code in a header and I copied it into my object file--that's your problem, not mine! doesn't make any sense at all. The only reasonable way to use your library (which for this discussion shall be assumed to have been legally obtained), is to compile programs using its header files, and link these programs against it. That's fine if you're building the program for your own use -- absent a EULA prohibiting certain uses of the work, you've got no problems (since copyright law doesn't dictate use). However, if you attempt to distribute your compiled work, with my implementation bits in them, you do need to comply with the licence of my implementation in regards to your redistribution of my copyrighted work. The issue at hand is whether the compilation phase creates an anthology work (AKA mere aggregation, I believe), or a derivative work. Are you taking the position that not even aggregation takes place during compilation? - Matt signature.asc Description: Digital signature
Re: Linux and GPLv2
On Wed, Mar 23, 2005 at 08:10:57PM +0100, M?ns Rullg?rd wrote: Henning Makholm [EMAIL PROTECTED] writes: Concluding: when you write a .c file, it is or not a derivative work on another original work independently of what the compiler and linker will do in the future. I repeat: No, but the resulting .o file may be derived from another work that the compiler also read while producing it. The object file may contain bits from header files, or whatever, but this has no bearing on the distributability of it. Nonsense. Literal copying is always copyright infringement. They only found their way there as the result of implementation details. Under your rather strange theory, copying a file can never be copyright infringement, because the way cp moves the bits around is just an 'implementation detail'. So presumably you don't think copyright infringement using a computer is possible. Allowing the a particular method of implementation to stretch the reach of copyright in such a way would be absurd, and this is what fair use is about. Fair use is an American perversion. It does not exist in most of the rest of the world in anything like the same form. Anything that relies on the American notion of fair use is non-free, because in the UK that means Non-commercial use only. -- .''`. ** Debian GNU/Linux ** | Andrew Suffield : :' : http://www.debian.org/ | `. `' | `- -- | signature.asc Description: Digital signature
Re: Linux and GPLv2
On Mon, Mar 14, 2005 at 09:50:54PM +, Anthony W. Youngman wrote: And doesn't the GPL contain a promise that any future GPL will be identical in spirit to the original? Nope. It says similar in spirit, which is much weaker to me. Certainly it's also not a major stretch to claim that a license which says if you use this program as part of your webpage, you must make source available is similar in spirit, since the spirit of the GPL is making source available and reusable. (Of course, there are plenty of potential restrictions in that spirit that go too far and aren't free.) -- Glenn Maynard -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Linux and GPLv2
Anthony W. Youngman [EMAIL PROTECTED] writes: And doesn't the GPL contain a promise that any future GPL will be identical in spirit to the original? It uses the phrase similar in spirit, which has yet to be given an exact definition. Of course, this assumes you actually want to take the matter to court... an act often prohibitively expensive for most FOSS developers... but then again, most of this conversation is academic anyway because it assumes people will actually dislike v3 AND that there is infringement ABD that the infringement is authorized under v3 but not v2. If the new GPL breaks that promise, then the original licensor has a very good case in law that the new GPL is *not* a later version, but a different version to which the or later wording doesn't apply... That would be a, maybe not desirable, but at least very interesting case. -- Måns Rullgård [EMAIL PROTECTED] -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Linux and GPLv2
On Tuesday 15 March 2005 01:34 pm, Glenn Maynard wrote: Nope. It says similar in spirit, which is much weaker to me. Certainly it's also not a major stretch to claim that a license which says if you use this program as part of your webpage, you must make source available is similar in spirit, since the spirit of the GPL is making source available and reusable. (Of course, there are plenty of potential restrictions in that spirit that go too far and aren't free.) Such languages is not very useful anyway. Against whom does this language apply? The FSF?! Doubtful, they are not participants in the license, nor beneficiaries in the contract (note that I am politely avoiding the important legal ambiguity as to whether this is a license or a contract). The phrase is a very nice pledge of intent, but its not going to be enforceable in a court of law. -Sean -- Sean Kellogg 2nd Year - University of Washington School of Law GPSS Senator - Student Bar Association Editor-at-Large - National ACS Blog [http://www.acsblog.org] c: 206.498.8207 e: [EMAIL PROTECTED] So, let go ...Jump in ...Oh well, what you waiting for? ...it's all right ...'Cause there's beauty in the breakdown
Re: Linux and GPLv2
Arnoud Engelfriet wrote: Interesting point. But the statement would apply certainly to Linus' own contributions. And that would preclude distribution of anything containing those contributions under anything but GPLv2 I think. But if you can take out his code (and any other that's GPLv2 only), you'd be free to apply GPLv3 if and when it comes out. Arnoud Sorry, but no, no, no. Everything that is not completely independent and extractable and beyond any doubt non-historically-derived of Linus code is a derivative work and, as such, can only be distributed under the terms of the GPLv2. To prove something not derivative, you would have to show that historically, it was made for other kernel, and that there is no tranformation of the linux kernel that resulted in that something. There *is* at least one example in the tree: the ppp code is derivative from one of the BSDs. So, you could take ppp and distribute under the BSD but not a lot else. HTH, Massa -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Linux and GPLv2
Glenn Maynard [EMAIL PROTECTED] writes: Such language is fine as long as the copyright holder and the license author are the same entity. New versions of the license are likely to reflect changes in the opinions of the authors, and they have every right to make provisions for a modified license to automatically apply to already released works. The danger arises when people start out-sourcing the writing of licenses to third parties. The FSF has its own agenda, and has little reason to consider the opinions of others, who just happen to use their license texts, when writing the next version. A couple years ago, this would all have been patently false. The FSF has a very strong interest in their notion of freedom being considered reliable, and having the community trust them to remain consistent There is no single the community, sharing a single opinion on freedom. There are many different views out there, and some recent moves from FSF have been in a direction away from a large enough number of people, with loud enough voices, to make it noticeable. -- Måns Rullgård [EMAIL PROTECTED] -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Linux and GPLv2
On Sun, Mar 13, 2005 at 03:30:28PM +0100, Måns Rullgård wrote: Arnoud Engelfriet [EMAIL PROTECTED] writes: And probably it will also deal with running the code on a publicly accessible server. The question is if a license based on copyright can legally place such restrictions on use of the program. Some idea of how the FSF may attempt this can be seen from the Affero General Public License. Apparantly the Affero GPL is a modified version of the GNU GPL, it adds Section 2(d): * d) If the Program as you received it is intended to interact with users through a computer network and if, in the version you received, any user interacting with the Program was given the opportunity to request transmission to that user of the Program's complete source code, you must not remove that facility from your modified version of the Program or work based on the Program, and must offer an equivalent opportunity for all users interacting with your Program through a computer network to request immediate transmission by HTTP of the complete source code of your modified version or other derivative work. It also adds an interesting twist on the or later thing often used with the GPLv2: Affero Inc. may publish revised and/or new versions of the Affero General Public License from time to time. Such new versions will be similar in spirit to the present version, but may differ in detail to address new problems or concerns. Each version is given a distinguishing version number. If the Program specifies a version number of this License which applies to it and any later version, you have the option of following the terms and conditions either of that version or of any later version published by Affero, Inc. If the Program does not specify a version number of this License, you may choose any version ever published by Affero, Inc. You may also choose to redistribute modified versions of this program under any version of the Free Software Foundation's GNU General Public License version 3 or higher, so long as that version of the GNU GPL includes terms and conditions substantially equivalent to those of this license. So, if you wish to use the AGPL, you as copyright holder can choose between AGPLv1 and AGPLv1 or later. But whichever you choose, you cannot remove the option to 'upgrade' to GNU GPLv3. -- Kuno. (ps. this is probably my first post to this list, so.. hi! everyone :). -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Linux and GPLv2
Humberto Massa [EMAIL PROTECTED] writes: Arnoud Engelfriet wrote: Interesting point. But the statement would apply certainly to Linus' own contributions. And that would preclude distribution of anything containing those contributions under anything but GPLv2 I think. But if you can take out his code (and any other that's GPLv2 only), you'd be free to apply GPLv3 if and when it comes out. Arnoud Sorry, but no, no, no. Everything that is not completely independent and extractable and beyond any doubt non-historically-derived of Linus code is a derivative work and, as such, can only be distributed under the terms of the GPLv2. To prove something not derivative, you would have to show that historically, it was made for other kernel, and that there is no tranformation of the linux kernel that resulted in that something. There *is* at least one example in the tree: the ppp code is derivative from one of the BSDs. Some of the filesystems (XFS and JFS, at least) have external origins, although they must have been somewhat adapted to the Linux VFS layer. -- Måns Rullgård [EMAIL PROTECTED] -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Linux and GPLv2
Kuno Woudt [EMAIL PROTECTED] writes: On Sun, Mar 13, 2005 at 03:30:28PM +0100, Måns Rullgård wrote: Arnoud Engelfriet [EMAIL PROTECTED] writes: And probably it will also deal with running the code on a publicly accessible server. The question is if a license based on copyright can legally place such restrictions on use of the program. Some idea of how the FSF may attempt this can be seen from the Affero General Public License. Apparantly the Affero GPL is a modified version of the GNU GPL, it adds Section 2(d): * d) If the Program as you received it is intended to interact with users through a computer network and if, in the version you received, any user interacting with the Program was given the opportunity to request transmission to that user of the Program's complete source code, you must not remove that facility from your modified version of the Program or work based on the Program, and must offer an equivalent opportunity for all users interacting with your Program through a computer network to request immediate transmission by HTTP of the complete source code of your modified version or other derivative work. This appears to only apply to self-distributing programs. If the program does not have a send-the-source function, I don't see any requirement that source be provided to users of a service based on the program. It also adds an interesting twist on the or later thing often used with the GPLv2: Affero Inc. may publish revised and/or new versions of the Affero General Public License from time to time. Such new versions will be similar in spirit to the present version, but may differ in detail to address new problems or concerns. I've always wondered what similar in spirit is supposed to mean. AFAIK, that phrase has no established legal interpretation. Each version is given a distinguishing version number. If the Program specifies a version number of this License which applies to it and any later version, you have the option of following the terms and conditions either of that version or of any later version published by Affero, Inc. If the Program does not specify a version number of this License, you may choose any version ever published by Affero, Inc. This looks similar to the language used in the GNU GPL. You may also choose to redistribute modified versions of this program under any version of the Free Software Foundation's GNU General Public License version 3 or higher, so long as that version of the GNU GPL includes terms and conditions substantially equivalent to those of this license. It would be interesting to see the reaction of these people, if the GNU GPLv3 does not include a source-for-service clause, after all. -- Måns Rullgård [EMAIL PROTECTED] -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Linux and GPLv2
Humberto Massa wrote: Everything that is not completely independent and extractable and beyond any doubt non-historically-derived of Linus code is a derivative work and, as such, can only be distributed under the terms of the GPLv2. You're correct in that anything that's a derivative work of any GPLv2 code also cannot be distributed under GPLv3 or later. But it's going to be very interesting to figure out what code is a derivative work of what. Anyway, this seems rather theoretical. Arnoud -- Arnoud Engelfriet, Dutch patent attorney - Speaking only for myself Patents, copyright and IPR explained for techies: http://www.iusmentis.com/ -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Linux and GPLv2
Arnoud Engelfriet wrote: You're correct in that anything that's a derivative work of any GPLv2 code also cannot be distributed under GPLv3 or later. But it's going to be very interesting to figure out what code is a derivative work of what. Anyway, this seems rather theoretical. Arnoud Yeah, but in a back-of-envelop calculation, to completely determine what code is derivative of what in Linux, one would take approximately 19 man-years. Maybe some automated way to study cvs.kernel.org changesets would shorten it up a bit, but ... ;-) Massa -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Linux and GPLv2
Kuno Woudt wrote: * d) If the Program as you received it is intended to interact with users through a computer network and if, in the version you received, any user interacting with the Program was given the opportunity to request transmission to that user of the Program's complete source code, you must not remove that facility from your modified version of the Program or work based on the Program, and must offer an equivalent opportunity for all users interacting with your Program through a computer network to request immediate transmission by HTTP of the complete source code of your modified version or other derivative work. Incidentally, this is pretty badly drafted, IMO. For example: - The requirement to not remove certain particular code is probably non-free; - The general requirement to make code available for download could be asserted without the don't remove code X clause; - Specifying HTTP is not future-proof, and may not be appropriate for some programs or environments; - What happens if the program interacts with other programs over a network? How do you define interacting with a user? Gerv -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Linux and GPLv2
On Sun, 13 Mar 2005 18:29:36 -0500 Glenn Maynard wrote: There's a more significant problem: if you say or later, and you don't like GPLv3--either because it allows things you don't like, or (as recent events suggest may be more likely) includes restrictions you don't like, people can take your work, modify it, and place their modifications under GPLv3-only. If you want to keep your code available under GPLv2, you can't incorporate those changes, since they're not available under v2. Considering that a primary motivator of the GPL is to prevent the case where you can't incorporate other people's enhancements to your work due to licensing, this seems like a potentially major failure. Indeed. [...] The or later gives the FSF more flexibility to change the license terms for a vast amount of software they really have no connection at all with, with or without the approval of the copyright holders of said software. The or later is intended, as I understand--from rational logic, as I don't believe I've seen any statement from the FSF--to allow the FSF to fix problems in the GPL. You've seen it, believe me! It's in the GPLv2 text itself! ;-) | 9. The Free Software Foundation may publish revised and/or new | versions of the General Public License from time to time. Such new | versions will be similar in spirit to the present version, but may | differ in detail to address new problems or concerns. ^^^ Without or later, it's impossible: the only way a bugfixed GPL could be used is after getting explicit permission from every copyright holder of a GPL work. Further, and just as importantly, the bugfixed GPLv3 would be incompatible with the original GPLv2! That would fragment the GPL at a fundamental level. Yes, at level that only dual licensing (which is what GPLv2 or later actually is) can cure... That would be fine, if the FSF had maintained its traditional consistency. Frankly, they broke that trust with the GFDL, and so I'd sympathise with people no longer willing to use the or later language. The above problem doesn't go away, though. Agreed entirely. The or later trick works as long as the FSF is trusted to actually fix problems and release better and better GPL versions. But I'm not sure that this trust is deserved any longer... :-( and :-(( and then :-((( -- Today is the tomorrow you worried about yesterday. .. Francesco Poli GnuPG Key ID = DD6DFCF4 Key fingerprint = C979 F34B 27CE 5CD8 DC12 31B5 78F4 279B DD6D FCF4 pgpJdE7XjSW0A.pgp Description: PGP signature
Re: Linux and GPLv2
On Mon, 14 Mar 2005 17:53:35 + Gervase Markham wrote: [about the don't remove get_source feature] - The requirement to not remove certain particular code is probably non-free; I don't think it's forbidding to remove the code: it's merely forbidding to drop a feature. You could reimplement it in a better (or even worse) way, but you must support it. Anyway I agree it's non-free. Suppose for example that my derivative work is intended for use on an embedded system with very limited hardware resources. I could fail to comply with my constraints, due to this prohibition to drop a functionality (in the meanwhile, perhaps, I'm distributing my derivative work separately, through my website, in both source and binary forms and even through Debian archives, since I'm a DD myself and have packaged my derivative work for Debian! Thus I'm a very good guy!). Obviously, this is a thoroughly hypothetical example (I don't write programs for embedded systems, IANADD, and, above all things, I wouldn't create derivative works of AGPL'd programs!) - The general requirement to make code available for download could be asserted without the don't remove code X clause; Yes, though I don't think such clauses could be made DFSG-free... :( - Specifying HTTP is not future-proof, and may not be appropriate for some programs or environments; Definitely agreed. - What happens if the program interacts with other programs over a network? How do you define interacting with a user? Who knows? I agree that this is very gray and unclear. -- Today is the tomorrow you worried about yesterday. .. Francesco Poli GnuPG Key ID = DD6DFCF4 Key fingerprint = C979 F34B 27CE 5CD8 DC12 31B5 78F4 279B DD6D FCF4 pgproudW5yfWV.pgp Description: PGP signature
Re: Linux and GPLv2
Francesco Poli [EMAIL PROTECTED] writes: On Sun, 13 Mar 2005 19:14:09 -0500 Glenn Maynard wrote: [...] There havn't been any such bugs, though, fortunately. Some people don't like the PHP loophole or whatever you want to call it, but I don't believe that's fixable in a free license, Could you please elaborate on the PHP loophole? I've never heard of it: what do you mean by that? (feel free to change the subject or even to reply to me in private, if you think it's better) I'm also curious about this one. -- Måns Rullgård [EMAIL PROTECTED] -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Linux and GPLv2
Francesco Poli [EMAIL PROTECTED] writes: Could you please elaborate on the PHP loophole? I've never heard of it: what do you mean by that? It's the whole web-as-platform idea. Let's say I take a copy of latex (assuming for the moment that it were GPL, which it isn't IIRC), and I enhance it (maybe I integrate it with word somehow, under NDA) so that it is no longer remotely compatible with the standard latex, and then set up a web site offering a document processing service for a fee. I never distribute the program in binary form, so I never have to distribute code either. I'm therefore able to take advantage of all the work the latex tex folks have done without contributing my own changes back to the community -- or to the users of the software. A valid concern, arguably, even if it does hinge on certain ideas about how the computing field will evolve that I doubt will turn out to be accurate. But the only licenses we've seen so far that deal with this problem (if it is a problem) give up too much freedom in exchange. At least, IMHO. -- Jeremy Hankins [EMAIL PROTECTED] PGP fingerprint: 748F 4D16 538E 75D6 8333 9E10 D212 B5ED 37D0 0A03 -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Linux and GPLv2
On Mon, Mar 14, 2005 at 05:53:35PM +, Gervase Markham wrote: Kuno Woudt wrote: * d) If the Program as you received it is intended to interact with users through a computer network and if, in the version you received, any user interacting with the Program was given the opportunity to request transmission to that user of the Program's complete source code, you must not remove that facility from your modified version of the Program or work based on the Program, and must offer an equivalent opportunity for all users interacting with your Program through a computer network to request immediate transmission by HTTP of the complete source code of your modified version or other derivative work. Incidentally, this is pretty badly drafted, IMO. For example: Add another one -- must offer an [...] opportunity for all users [...] to request immediate transmission by HTTP -- doesn't mean that the request must be successfully honoured... grin - Matt signature.asc Description: Digital signature
Re: Linux and GPLv2
On Mon, 14 Mar 2005, Jeremy Hankins wrote: Francesco Poli [EMAIL PROTECTED] writes: Could you please elaborate on the PHP loophole? I've never heard of it: what do you mean by that? It's the whole web-as-platform idea. This is commonly refered to as the ASP[1] loophole not the PHP loophole for the obvious reasons that the former describes the actual problem, whereas the latter is just a language that isn't restricted to usage by ASPs. Search for affero and asp loophole from somewhere around 2003 on -legal if you want more information on why closing this loophole is probably not possible to do in a free manner. Don Armstrong 1: Where ASP is application service provider. -- The solution to a problem changes the problem. -- Peer's Law http://www.donarmstrong.com http://rzlab.ucr.edu -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Linux and GPLv2
On Mon, Mar 14, 2005 at 02:44:02PM +0100, Måns Rullgård wrote: There is no single the community, sharing a single opinion on freedom. There are many different views out there, and some recent moves from FSF have been in a direction away from a large enough number of people, with loud enough voices, to make it noticeable. Historically, the FSF had been very consistent, enough so that many people have been willing to trust the FSF to act in good faith with the will be similar in spirit to the present version of GPL#9. Given the relatively recent developments, we're in agreement, though. -- Glenn Maynard -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Linux and GPLv2
Kuno Woudt [EMAIL PROTECTED] writes: On Mon, Mar 14, 2005 at 08:00:24PM -0500, Jeremy Hankins wrote: A valid concern, arguably, even if it does hinge on certain ideas about how the computing field will evolve that I doubt will turn out to be accurate. But the only licenses we've seen so far that deal with this problem (if it is a problem) give up too much freedom in exchange. At least, IMHO. Can you be specific on which licenses you think attempt to deal with this problem? So far I only know of the Affero GPL, which I already mentioned elsewhere in this thread, and I am curious how other license authors have attempted to fix this problem. The Affero GPL is the main one. Back when it came up on this list I think there was some discussion about possibly clauses that might serve the same purpose, but I don't think we came up with anything satisfactory. The APSL tries to do this, I think, through the use of the term externally deploy. I think it does a somewhat better job than the Affero license does, but is still subject to a lot of confusion about what sorts of things count as providing a service (which is part of the externally deploy definition). It's also not clear where it gets the legal authority to place restrictions on providing services that it does. Possibly by claiming that it would be a public performance. You can see the discussion in the archives, if you're interested. It was in August of '03 came up again in June of '04. -- Jeremy Hankins [EMAIL PROTECTED] PGP fingerprint: 748F 4D16 538E 75D6 8333 9E10 D212 B5ED 37D0 0A03 -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Linux and GPLv2
Kuno Woudt wrote: Can you be specific on which licenses you think attempt to deal with this problem? So far I only know of the Affero GPL, which I already mentioned elsewhere in this thread, and I am curious how other license authors have attempted to fix this problem. Larry Rosen's Open Software License also tries to cover this. In article 5 it defines External Deployment as follows: 5. External Deployment. The term External Deployment means the use or distribution of the Original Work or Derivative Works in any way such that the Original Work or Derivative Works may be accessed or used by anyone other than You, whether the Original Work or Derivative Works are distributed to those persons or made available as an application intended for use over a computer network. This quite clearly is intended to cover usage on a website by an ASP. And then article 5 goes on to say: 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. Article 1(c) is the you may distribute, but derivatives must be OSL as well article. Arnoud -- Arnoud Engelfriet, Dutch patent attorney - Speaking only for myself Patents, copyright and IPR explained for techies: http://www.iusmentis.com/ -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Linux and GPLv2
On Mar 15, 2005, at 03:24, Kuno Woudt wrote: On Mon, Mar 14, 2005 at 08:00:24PM -0500, Jeremy Hankins wrote: A valid concern, arguably, even if it does hinge on certain ideas about how the computing field will evolve that I doubt will turn out to be accurate. But the only licenses we've seen so far that deal with this problem (if it is a problem) give up too much freedom in exchange. At least, IMHO. Can you be specific on which licenses you think attempt to deal with this problem? The license of POV-Ray 3.0 and 3.1 (free as in beer, source available; not free in the DFSG sense) addressed this issue in the section called ONLINE OR REMOTE EXECUTION OF POV-Ray. Charging for CPU time was allowed provided that the charge for POV-Ray CPU time was the same as for other CPU time. Obscuring the fact the POV-Ray was being run was prohibited. The users had to be provided with access to the files of the POV-Ray package. -- Henri Sivonen [EMAIL PROTECTED] http://hsivonen.iki.fi/ -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Linux and GPLv2
Hello, My understanding is that Linux is distributed under the GPLv2 exclusively. That is, instead of the usual GPL version 2 or later it just says GPL version 2. Given the vast number of Linux contributors, this means that Linux won't be able to migrate to the GPLv3 when it comes out, correct? Cheers, -- Daniel Carrera | I don't want it perfect, Join OOoAuthors today! | I want it Tuesday. http://oooauthors.org | -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Linux and GPLv2
Daniel Carrera [EMAIL PROTECTED] writes: Hello, My understanding is that Linux is distributed under the GPLv2 exclusively. That is, instead of the usual GPL version 2 or later it just says GPL version 2. That's what it says, yes. People occasionally question the validity of that, though, arguing that the statement was added after many people had already made contributions. Given the vast number of Linux contributors, this means that Linux won't be able to migrate to the GPLv3 when it comes out, correct? That would be the case. Is this a problem? Personally, I'd be very sceptical about releasing code under a license containing a blanket permission to use it under another yet to be written license. What if I don't at all agree with GPLv3? -- Måns Rullgård [EMAIL PROTECTED] -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Linux and GPLv2
On Sun, Mar 13, 2005 at 02:09:10PM +0100, M?ns Rullg?rd wrote: Daniel Carrera [EMAIL PROTECTED] writes: Given the vast number of Linux contributors, this means that Linux won't be able to migrate to the GPLv3 when it comes out, correct? That would be the case. Is this a problem? Personally, I'd be very sceptical about releasing code under a license containing a blanket permission to use it under another yet to be written license. What if I don't at all agree with GPLv3? The theory goes, apparently, that if you don't like the GPLv3 you can simply remove the 'or later' from all copies you distribute, and you're effectively exercising your granted right under version 2 or, at your option, any later version. I'm a little skeptical about how well this is or isn't going to work. I fear a lot of unpleasant forking action when the GPLv3 comes out, if it contains significant language changes along the lines of the GFDL (which I believe it will, from statements I've read from RMS and Hasn Reiser, amongst others), between people who decide to go v2 only and those who like v3 and will go v2 or later or even v3 or later, which will effectively produce licence-incompatible forks. - Matt signature.asc Description: Digital signature
Re: Linux and GPLv2
Måns Rullgård wrote: Given the vast number of Linux contributors, this means that Linux won't be able to migrate to the GPLv3 when it comes out, correct? That would be the case. Is this a problem? For a large colaborative project, possibly. Using only the GPLv2 means you are trapped in that license. Having an or later allows some measure of adaptability. Suppose that there is a good reason why the GPLv2 needs to be updated (e.g. to deal with software patents). Then you would like to have the choice of moving to the GPLv3 if you want. What if I don't at all agree with GPLv3? Well, then it means you gave people more freedoms than you intended. You can still make a GPLv2 fork and make all subsequent releases GPLv2 only. The point is, the or later gives you more flexibility and choice. I think it's a prudent precaution. Cheers, -- Daniel Carrera | I don't want it perfect, Join OOoAuthors today! | I want it Tuesday. http://oooauthors.org |
Re: Linux and GPLv2
M?ns Rullg?rd wrote: Daniel Carrera [EMAIL PROTECTED] writes: My understanding is that Linux is distributed under the GPLv2 exclusively. That is, instead of the usual GPL version 2 or later it just says GPL version 2. That's what it says, yes. People occasionally question the validity of that, though, arguing that the statement was added after many people had already made contributions. Interesting point. But the statement would apply certainly to Linus' own contributions. And that would preclude distribution of anything containing those contributions under anything but GPLv2 I think. But if you can take out his code (and any other that's GPLv2 only), you'd be free to apply GPLv3 if and when it comes out. Arnoud -- Arnoud Engelfriet, Dutch patent attorney - Speaking only for myself Patents, copyright and IPR explained for techies: http://www.iusmentis.com/ -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Linux and GPLv2
Arnoud Engelfriet wrote: I'm interested in why you think it's not. Wow, hey. I was just asking a question. I didn't know there was an issue here. I certainly haven't thought about it half as much as you have. Cheers, -- Daniel Carrera | I don't want it perfect, Join OOoAuthors today! | I want it Tuesday. http://oooauthors.org | -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Linux and GPLv2
On Sun, Mar 13, 2005 at 08:31:38AM -0500, Daniel Carrera wrote: Henning Makholm wrote: Yes, probably. (Which, if the signals we've been getting from FSF the last few years are to be trusted, does not strike me as a bad thing at all). This issue is new to me. What are those signals? What are you talking about? Do you have a URL that might help me get up to speed? I'm too tired to dig up the exact reference, but in a large heated discussion between Hans Reiser and many other people on d-devel last year (or maybe the year before) about removing or obscuring credits in mkreiserfs, Hans Reiser stated that he had information from RMS that there would be some sort of invariant section-like clause in GPLv3. Earlier than that, in a thread here on d-legal regarding the GFDL, RMS himself made a few sideways comments regarding the content of the GPLv3. More generally, the direction that the FSF appears to have been moving in the last few years has, in some peoples' eyes, been diverging somewhat from the Debianistic view of freedom. - Matt signature.asc Description: Digital signature
Re: Linux and GPLv2
Daniel Carrera [EMAIL PROTECTED] writes: Måns Rullgård wrote: Given the vast number of Linux contributors, this means that Linux won't be able to migrate to the GPLv3 when it comes out, correct? That would be the case. Is this a problem? For a large colaborative project, possibly. Using only the GPLv2 means you are trapped in that license. Having an or later allows some measure of adaptability. Suppose that there is a good reason why the GPLv2 needs to be updated (e.g. to deal with software patents). Then you would like to have the choice of moving to the GPLv3 if you want. We have to consider the possibility that GPLv3 will say something we don not want. Then we do not want people distributing it under those terms. Never give permission to do something you do not know what it is. What if I don't at all agree with GPLv3? Well, then it means you gave people more freedoms than you intended. You can still make a GPLv2 fork and make all subsequent releases GPLv2 only. Only if all the copyright holders agree. Suppose A has accepted contributions from B, with the or later option, and it turns out that A does not approve of v3. Now B refuses to drop this, so A is effectively forced to distribute his code under a license he does not approve of. The point is, the or later gives you more flexibility and choice. I think it's a prudent precaution. The or later gives the FSF more flexibility to change the license terms for a vast amount of software they really have no connection at all with, with or without the approval of the copyright holders of said software. I'd be very cautious about placing my code at the hands of a third party in such a manner, and I think it is unfortunate that so many authors release code under the GPL (with or later option), without properly considering the implications. -- Måns Rullgård [EMAIL PROTECTED] -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Linux and GPLv2
Arnoud Engelfriet [EMAIL PROTECTED] writes: And probably it will also deal with running the code on a publicly accessible server. The question is if a license based on copyright can legally place such restrictions on use of the program. -- Måns Rullgård [EMAIL PROTECTED] -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Linux and GPLv2
Måns Rullgård wrote: Well, then it means you gave people more freedoms than you intended. You can still make a GPLv2 fork and make all subsequent releases GPLv2 only. Only if all the copyright holders agree. Suppose A has accepted contributions from B, with the or later option, and it turns out that A does not approve of v3. Now B refuses to drop this, so A is effectively forced to distribute his code under a license he does not approve of. No, if you get code as GPLv2 or later, you can pick either GPLv2 or pick something later. In your example, A can say I pick GPLv2 and make a GPLv2 fork. It's like dual licensing. At least, that's what the Debian FAQ says :-) I'd be very cautious about placing my code at the hands of a third party in such a manner, and I think it is unfortunate that so many authors release code under the GPL (with or later option), without properly considering the implications. I guess it comes down to whether those copyright holders trust the FSF to not totally screw it up. I'd think they would have to do something really bad with the GPLv3 for this to be a problem. Consider an extreme case. Suppose GPLv3 is non-free, propietary. That means that your GPLv2 or later work is now dual licensed: GPLv2/proprietary But that is still free. It's like MySQL for example (GPL/proprietary). As long as the GPLv2 is an option, the work is free. Or am I just confused? Cheers, -- Daniel Carrera | I don't want it perfect, Join OOoAuthors today! | I want it Tuesday. http://oooauthors.org |
Re: Linux and GPLv2
Daniel Carrera [EMAIL PROTECTED] writes: Måns Rullgård wrote: Well, then it means you gave people more freedoms than you intended. You can still make a GPLv2 fork and make all subsequent releases GPLv2 only. Only if all the copyright holders agree. Suppose A has accepted contributions from B, with the or later option, and it turns out that A does not approve of v3. Now B refuses to drop this, so A is effectively forced to distribute his code under a license he does not approve of. No, if you get code as GPLv2 or later, you can pick either GPLv2 or pick something later. In your example, A can say I pick GPLv2 and make a GPLv2 fork. It's like dual licensing. At least, that's what the Debian FAQ says :-) You're probably right, but I still don't like the look of it. I'd be very cautious about placing my code at the hands of a third party in such a manner, and I think it is unfortunate that so many authors release code under the GPL (with or later option), without properly considering the implications. I guess it comes down to whether those copyright holders trust the FSF to not totally screw it up. I'd think they would have to do something really bad with the GPLv3 for this to be a problem. Take a look out there. Freshmeat.net lists approximately 2 projects using the GPL. How many of these authors do you think have thoroughly read and understood the GPL? I've seen several cases where the GPL has been chosen for no reason in particular, other than it's the most popular free software license, or similar. I do not claim to understand it myself either, far from it, which is one reason I do not use it for my own work. Consider an extreme case. Suppose GPLv3 is non-free, propietary. That means that your GPLv2 or later work is now dual licensed: GPLv2/proprietary But that is still free. It's like MySQL for example (GPL/proprietary). As long as the GPLv2 is an option, the work is free. You are seeing it from the licensee's point of view. There, there will indeed be little difference after the introduction of a GPLv3. From the licensors point of view, it is quite different. Suppose the GPLv3 relaxed the requirements to more resemble the BSD license. This would suddenly allow someone to take your (already released) code, and incorporate it in a proprietary program. Many have chosen the GPL for the precise reason that it disallows this. Consider next a (more likely) stricter GPLv3, with restrictions on use as well as distribution (e.g. running a public server, as has been mentioned (and debated at length)). It may have been the author's intent to allow unrestricted use of his program by anyone, and he may wish that programs based on his give the users the same rights. With this version of the GPLv3, it could also be used against the authors' will. If, one might argue, the author wishes for the terms to remain those of the GPLv2, why does he not remove the or any later version option? The answer is simple. Such a license is not compatible with the standard GPL (with the upgrade option), since it has further restrictions, compared to the version allowing a switch to a later version. One common reason to use the GPL in the first place, is precisely to be compatible with other GPL licensed software. Remember that few (none?) copyleft licenses are compatible with the GPL, be it by design or by chance. Placing your code under the GPL, is placing a large faith in the FSF not to change the license terms in a manner you might disagree with, a faith which in many case may be broken, should some of the rumored clauses end up in the final GPLv3 text. -- Måns Rullgård [EMAIL PROTECTED] -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Linux and GPLv2
On Sun, 13 Mar 2005 09:31:51 -0500 Daniel Carrera wrote: Måns Rullgård wrote: Well, then it means you gave people more freedoms than you intended. You can still make a GPLv2 fork and make all subsequent releases GPLv2 only. Only if all the copyright holders agree. Suppose A has accepted contributions from B, with the or later option, and it turns out that A does not approve of v3. Now B refuses to drop this, so A is effectively forced to distribute his code under a license he does not approve of. No, if you get code as GPLv2 or later, you can pick either GPLv2 or pick something later. In your example, A can say I pick GPLv2 and make a GPLv2 fork. It's like dual licensing. At least, that's what the Debian FAQ says :-) It seems to me that you are right. I can accept contributions under GPLv2 or later and then decide I want to accept them under GPLv2 only: once I begin to release new versions (or forks...) under GPLv2 only no one (apart from the totality of the copyright holders!) can revert back to GPLv2 or later or switch to GPLv3. But... there is a but. See below! I'd be very cautious about placing my code at the hands of a third party in such a manner, and I think it is unfortunate that so many authors release code under the GPL (with or later option), without properly considering the implications. I guess it comes down to whether those copyright holders trust the FSF to not totally screw it up. I'd think they would have to do something really bad with the GPLv3 for this to be a problem. Well, I learned the trust no 1 rule the hard way, by following the unfortunate GFDL affair... :-( [That is to say: I used to trust the FSF to always write and promote Free and good licenses, but then I learned about the issues with the GFDL and changed my mind...] Consider an extreme case. Suppose GPLv3 is non-free, propietary. That means that your GPLv2 or later work is now dual licensed: GPLv2/proprietary But that is still free. It's like MySQL for example (GPL/proprietary). As long as the GPLv2 is an option, the work is free. Proprietary means less freedoms to the public, more rights reserved for the copyright holders. So that would not be an issue, as long as the public can still choose to redistribute and/or modify the work under the GPLv2. Pains could instead arise, in the opposite case (that is if GPLv3 is *more* permissive than GPLv2). Suppose GPLv3 is a non-copyleft license (let's say GPLv3 becomes equivalent to 2-clause BSD). I really doubt FSF would do such a move, but let's assume it does for the sake of example simplicity. Maybe you chose the GPLv2 or later because you *want* copyleft and do not want anyone be permitted to proprietarize your work: bang! they pick GPLv3 and create a proprietary derivative work! You cannot prevent this for already released versions, only for future ones (by switching to GPLv2 only)... :-( Now a more concrete example. Suppose GPLv3 allows something similar to GFDL Invariant Sections to be included in the work. Maybe you chose the GPLv2 or later because you believe that a Free work should *not* include unmodifiable and unremovable parts: bang! they pick GPLv3 and create a non-free derivative work containing some such parts! You cannot prevent this for already released versions, and, if you want to incorporate some useful modifications from that derivative work back to your official version, you cannot, unless you switch to GPLv3 or later *and* include the Invariant Sections (that you will never be able to drop!). :-( To conclude, IMHO, it's a matter of trust. Release under GPLv2 or later if you trust the FSF to never screw things up with future GPL versions. Release under GPLv2 only if you don't. Of course, this is valid as long as you like GPLv2: if you don't, choose another license (but, please, choose a DFSG-free and possibly GPL-compatible one!). ;-) -- Today is the tomorrow you worried about yesterday. .. Francesco Poli GnuPG Key ID = DD6DFCF4 Key fingerprint = C979 F34B 27CE 5CD8 DC12 31B5 78F4 279B DD6D FCF4 pgpNTmud0ht6Q.pgp Description: PGP signature
Re: Linux and GPLv2
On Sun, 13 Mar 2005 14:33:36 +0100 Arnoud Engelfriet wrote: Interesting point. But the statement would apply certainly to Linus' own contributions. And that would preclude distribution of anything containing those contributions under anything but GPLv2 I think. But if you can take out his code (and any other that's GPLv2 only), you'd be free to apply GPLv3 if and when it comes out. Indeed. There are parts under * GPLv2 only * GPLv2 or later * other GPLv2-compatible licenses (such as, if I recall correctly, 3-clause BSD, X11, ...) As a consequence, Linux (as a whole) is under GPLv2 only. -- Today is the tomorrow you worried about yesterday. .. Francesco Poli GnuPG Key ID = DD6DFCF4 Key fingerprint = C979 F34B 27CE 5CD8 DC12 31B5 78F4 279B DD6D FCF4 pgpbzxK7Xa3fG.pgp Description: PGP signature
Re: Linux and GPLv2
On Sun, 13 Mar 2005 16:50:39 +0100 Måns Rullgård wrote: If, one might argue, the author wishes for the terms to remain those of the GPLv2, why does he not remove the or any later version option? The answer is simple. Such a license is not compatible with the standard GPL (with the upgrade option), since it has further restrictions, compared to the version allowing a switch to a later version. I don't think so. GPLv2 only is compatible with GPLv2 or later. You can take work W_1 under GPLv2 only and work W_2 under GPLv2 or later, combine them into derivative work W_d and distribute W_d under GPLv2 only. You received W_2 under GPLv2 or GPLv3 or ... at your option and you simply chose GPLv2 (rather than all of them!); then you combined two GPLv2-licensed works into one derivative and complied with both instances of GPLv2, by releasing W_d source code under the GPLv2 itself, not adding further restrictions, and so forth... -- Today is the tomorrow you worried about yesterday. .. Francesco Poli GnuPG Key ID = DD6DFCF4 Key fingerprint = C979 F34B 27CE 5CD8 DC12 31B5 78F4 279B DD6D FCF4 pgpAOVWedSGsI.pgp Description: PGP signature
Re: Linux and GPLv2
Francesco Poli [EMAIL PROTECTED] writes: On Sun, 13 Mar 2005 16:50:39 +0100 Måns Rullgård wrote: If, one might argue, the author wishes for the terms to remain those of the GPLv2, why does he not remove the or any later version option? The answer is simple. Such a license is not compatible with the standard GPL (with the upgrade option), since it has further restrictions, compared to the version allowing a switch to a later version. I don't think so. GPLv2 only is compatible with GPLv2 or later. You can take work W_1 under GPLv2 only and work W_2 under GPLv2 or later, combine them into derivative work W_d and distribute W_d under GPLv2 only. You received W_2 under GPLv2 or GPLv3 or ... at your option and you simply chose GPLv2 (rather than all of them!); then you combined two GPLv2-licensed works into one derivative and complied with both instances of GPLv2, by releasing W_d source code under the GPLv2 itself, not adding further restrictions, and so forth... On second thoughts, you seem to be right, fortunately. The no further restrictions must be as compared to the actual license text (the COPYING file), not the note at the top of the source files. -- Måns Rullgård [EMAIL PROTECTED] -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Linux and GPLv2
On Mon, 14 Mar 2005 00:38:21 +1100, Matthew Palmer [EMAIL PROTECTED] wrote: I'm too tired to dig up the exact reference, but in a large heated discussion between Hans Reiser and many other people on d-devel last year (or maybe the year before) about removing or obscuring credits in mkreiserfs, Hans Reiser stated that he had information from RMS that there would be some sort of invariant section-like clause in GPLv3. FWIW, RMS later posted to the list[1] stating that Hans was mistaken: Reiser's statement was inaccurate. For GPL version 3 we are considering requirements for preserving certain limited author information in the source code, and making explicit that other GPL-compatible licenses that are present on parts of the code cannot be removed from the source, but nothing beyond that. [1] http://lists.debian.org/debian-legal/2003/06/msg00173.html -- Andrew Saunders -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Linux and GPLv2
M?ns Rullg?rd wrote: If, one might argue, the author wishes for the terms to remain those of the GPLv2, why does he not remove the or any later version option? The answer is simple. Such a license is not compatible with the standard GPL (with the upgrade option), since it has further restrictions, compared to the version allowing a switch to a later version. The GPLv2 does not have an upgrade option. Authors may decide to offer a kind of dual license: one is GPLv2, another is any later version of the GPL as published by the FSF. I really don't see how I am imposing any further restrictions on the rights granted by the GPLv2 by not offering a dual license under a future GPLv3. Arnoud -- Arnoud Engelfriet, Dutch patent attorney - Speaking only for myself Patents, copyright and IPR explained for techies: http://www.iusmentis.com/ -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Linux and GPLv2
M?ns Rullg?rd wrote: Arnoud Engelfriet [EMAIL PROTECTED] writes: And probably it will also deal with running the code on a publicly accessible server. The question is if a license based on copyright can legally place such restrictions on use of the program. I've heard people speculate that this could be called a public performance of the work, like singing a song in front of an audience. And the right of public performance is in copyright law. Arnoud -- Arnoud Engelfriet, Dutch patent attorney - Speaking only for myself Patents, copyright and IPR explained for techies: http://www.iusmentis.com/ -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Linux and GPLv2
Arnoud Engelfriet [EMAIL PROTECTED] writes: M?ns Rullg?rd wrote: Arnoud Engelfriet [EMAIL PROTECTED] writes: And probably it will also deal with running the code on a publicly accessible server. The question is if a license based on copyright can legally place such restrictions on use of the program. I've heard people speculate that this could be called a public performance of the work, like singing a song in front of an audience. And the right of public performance is in copyright law. I don't think this a very good analogy. This is like saying that by playing the guitar with an audience, I am publicly performing the book I used when I learned how to play. When performing a song, the actual song, albeit somewhat altered, is what reaches the audience. When serving web pages, no information from the server software reaches the clients. Anyhow, this has been discussed at length here not long ago, so there is no need to do it over again. -- Måns Rullgård [EMAIL PROTECTED] -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Linux and GPLv2
Missing from this discussion is a rather important aspect of this license... the law. If GPL v3 comes out with provisions that are even arguablly different from GPL v2 there will be all sorts of grounds for developers to strike out the 'or later' language from all prior grants of access to their code. It is a matter of equity that is a) critical to any issue like this, and b) all too often over looked by this list. It is quite difficult for someone to agree to terms they have not seen before. More importantly, I don't see how I could possibly agree to terms propagated by a body that does not have privity in the contract (FSF). Unless you have assigned your copyright over to them (and may programs have) I don't think that language is going to be enforceable. Of course, this assumes you actually want to take the matter to court... an act often prohibitively expensive for most FOSS developers... but then again, most of this conversation is academic anyway because it assumes people will actually dislike v3 AND that there is infringement ABD that the infringement is authorized under v3 but not v2. -Sean p.s. For those waiting for my comments on treble damages and patent infringement, it is coming, I assure you... unfortunately finals have reared their ugly head and I haven't had time to write something of reasonable quality on the subject. -- Sean Kellogg 2nd Year - University of Washington School of Law GPSS Senator - Student Bar Association Editor-at-Large - National ACS Blog [http://www.acsblog.org] c: 206.498.8207 e: [EMAIL PROTECTED] So, let go ...Jump in ...Oh well, what you waiting for? ...it's all right ...'Cause there's beauty in the breakdown
Re: Linux and GPLv2
Mns Rullgrd [EMAIL PROTECTED] writes: If, one might argue, the author wishes for the terms to remain those of the GPLv2, why does he not remove the or any later version option? The answer is simple. Such a license is not compatible with the standard GPL (with the upgrade option), since it has further restrictions, compared to the version allowing a switch to a later version. That's not my understanding. The GPLv2 GPLv2 are logically distinct licenses, one can simply decide to drop the or any later version in code you distribute based on code with that clause. The no extra restrictions bit refers to the GPLv2 *or* the GPLv3 -- not both together. Of course, if someone contributes modifications under only GPLv3 (for example) you then have to make a choice about accepting that code, or retaining the GPLv2 license. But that's not a new problem. One common reason to use the GPL in the first place, is precisely to be compatible with other GPL licensed software. Remember that few (none?) copyleft licenses are compatible with the GPL, be it by design or by chance. This is less a characteristic of the GPL in particular than of licensing in general. Placing your code under the GPL, is placing a large faith in the FSF not to change the license terms in a manner you might disagree with, a faith which in many case may be broken, should some of the rumored clauses end up in the final GPLv3 text. True, and given some of the rumors I'm rather skeptical about the freeness of the GPLv3 to be. But all in all I don't think the risks are that great of using the or any later version language. Worst case scenario is that folks discover they've given more permissions than they meant to. But as for folks understanding better how they're licensing their software, I couldn't agree more. -- Jeremy Hankins [EMAIL PROTECTED] PGP fingerprint: 748F 4D16 538E 75D6 8333 9E10 D212 B5ED 37D0 0A03
Re: Linux and GPLv2
Le dimanche 13 mars 2005 14:09 +0100, Mns Rullgrd a crit : Personally, I'd be very sceptical about releasing code under a license containing a blanket permission to use it under another yet to be written license. What if I don't at all agree with GPLv3? Given that the FSF has already written and recommended blatantly non-free licenses, I generally my software under the GPL v2 only. -- .''`. Josselin Mouette/\./\ : :' : [EMAIL PROTECTED] `. `'[EMAIL PROTECTED] `- Debian GNU/Linux -- The power of freedom signature.asc Description: Ceci est une partie de message =?ISO-8859-1?Q?num=E9riquement?= =?ISO-8859-1?Q?_sign=E9e?=
Re: Linux and GPLv2
M?ns Rullg?rd wrote: Arnoud Engelfriet [EMAIL PROTECTED] writes: I've heard people speculate that this could be called a public performance of the work, like singing a song in front of an audience. And the right of public performance is in copyright law. I don't think this a very good analogy. This is like saying that by playing the guitar with an audience, I am publicly performing the book I used when I learned how to play. You're publicly performing the work on the sheet music in front of you. So under this hypothetical GPLv3, you have to hand out copies of the sheet music. At least that was my analogy. Arnoud -- Arnoud Engelfriet, Dutch patent attorney - Speaking only for myself Patents, copyright and IPR explained for techies: http://www.iusmentis.com/ -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Linux and GPLv2
Sean Kellogg [EMAIL PROTECTED] writes: Missing from this discussion is a rather important aspect of this license... the law. If GPL v3 comes out with provisions that are even arguablly different from GPL v2 there will be all sorts of grounds for developers to strike out the 'or later' language from all prior grants of access to their code. It is a matter of equity that is a) critical to any issue like this, and b) all too often over looked by this list. It is quite difficult for someone to agree to terms they have not seen before. In this case, terms that have not even been written yet, and over which the someone has no influence whatsoever. More importantly, I don't see how I could possibly agree to terms propagated by a body that does not have privity in the contract (FSF). Unless you have assigned your copyright over to them (and may programs have) I don't think that language is going to be enforceable. I'd go as far as to call it outright stupid to release something under terms yet to be decided by a third party, legal or not. Of course, this assumes you actually want to take the matter to court... an act often prohibitively expensive for most FOSS developers... but then again, most of this conversation is academic anyway because it assumes people will actually dislike v3 AND that there is infringement ABD that the infringement is authorized under v3 but not v2. Well, there are a few that dislike v2 already, or at least some of the more far-reaching interpretations of it. Seeing as v3 will attempt to extend its reach even further, I see it as inevitable that a fair amount of people will have a word or two to say about it. -- Måns Rullgård [EMAIL PROTECTED] -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Linux and GPLv2
Josselin Mouette [EMAIL PROTECTED] writes: Le dimanche 13 mars 2005 à 14:09 +0100, Måns Rullgård a écrit : Personally, I'd be very sceptical about releasing code under a license containing a blanket permission to use it under another yet to be written license. What if I don't at all agree with GPLv3? Given that the FSF has already written and recommended blatantly non-free licenses, I generally my software under the GPL v2 only. -- .''`. Josselin Mouette/\./\ : :' : [EMAIL PROTECTED] `. `'[EMAIL PROTECTED] `- Debian GNU/Linux -- The power of freedom -- Måns Rullgård [EMAIL PROTECTED] -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Linux and GPLv2
On Sunday 13 March 2005 01:21 pm, Måns Rullgård wrote: Well, there are a few that dislike v2 already, or at least some of the more far-reaching interpretations of it. Seeing as v3 will attempt to extend its reach even further, I see it as inevitable that a fair amount of people will have a word or two to say about it. Well, if you don't like the interpretation of what v2 says then take it to court. No one holds a monopoly on the right to decide what it says, and only the court reviewing the GPL can determine what its terms actually mean, and only within the jurisdiction in which it sits. Of course, you may not end up liking what the judge says :) Its actually quite a shame that there haven't been any court cases on the terms of the GPL... would make for some fascinating reading. -- Måns Rullgård [EMAIL PROTECTED] -- Sean Kellogg 2nd Year - University of Washington School of Law GPSS Senator - Student Bar Association Editor-at-Large - National ACS Blog [http://www.acsblog.org] c: 206.498.8207 e: [EMAIL PROTECTED] So, let go ...Jump in ...Oh well, what you waiting for? ...it's all right ...'Cause there's beauty in the breakdown
Re: Linux and GPLv2
Josselin Mouette [EMAIL PROTECTED] writes: Le dimanche 13 mars 2005 à 14:09 +0100, Måns Rullgård a écrit : Personally, I'd be very sceptical about releasing code under a license containing a blanket permission to use it under another yet to be written license. What if I don't at all agree with GPLv3? Given that the FSF has already written and recommended blatantly non-free licenses, I generally my software under the GPL v2 only. I usually use the MIT/X11/BSD license. I can see some situations where I might have preferred a copyleft license. However, the BSD license is simple enough that I can understand it (I think), and I don't need to worry about which libraries I link with, as long as I don't distribute binaries (which is far too cumbersome anyway). PS. Sorry for the blank mail. I slipped on the send button. -- Måns Rullgård [EMAIL PROTECTED] -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Linux and GPLv2
Arnoud Engelfriet [EMAIL PROTECTED] writes: M?ns Rullg?rd wrote: Arnoud Engelfriet [EMAIL PROTECTED] writes: I've heard people speculate that this could be called a public performance of the work, like singing a song in front of an audience. And the right of public performance is in copyright law. I don't think this a very good analogy. This is like saying that by playing the guitar with an audience, I am publicly performing the book I used when I learned how to play. You're publicly performing the work on the sheet music in front of you. So under this hypothetical GPLv3, you have to hand out copies of the sheet music. At least that was my analogy. The music sheets would correspond to the web pages, not the web server software. -- Måns Rullgård [EMAIL PROTECTED] -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]