Re: [Python-Dev] VC 7.1 maintenance?
On 16 Jun 2004, at 22:46, Tim Peters wrote:
Does this mean that the final answer to the question of whether VC7 was GPL-compatible was "yes"? -- Jack Jansen, <Jack.Jansen@cwi.nl>, http://www.cwi.nl/~jack If I can't dance I don't want to be part of your revolution -- Emma Goldman
[Jack Jansen]
Does this mean that the final answer to the question of whether VC7 was GPL-compatible was "yes"?
Sorry, if there was some debate about that, I missed it. That's not an invitation to start another one here <wink> -- since "GPL compatible" is determined by pronouncements from the FSF, I strongly advise asking them directly whether a thing is GPL-compatible.
Jack Jansen wrote:
I don't recall that the question has been asked before - let alone recalling an answer to it. If you were wondering about the question "is it possible to build extension modules for a VC7 compiled Python without having to buy VC7", then this question would have nothing to do with legal issues, but much more with technical ones. Regards, Martin
On 24 Jun 2004, at 07:15, Martin v. Löwis wrote:
The question I mean is that the GPL says "You shall not link against commercial libraries, unless they are part of the base OS". The second part of that sentence could be considered true for the VC6 version of MSVCRT, but not for the VC7 one. Actually, Microsoft strongly suggest you ship msvcr71.dll (or whatever the name) with the application, and not put it into /System. That appears to me to be a definite no-no for GPL software. -- Jack Jansen, <Jack.Jansen@cwi.nl>, http://www.cwi.nl/~jack If I can't dance I don't want to be part of your revolution -- Emma Goldman
[Jack Jansen]
Python isn't released under the GPL, so what specifically are you asking about? Whether an author of a Python 2.4 derivative work targeting Windows can put their work under the GPL? I expect that's more a question of the conditions under which Microsoft licenses msvcr71.dll. They've had some goofy language in the past purporting to prevent redistribution of MS redistributables in packages under "viral" licenses. Maybe they still do. I expect that by "You shall not link ..." (which doesn't appear in the GPL), you have in mind the confusing requirements of GPL Version 2 Section 3. That exempts copiers and distributors from including full source code for "major components ... of the operating system on which the executable runs", *except* "unless that component itself accompanies the executable". Python installers have always included the MS C runtime DLLs, so there was no relief in the GPL for VC6. I suggest asking the FSF directly what they believe. Regardless of internal consistency, I expect they'll develop a rationale for saying "sure, fine". The questions really have nothing to do with Python.
Phil Thompson wrote:
As Tim says, if you want to know for sure, ask the FSF for clarification. It is not true that GPL code cannot be linked against proprietary libraries. If you have the code, you can link it against whatever you want. The GPL does not restrict your right to link the application against libraries. The GPL only talks about distribution of things. If you include MSVCR71.DLL in your binary Python module, you would need to provide source code to MSVCR71.DLL as well. The solution is very simple: just don't ship MSVCR71.DLL with the binary Python module. It will work on the target system because Python has already installed that DLL. Regards, Martin
On Saturday 26 June 2004 12:37 pm, "Martin v. Löwis" wrote:
Actually, it's simpler than that. From the GPL FAQ: "I'm writing a Windows application with Microsoft Visual C++ (or Visual Basic) and I will be releasing it under the GPL. Is dynamically linking my program with the Visual C++ (or Visual Basic) run-time library permitted under the GPL? Yes, because that run-time library normally accompanies the compiler or interpreter you are using." My (incorrect) understanding was that "operating system" was meant rather than "compiler or interpreter". Phil
[Phil Thompson]
Heh. That's what I meant by Regardless of internal consistency, I expect they'll develop a rationale for saying "sure, fine". I quoted the relevant part of the GPL verbatim before, and it does indeed say specifically and only "operating system". This is why you have to ask the FSF directly about what they mean: since the meaning of the GPL has never been clarified via court rulings, in practice it means whatever the FSF says it means, and that can't always be predicted from what the GPL actually says.
"Tim" == Tim Peters <tim.peters@gmail.com> writes:
Tim> [Phil Thompson] >> Actually, it's simpler than that. From the GPL FAQ: >> "I'm writing a Windows application with Microsoft Visual C++ >> (or Visual Basic) and I will be releasing it under the GPL. Is >> dynamically linking my program with the Visual C++ (or Visual >> Basic) run-time library permitted under the GPL? >> Yes, because that run-time library normally accompanies the >> compiler or interpreter you are using." That's weird; that kind of blanket statement simply is not compatible with the letter of the GPL. I wonder if Eben Moglen vets that FAQ. They must be assuming that the run-time library is "compiler helper" stuff, eg-but-not-limited-to code to handle long long ints and set up the C++ exception and global constructors, etc. So msvcrt should be fine. The interpretation is that this is code whose presence is mandated in one form or another by the language definition. Then it could be generated at build time by the compiler, but instead it's "pregenerated" and supplied as a library. The GPL doesn't care whether the translator is implemented via inline code generation or by library calls. In other words, if you call any functions in mscvrt directly, then you've probably got GPL issues. If you use proprietary language extensions implemented as library calls, you may have GPL issues. But if all calls are implicit in standard C code that could just as well be open-coded, or linked with cygwin.dll or libgcc1 and crt.o, no problem. >> My (incorrect) understanding was that "operating system" was >> meant rather than "compiler or interpreter". Unfortunately, not all official statements by the FSF are vetted by lawyers. Don't conclude you were incorrect until you've gotten legal advice from a lawyer with his lawyer hat on (either yours or the FSF's). However, in this case I would guess that this has nothing to do with the "operating system exception." Tim> Heh. That's what I meant by Tim> Regardless of internal consistency, I expect they'll Tim> develop a rationale for saying "sure, fine". Heh, good luck. Ask Peter Deutsch or the guys at TrollTech about that. Tim> This is why you have to ask the FSF directly about what they Tim> mean: since the meaning Of the GPL has never been clarified Tim> via court rulings, in practice it means whatever the FSF says Tim> it means, and that can't always be predicted from what the Tim> GPL actually says. The real reason for asking the FSF, of course, is that legal advice from the FSF is free-as-in-free-beer, while your lawyer will get a vacation on the Riviera for answering that question, despite being far less expert than Eben Moglen. [Aside: It's both not as bad as you suggest, and far worse than that. But that's not relevant today.] Of course, the usual dysclaimer: I ANAL. But-I'm-writing-this-to-avoid-my-real-work-ly y'rs, -- Institute of Policy and Planning Sciences http://turnbull.sk.tsukuba.ac.jp University of Tsukuba Tennodai 1-1-1 Tsukuba 305-8573 JAPAN Ask not how you can "do" free software business; ask what your bu
Stephen J. Turnbull wrote:
I think the reasoning is different: in order to run software, you need an operating system. It is ok if free software doesn't ship with its own operating system, but if users of the software already have an operating system. Likewise, it is ok if free software doesn't ship with the compiler to compile it - users who want to build the software are expected to obtain a copy of the compiler. Therefore, users can be expected to have everything that comes with the operating system, and everything that comes with the compiler. Therefore, it is ok to assume that people already have a copy of msvcr71.dll if they want to use VC7.1 to build something. Regards, Martin
"Martin" == Martin v Löwis <martin@v.loewis.de> writes:
Martin> Stephen J. Turnbull wrote: >> They must be assuming that the run-time library is "compiler >> helper" stuff, eg-but-not-limited-to code to handle long long >> ints and set up the C++ exception and global constructors, etc. >> So msvcrt should be fine. The interpretation is that this is >> code whose presence is mandated in one form or another by the >> language definition. Martin> I think the reasoning is different: in order to run Martin> software, you need an operating system. It is ok if free Martin> software doesn't ship with its own operating system, but Martin> if users of the software already have an operating Martin> system. Likewise, it is ok if free software doesn't ship Martin> with the compiler to compile it - users who want to build Martin> the software are expected to obtain a copy of the Martin> compiler. No. I've never seen either rms or Eben Moglen admit such reasoning that I know of (of course one or both may have approved the FAQ, but I don't know that). The operating system is a _specific_ exception written into the GPL for those reasons, but that doesn't mean the logic can be extended to compilers at the licensee's option. There is never a problem with building GPL sources. You are always permitted to do that, and as far as the GPL is concerned, you may link with anything (although I've heard that some MSFT licenses prohibit the reverse). What is restricted by the GPL is _redistribution_ of the product of linking (which includes source that is "intended" to be linked). Thus, under copyright law, the binary may be a derivative of the runtime library. Determining that is a delicate matter. If you ship only source which does not depend on extensions in the runtime, your logic is correct. However, if you distribute Python binaries that require that DLL, your logic that users already have it if they're going to build Python is irrelevant. Isn't distribution of binaries, including the runtime DLL, what is being discussed here? -- Institute of Policy and Planning Sciences http://turnbull.sk.tsukuba.ac.jp University of Tsukuba Tennodai 1-1-1 Tsukuba 305-8573 JAPAN Ask not how you can "do" free software business; ask what your business can "do for" free software.
Stephen J. Turnbull wrote:
Well, I won't distribute Python binaries: the PSF, or, more specifically, Thomas Heller will. For the PSF, the GPL is not binding, since Python is not licensed under the terms of the GPL. My logic was talking about Python *extension* modules. Those don't have to include a copy of the runtime, as the Python installation will already have it. However, this is all theoretical, as apparently nobody is concerned anymore that they couldn't do what they wanted to do (they may still be concerned that others couldn't do what they wanted to do, but this is still theoretical unless these others step forward). Regards, Martin
If you include MSVCR71.DLL in your binary Python module, you would need to provide source code to MSVCR71.DLL as well.
If that's true, then it's illegal to distribute *any* GPL software compiled with VC7 -- or with any other language implementation that links in a non-open-source runtime library -- and has been for a long time. I find it hard to believe that's the intent of the GPL. There must be a lot of software out there that does this -- is it all breaking the law? Greg Ewing, Computer Science Dept, +--------------------------------------+ University of Canterbury, | A citizen of NewZealandCorp, a | Christchurch, New Zealand | wholly-owned subsidiary of USA Inc. | greg@cosc.canterbury.ac.nz +--------------------------------------+
Greg Ewing wrote:
As always, Tim is right: if you want to know for sure, ask the FSF lawyer. My interpretation is that in the strict sense of the letter these programs violate the GPL. The entire binary distribution forms "work based on the Program"; inclusion of msvcr71.dll is not "mere aggregation". Then, according to clause 3, you are required to accompany the program with a machine-readable source code (or equivalent offers), which means "complete source code means all the source code for all modules it contains". The special exemption for things normally distributed with the operating system does not apply, as msvcr71.dll is not normally distributed with "major components (compiler, kernel, and so on) of the operating system". However, it appears that the FSF FAQ has taken the stance that VC7.1 is indeed a major component of the operating system (being the system compiler), and that therefore, the exemption clause does apply. Under U.S. law, it seems that only a court could interpret this license for good. But what do I care about U.S. law. Regards, Martin
participants (7)
-
"Martin v. Löwis" -
Greg Ewing -
Jack Jansen -
Phil Thompson -
Stephen J. Turnbull -
Tim Peters -
Tim Peters