GFDL and Debian Logo
Sorry, I pasted the mail into the subject line by accident. Here is a secondy try: Hi, there is a discussion in the German Wikipedia whether the Debian Open Use Logo http://de.wikipedia.org/wiki/Bild:Debian_logo.png may be subjected to the GFDL. The German Wikipedia does not accept any content not licensed as public domain or GFDL (e. g. no fair use because this is an American law only). This logo or a modified version may be used by anyone to refer to the Debian project, but does not indicate endorsement by the project As far as I understand this sentence, the logo may not be used for other things and is not free according to the GFDL. The possible copyright violation notice, however, was removed by the uploader today. http://de.wikipedia.org/wiki/Bild_Diskussion:Debian_logo.png (Although this page is in German, you may add English comments to it). Hendrik
Re: GFDL and Debian Logo
On Wed, Sep 22, 2004 at 07:57:40AM +0200, Hendrik Brummermann wrote: there is a discussion in the German Wikipedia whether the Debian Open Use Logo http://de.wikipedia.org/wiki/Bild:Debian_logo.png may be subjected to the GFDL. The German Wikipedia does not accept any content not licensed as public domain or GFDL (e. g. no fair use because this is an American law only). Wikis are still using the GFDL? That's discouraging, considering how badly broken it is; given the number of flaws that have been found in it, it's discouraging that any rational person would use it for anything at all. This logo or a modified version may be used by anyone to refer to the Debian project, but does not indicate endorsement by the project As far as I understand this sentence, the logo may not be used for other things and is not free according to the GFDL. The possible copyright violation notice, however, was removed by the uploader today. The GFDL is a license; it's not something used to determine freeness. The DFSG does that. Are you mixing up abbreviations? -- Glenn Maynard
Re: GFDL and Debian Logo
On 09/22/04 01:57:40, Hendrik Brummermann wrote: there is a discussion in the German Wikipedia whether the Debian Open Use Logo http://de.wikipedia.org/wiki/Bild:Debian_logo.png may be subjected to the GFDL. I'm not a lawyer and I don't speak for Debian, but I don't think that you can re-license the Open Use Logo under the GFDL. As I read it, the Open Use Logo license requires that you use the logo and derivatives _to_refer_to_the_Debian_project_. The GFDL doesn't preserve this requirement. If the logo were put under the GFDL, someone could use it to refer to an ice cream shop, an electrical component, or an astrological sign. What does Debian care if someone does that? Because under American (and I think some other) trademark regimes, if we allow people to use our trademark image to refer to just anything, it dilutes our trademark and makes it difficult for us to enforce. If someone put the Debian logo on, say, a Linux distribution without any Debian components inside, we would have a hard time making them stop. That logo can mean anything you want, they could say. This actually matters in US courts. So, my inexpert answer: NO. As a Wikipedian, I'd recommend removing the logo from German Wikipedia (although en: is a lot less rigorous about information freedom -- bravo to de:!). ~ESP
Re: Open Software License v2.1
On 2004-09-22 02:13:04 +0100 Glenn Maynard [EMAIL PROTECTED] wrote: Bob creates Emacs, under a claim patent infringement in this work and lose your license to it license, which includes GIF decoding. Lose your patent licence or all licence? -- MJR/slefMy Opinion Only and not of any group I know
Re: Open Software License v2.1
On 2004-09-21 23:16:47 +0100 Josh Triplett [EMAIL PROTECTED] wrote: For what it's worth, I agree entirely. No software patent is legitimate, and clauses stating that you can't continue to use a piece of Free Software while claiming that software infringes your patent are both Free and desirable. That is a bug with some jurisdictions, not a problem in copyright law. Attempting to use copyright law to fix all the world's ills is a short road to madness. If you are in a swpat-afflicted place, then losing your patent licence is sufficient to forbid use, as I understand it. There seems little need to confuse copyright and patent laws. Do these terms try to change the usual presumption that no-one has wronged anyone? Bogus copyright accusations are also illegitimate, yet the GPL does not state that you can't continue to use a piece of free software while merely *claiming* that it infringes your copyright. Does any free software copyright licence contain a copyright defence clause similar to these patent defence clauses? Are there copyright clauses which forbid even participating in copyright cases about that work? Have they been or would they be regarded as following DFSG? -- MJR/slefMy Opinion Only and not of any group I know Creative copyleft computing - http://www.ttllp.co.uk/ http://www.thewalks.co.uk stand 13,Lynn Carnival,12 Sep
Re: Open Software License v2.1
On Wed, Sep 22, 2004 at 09:04:29AM +0100, MJ Ray wrote: Bob creates Emacs, under a claim patent infringement in this work and lose your license to it license, which includes GIF decoding. Lose your patent licence or all licence? Both patent and copyright. -- Glenn Maynard
Re: Open Software License v2.1
On Wed, Sep 22, 2004 at 11:44:13AM +0100, Andrew Suffield wrote: Bob creates Emacs, under a claim patent infringement in this work and lose your license to it license, which includes GIF decoding. Joe derives XEmacs from that work. This inherits, among many other things, GIF decoding. Bill sues Joe, claiming that XEmacs infringes his GIF patent. Does and should Bill lose his license to Emacs, in addition to XEmacs? I think the answer to both is yes. The copyright and patent holder has no need for a license. Bill is not a copyright holder at all in this scenario. -- Glenn Maynard
Re: missing licenses in gnulib
On 2004-09-22 14:58:24 +0100 Robert Millan [EMAIL PROTECTED] wrote: [ putting debian-legal on CC ] For what end? There's also the problem with non-free documentation in doc directory (3 files), but I'm aware that for the FSF freedom isn't important for documentation so I'm ommiting the list here. To me, this looks like you are trolling a list @gnu.org - please be more precise with your wording on this topic. My current understanding is that the FSF does not consider is this documentation also free software? to be a sensible question, while many debian developers do. I agree with the conclusion that it's better to omit them from discussion for now. -- MJR/slefMy Opinion Only and not of any group I know Creative copyleft computing - http://www.ttllp.co.uk/ LinuxExpo.org.uk village 6+7 Oct http://www.affs.org.uk
Re: Open Software License v2.1
Scripsit Brian Thomas Sniffen [EMAIL PROTECTED] PS You know, I just thought of something. If these clauses cancelled the copyright license to *everybody* as soon as *anybody* *wins* a patent lawsuit over the software, I wouldn't mind them so much. That would spectacularly fail the tentacles-of-evil test. If the author, Foobar Ltd. happens to be acquired by Evil Megacorp, E.M. could have one of their other subsidiaries sue Foobar for with a claim that their xor-cursor patent is violated, and deliberately let Foobar put up no competent defense at all in court. Poof, everybody's copyright license is gone. It's the cancellation of the license for even seeking impartial justice that bothers me. The situation the clause aims at is one where a patent owner seeks to gain a monopoly on the original author's work by preventing everybody else - including the original author himself - from using it. I don't think justice, impartial or not, has anything to do with that. My intuition is that it is fair for free software to say, if you want to have a monopoly on implementations of your patented gadget, you have to write the code yourself. -- Henning Makholm In my opinion, this child don't need to have his head shrunk at all.
Re: Open Software License v2.1
Brian Thomas Sniffen wrote: Raul Miller [EMAIL PROTECTED] writes: On Tue, Sep 21, 2004 at 09:44:47PM -0400, Brian Thomas Sniffen wrote: : A writes some software, and GPLs it. B claims that the software is : on his hard drive, and sues A for that drive. B wins, and now only : B can distribute the software -- A can't [and no one else can] : without getting a license from B. This scenario of yours has nothing to do with the freeness of the license. It's just your patent scenario from above, rephrased to deal with physical property. Are you claiming that a license which prevents this scenario, by terminating the copyright license to anyone who sues regarding physical media containing the software, is free? Patents are not property. Patents do not behave like property, despite the unfortunate terms used to refer to a broad class of law that includes patents. Claiming that someone stole your property in no way approximates a patent lawsuit. Using your analogical situation, the comparable case with a patent lawsuit would be B claiming that A's hard drive is similar to B's hard drive, and that B thinks he should control all hard drives similar to his. - Josh Triplett signature.asc Description: OpenPGP digital signature
Re: Bug#265352: grub: Debian splash images for Grub
Matthew Garrett wrote: Josh Triplett [EMAIL PROTECTED] wrote: Both of these licenses seem clearly non-free to me, since they restrict the uses of unmodified or insufficiently different versions. Trademark law limits what can be done here. Granting a trademark license (explicitly or implicitly) that allows people to use the trademark for any purpose is likely to result in the loss of the trademark. The trademark license would only apply in cases where there /is/ a potential trademark issue - in all other cases, copyright law would already apply. Granting the right to use and modify an image of a trademark without providing a license to use the trademark would be more restrictive than the licenses suggested above, but is something that I don't think we've really considered in the past. First of all, even if it is the case that we can't offer a DFSG-free license for the logo without allowing it to become diluted, then that does not exempt it from being DFSG-free. I believe the suggested licenses were very clearly non-DFSG-free. Second, I'm not suggesting that we put no restrictions on the logo. I would suggest that we require people who copy, modify, or distribute the logo to acknowledge the origin of the logo, and not misrepresent it as being written by them. We should probably also include a copyleft. Such restrictions would be DFSG-free, and probably GPL-compatible for that matter. I also believe such a restriction, if actually enforced, ought to be sufficient to maintain a trademark. (IANAL, hence ought to be.) Third, if we want a logo with a restrictive license, there is always the Official Use logo. (I dislike the idea that any of Debian's logos would be non-DFSG-free, however.) - Josh Triplett signature.asc Description: OpenPGP digital signature
Re: GFDL and Debian Logo
Walter Landry [EMAIL PROTECTED]: The Debian Open Use Logo is not compatible with the GFDL. If fair use is really that limited in Germany, then the German wikipedia is going to have to purge all logos. I doubt that any have anything approaching a free license. As a comparison, the English entries for IBM and HP have their logos, while the German entries do not. So at least that is consistent. Perhaps I'm being thick here, but what legal difference does the language make? Doesn't the German Wikipedia use the same licence as the English Wikipedia, and aren't they both accessible in Germany?
Re: Bug#265352: grub: Debian splash images for Grub
Josh Triplett [EMAIL PROTECTED]: First of all, even if it is the case that we can't offer a DFSG-free license for the logo without allowing it to become diluted, then that does not exempt it from being DFSG-free. I believe the suggested licenses were very clearly non-DFSG-free. Does it qualify as DFSG-free if you give it a free copyright licence without granting any kind of trademark licence? If it doesn't, what exactly are the situations in which a trademark licence is required in addition to a copyright licence? I note that many packages contain the word Microsoft without there being a DFSG-free licence for that trademark, so a line would have to be drawn somewhere between the two cases; I can't immediately see how or why.
Re: Bug#265352: grub: Debian splash images for Grub
On Wed, Sep 22, 2004 at 06:41:29PM +0100, Edmund GRIMLEY EVANS wrote: Does it qualify as DFSG-free if you give it a free copyright licence without granting any kind of trademark licence? Trademark license is orthogonal to copyright license. Trademark laws are more like truth in advertising laws than anything else -- they're aimed at preventing deception -- they don't prohibit any non-deceitful uses or changes. [That said, it's the trademark owner who gets to define what's truthful use of the trademark, and what's not.] Trademark license isn't specific to a piece of software -- it's specific to a context. -- Raul
Re: CeCILL again...
Glenn Maynard wrote: On Sun, Aug 22, 2004 at 09:29:13PM -0700, Steve Langasek wrote: Agreed; I think at a minimum we need either a legally-binding license in English, or an English translation vouched for by someone debian-legal subscribers are willing to trust, to be able to make any determination that a license is DFSG-free. But we have a few goals: for us to be able to determine that it's DFSG-free; for us to be able to determine that Debian and its mirrors can safely distribute it even if so (eg. license compatibility); and for users to be able to know and understand what they can do to a given work. I'd be willing, for example, to trust others to the first, since it doesn't put me in actual risk; but I'd be much slower to trust others to the rest, since an error can cause me to violate the license. Generally, I'd only personally accept the former (a legally-binding license in English) before distributing modifications to a program. I want to be able to read my rights for myself. (I'm not a lawyer, so I might mess up and read it wrong, but at least it'd be my own fault.) Well, we only need someone fluent in French legalese to verify that the GPL conversion is valid. After that, we _have_ a legally-binding, English, DFSG-free license. - Josh Triplett signature.asc Description: OpenPGP digital signature
Re: CeCILL again...
Glenn Maynard wrote: On Tue, Aug 24, 2004 at 11:19:27AM +0100, Edmund GRIMLEY EVANS wrote: Glenn Maynard wrote: who should not have to be bound by a text in a language that they don't understand properly. (The only solution available to me, in that situation, is to not touch the software.) Then you have a solution. Use it. But please don't try to force your solution on other people who may be perfectly happy, or even happier, with a licence in French. I don't think just don't use the software is an acceptable solution. Your requirement that you be able to understand the license terms is perfectly reasonable. That does not mean Debian should have that requirement. Consider this: it is also a perfectly reasonable requirement that you be able to read the source of all the software you use, to ensure that it does not take undesirable actions. This might well cause you to avoid using software in programming languages you don't understand, or with comments and variable names in languages you don't understand. Nevertheless, such software is still Free. - Josh Triplett signature.asc Description: OpenPGP digital signature
Re: Open Software License v2.1
Raul Miller [EMAIL PROTECTED] writes: If the software is not free, regardless of the copyright license, then the reason it's not free is not the copyright license. Thus, this scenario has no bearing on the freeness of the license. I don't think that's true. Certainly, I see no reason it should be accepted as obviously true. Where, specifically, do you disagree? [Let's take it for granted that if you disagree with an antecedent that you feel that the consequent is illogical.] Something can be non-free for many reasons. If it is non-free regardless of the copyright license, and under a non-free copyright license, then it is nonsensical to speak of a single reason that it is non-free. I think I see what conclusion you'd like me to reach, though -- that a copyright license which is sometimes free and sometimes non-free, but only non-free in cases where the software in question is already non-free anyway, is a free license. Is that the basic idea? If so, I almost believe it. Almost, but not quite. By way of comparison, consider a GPL-like license which additionally prohibited private modification to include non-GPL-compatible works written by others -- that is, to prohibit making any thing you could not distribute under the GPL out of parts you have only under the GPL. That's non-free, and this is non-free for the same reason. A writes some software, and GPLs it. B claims that the software is patent restricted, and sues A. B wins, and now only B can distribute the software -- A can't [and no one else can] without buying a license from B. You seem to be claiming that a license which prevents this scenario is not good, and that the reason it's not good is that it prevents this scenario. It might be good. It isn't free. If it is free, why isn't the following free: : A writes some software, and GPLs it. B claims that : the software is on his hard drive, and sues A for : that drive. B wins, and now only B can distribute : the software -- A can't [and no one else can] : without getting a license from B. This scenario of yours has nothing to do with the freeness of the license. It's just your patent scenario from above, rephrased to deal with physical property. That's not my patent scenario, that's my paraphrase of your patent scenario. I'm claiming that the scenario is invalid, your rephrasing of it didn't make it valid. What's invalid about that scenario? I understand that you object to the conclusions drawn from it, but now you claim that the hypothetical itself is invalid. What does this mean? -Brian -- Brian Sniffen [EMAIL PROTECTED]
Re: GFDL and Debian Logo
Edmund GRIMLEY EVANS schrieb: Walter Landry [EMAIL PROTECTED]: The Debian Open Use Logo is not compatible with the GFDL. If fair use is really that limited in Germany, then the German wikipedia is going to have to purge all logos. I doubt that any have anything approaching a free license. As a comparison, the English entries for IBM and HP have their logos, while the German entries do not. So at least that is consistent. Perhaps I'm being thick here, but what legal difference does the language make? Doesn't the German Wikipedia use the same licence as the English Wikipedia, and aren't they both accessible in Germany? All wikipedias use the GFDL :-( The English community, however, accepts fair use images. The German community does not. For obvious reasons there are a lot of American in the English community but only a few in the German one. Although the wikipedia-servers are in America someone in Germany may get in trouble if he or she uploads images to a public server or uses them against the German law. As most people do not care about licenses the German community has decided to require GFDL or public domain. And of course the German Wikipedia should be free (ignoring the issues of the GFDL without invariant sections). Hendrik
Re: Open Software License v2.1
Henning Makholm [EMAIL PROTECTED] writes: Scripsit Brian Thomas Sniffen [EMAIL PROTECTED] PS You know, I just thought of something. If these clauses cancelled the copyright license to *everybody* as soon as *anybody* *wins* a patent lawsuit over the software, I wouldn't mind them so much. That would spectacularly fail the tentacles-of-evil test. If the author, Foobar Ltd. happens to be acquired by Evil Megacorp, E.M. could have one of their other subsidiaries sue Foobar for with a claim that their xor-cursor patent is violated, and deliberately let Foobar put up no competent defense at all in court. Poof, everybody's copyright license is gone. But as Poole and others have argued here, if something is less than perfectly free for any reason, non-freeness of any other sort matters not. Applied here, that means your example doesn't matter because they could do it anyway, by simply buying the patent and threatening suit. It's the cancellation of the license for even seeking impartial justice that bothers me. The situation the clause aims at is one where a patent owner seeks to gain a monopoly on the original author's work by preventing everybody else - including the original author himself - from using it. Your use of the term original author is misapplied. Either the copyright owner is not the original author, because the patent predates his work, or he is the original author and can win the suit easily. I don't think justice, impartial or not, has anything to do with that. My intuition is that it is fair for free software to say, if you want to have a monopoly on implementations of your patented gadget, you have to write the code yourself. Similarly, I think it's fair to say that Free Software licenses should not attempt to circumvent the courts, and that penalties for bringing law cases belong only in negotiated contracts. -Brian -- Brian Sniffen [EMAIL PROTECTED]
Re: Open Software License v2.1
If the software is not free, regardless of the copyright license, then the reason it's not free is not the copyright license. Thus, this scenario has no bearing on the freeness of the license. I don't think that's true. Certainly, I see no reason it should be accepted as obviously true. Raul Miller [EMAIL PROTECTED] writes: Where, specifically, do you disagree? [Let's take it for granted that if you disagree with an antecedent that you feel that the consequent is illogical.] On Wed, Sep 22, 2004 at 03:58:11PM -0400, Brian Thomas Sniffen wrote: Something can be non-free for many reasons. If it is non-free regardless of the copyright license, and under a non-free copyright license, then it is nonsensical to speak of a single reason that it is non-free. Sure. But you seem to have claimed that this scenario has no bearing on the freeness of the license might not be correct. Ok, there might be more than one reason for a license to be free or non-free -- but the possibility that more than one reason exists doesn't seem to make an irrelevant scenario relevant. I think I see what conclusion you'd like me to reach, though -- that a copyright license which is sometimes free and sometimes non-free, but only non-free in cases where the software in question is already non-free anyway, is a free license. Huh? I'm not asking you to draw any such conclusion. For all I know, you might be right about the osl-2.1 license being non-free. I just want to see a convincing argument that this is the case before I agree to any such thing. [And, I don't consider an irrelevant argument to be a convincing argument.] Is that the basic idea? No. I'm claiming that your argument -- which makes all instances of the program in question non-free, regardless of the license -- has no bearing on the freeness of the license. If X is irrelevant to Y, then X does not prove Y, and X does not disprove Y. If so, I almost believe it. Almost, but not quite. By way of comparison, consider a GPL-like license which additionally prohibited private modification to include non-GPL-compatible works written by others -- that is, to prohibit making any thing you could not distribute under the GPL out of parts you have only under the GPL. That's non-free, and this is non-free for the same reason. I'm not convinced that this example is related in any relevant sense. You've not identified the reason that this example is non-free, so I'm reserving judgement on whether that reason applies to the patent case. That's not my patent scenario, that's my paraphrase of your patent scenario. I'm claiming that the scenario is invalid, your rephrasing of it didn't make it valid. What's invalid about that scenario? I understand that you object to the conclusions drawn from it, but now you claim that the hypothetical itself is invalid. What does this mean? I identified several scenarios, and you ask about that scenario... I'm going to guess and go for this one: A writes some software, and GPLs it. B claims that the software is patent restricted, and sues A. B wins, and now only B can distribute the software -- A can't [and no one else can] without buying a license from B. As stated, the scenario is misleading: B can't satisfy the GPL while distributing the program without granting transitive rights to the patent. In other words, there's two potential cases: the software is not distributed at all [the typical case], or it's distributed under the original terms [with B not benefiting from the lawsuit]. Let's try this -- YOU present a scenario which illustrates the non-free nature of osl-2.1. -- Raul
Re: Open Software License v2.1
Brian Thomas Sniffen writes: Henning Makholm [EMAIL PROTECTED] writes: The situation the clause aims at is one where a patent owner seeks to gain a monopoly on the original author's work by preventing everybody else - including the original author himself - from using it. Your use of the term original author is misapplied. Either the copyright owner is not the original author, because the patent predates his work, or he is the original author and can win the suit easily. His use of the term is correct. The software was written by the original author. Software can infringe a patent that was issued before, after or concurrently with the software being published. The odds are that, even for a patent with a filing date after the software was first published, the defendant in a patent suit would have trouble winning on that basis. Among other reasons, many countries have a grace period that allows publication of an invention before filing a patent for it; the USA allows such publication by third parties. I don't think justice, impartial or not, has anything to do with that. My intuition is that it is fair for free software to say, if you want to have a monopoly on implementations of your patented gadget, you have to write the code yourself. Similarly, I think it's fair to say that Free Software licenses should not attempt to circumvent the courts, and that penalties for bringing law cases belong only in negotiated contracts. Can you relate this limitation on licenses to the DFSG? Michael Poole
Re: Open Software License v2.1
MJ Ray wrote: On 2004-09-21 23:16:47 +0100 Josh Triplett [EMAIL PROTECTED] wrote: For what it's worth, I agree entirely. No software patent is legitimate, and clauses stating that you can't continue to use a piece of Free Software while claiming that software infringes your patent are both Free and desirable. That is a bug with some jurisdictions, not a problem in copyright law. Yes, this is true. Fixing the jurisdictions ranges from impossible to a 20-year project, however. Attempting to use copyright law to fix all the world's ills is a short road to madness. Good thing we're not doing that then. If you are in a swpat-afflicted place, then losing your patent licence is sufficient to forbid use, as I understand it. There seems little need to confuse copyright and patent laws. The purpose, as I see it, is specifically to prevent a method of taking a program proprietary. Do these terms try to change the usual presumption that no-one has wronged anyone? Bogus copyright accusations are also illegitimate, yet the GPL does not state that you can't continue to use a piece of free software while merely *claiming* that it infringes your copyright. Claiming *in court*. Does any free software copyright licence contain a copyright defence clause similar to these patent defence clauses? No. I do not see it as a substantially different situation, though. Suppose B alleges that the distribution of work X by copyright holder A under the GPL is illegitimate because, according to B, B, not A, holds copyright in work X. In that case nobody has permission to distribute X, because the copyright holder (B) has not given permission. Now, if B alleges that both A and B hold copyrights in X, then the interesting result is that B has permission to use A's work, but nobody else does. This seems unreasonable and unfair. Terminating B's license evens things out again. The situation where this might be undesirable, and a retaliation clause bad, is if B is already issuing B's work under a free license to everyone. In this case B should perhaps be free to sue A for violation of the free license's terms, without losing his license to A's work. OK, I've convinced myself that copyright defence clauses are bad because of this situation. In the analogous case for patents, I'm wavering. The difference is that I'm not sure there's such a thing as a fair free license for a software patent which has *any* restrictions at *all*. Possibly an attribution requirement -- but perhaps not, given that a patent can control totally independent invention of the same thing, and in that case attribution is not warranted. Are there copyright clauses which forbid even participating in copyright cases about that work? Obviously not. The patent clauses don't do that either, so this is quite irrelevant. Have they been or would they be regarded as following DFSG? -- This space intentionally left blank.
Re: Open Software License v2.1
Scripsit Brian Thomas Sniffen [EMAIL PROTECTED] It think it's free to terminate a public license completely and universally as soon as anybody brings and wins any suit against any party that claims that the work using some patented technology. Still fails the Tentacles of Evil test, this time very spectacularly. Imagine that, say, Emacs was under such a license. Microsoft Denmark Aps then starts selling media with the source code for Emacs on them. Microsoft Germany GmbH files suit against MS Denmark, alleging that Emacs uses Microsoft's patented two's-complement technology for representing negative numbers, and demanding royalties payments. The defendant agrees that the technology is, in fact, used, but refuses to pay royalty on grounds that half of the profits is being donated to a charity. The judge, correctly, dismisses this defense as rubbish and has to rule in favor of the plaintiff. Poof. Suddenly everybody, has lost their license for Emacs, completely and universally. (This will have the most practical importance if Microsoft has previously acquired the FSF, or RMS has gone mad and started working for Redmond. But that's exactly the kind of things a license has to survive in order to be free). -- Henning Makholm (Og det er vasketøjet tit.)
Re: Open Software License v2.1
Glenn Maynard wrote: Ick. A, B, C, X, VD, MSC, π. I find these hypotheticals to be a lot easier to parse and process if I give these people names and use actual projects to put things in perspective with one another ... On Tue, Sep 21, 2004 at 03:08:04PM -0400, Nathanael Nerode wrote: (Essentially, by buying the copyright, they would have gotten themselves a special license to avoid the patent-termination clause, and that's all.) However, the question is if the patent grant is intrinsic to these patent defense clauses being free. Without a patent license grant, if John writes GIFEnc, and holds a GIF patent, he can distribute GIFEnc under a patent-defense license, and then sue users of it for violating his GIF patent. Those users can't as easily countersue by claiming that GIFEnc also violates their own GIF patent, because they'll lose their license to it. With the patent grant, John can't sue for that patent violation--or if he does, users have a trivial defense (he granted them a license). Is the first case free? If you think so, the copyright-transfer case is irrelevant. Hmm. No, the first case is non-free. However, if he distributed under a non-patent-defense license, it would *still* be non-free. So I'm not clear on how the so-called patent-defence clause makes any difference here. I'm inclined to say it's not. I'll agree not to sue you for patent violation for a license to this software, but only if it's mutual--don't give me a copyright license and then turn around and sue me for patent violation. I'm not going to give up my ability to countersue unless you agree to make it unnecessary. Now, I can still be sued by GIFCorp for patent violation. They'll lose their license to GIFEnc, but they may not care. In that case, I won't be able to sue them for violations due to their use of GIFEnc--they don't even use GIFEnc. So the case in question is: GIFCorp has bought GIFEnc and uses it heavily, and sues me for violating their own GIF patents (which the origial patent grant didn't include). They do use GIFEnc, but I can't countersue for their violations in that use--I'll lose my license. We're back to the above without a patent license grant case. Right but if the patent-defence clause is absent, then either: (1) GIFEnc is still free, because GIFCorp's patents will be defeated (In which case what's the need for the countersuit?) Or (2) GIFEnc isn't free So what difference does the patent-defence clause make? You are convincing me that patent license grants are a necessary component of *all* licenses, not just patent-defence licenses. :-) -- This space intentionally left blank.
Re: Open Software License v2.1
Glenn Maynard wrote: (Unrequested CC sent; it just seems like a good idea when sending mails concerning possible MUA problems ...) On Tue, Sep 21, 2004 at 01:16:51PM -0400, Nathanael Nerode wrote: You haven't been reading my postings? I doubt anyone is reading all of your postings, due to your well-known bad habit of spamming the list with a page of replies to old posts at a time. I wouldn't be surprised if some people had you killfiled for it. This is the fourth post I see from you on this list this month, followed by about twenty more. In this case, I am catching up on a month of postings. And I (a) haven't said anything redundant (b) have stuck to a strict rule of no more than one message per topic (I wouldn't bother asking this, since every time I have you've blown me off or ignored me, as I recall; except that this response seems to indicate that you think you've been posting My opinions on the software-patent retaliation clauses were made clear months ago during the discussion of the Apache License 2.0; I doubt that Andrew Suffield was killfiling me at that time. even though I see almost all of your posts *after* this one ...) -- This space intentionally left blank.
Re: Nathanael - your mail is broken [MAILER-DAEMON@zewt.org: Undelivered Mail Returned to Sender]
Glenn Maynard wrote: Nathanael, my CC to you bounced, because your ISP is using a bullshit spam filter. I'm only forwarding this to make sure that you're aware that you're losing legitimate mail because of it. Yes, I'm in the process of getting a new account. -- This space intentionally left blank.
Re: Bug#265352: grub: Debian splash images for Grub
Josh Triplett wrote: First of all, even if it is the case that we can't offer a DFSG-free license for the logo without allowing it to become diluted, then that does not exempt it from being DFSG-free. I believe the suggested licenses were very clearly non-DFSG-free. Second, I'm not suggesting that we put no restrictions on the logo. I would suggest that we require people who copy, modify, or distribute the logo to acknowledge the origin of the logo, and not misrepresent it as being written by them. The point in a traditional common-law trademark is that we don't want someone to go out and start Debian Computing, Inc., use the Debian open-use logo, and proceed to run a competing organization. A trademark license *has* to prohibit such things. Prohibiting misrepresenting the origin of the *logo* doesn't suffice. We have to require that the logo, and anything confusingly similar, is not used to identify things which aren't Debian. We can (and should), however, * allow it (and the word Debian) to be used in any manner which identifies Debian, the project or the distribution (including by Debian-based distributions, Debian redistributors, Debian developers, Debian promoters, Debian haters, Linux collages, etc.). This is in contrast to normal corporate trademark licenses. * allow it to be used in any manner which isn't identifying anything (abstract art, etc.) * allow modified versions which are not confusingly similar to be used for any purpose whatsoever We should probably also include a copyleft. Such restrictions would be DFSG-free, and probably GPL-compatible for that matter. I also believe such a restriction, if actually enforced, ought to be sufficient to maintain a trademark. (IANAL, hence ought to be.) IANAL but I think it isn't from what I know about trademarks. Trademarks are very old-fashioned things. Third, if we want a logo with a restrictive license, there is always the Official Use logo. (I dislike the idea that any of Debian's logos would be non-DFSG-free, however.) - Josh Triplett -- This space intentionally left blank.
Re: Bug#265352: grub: Debian splash images for Grub
On Wed, Sep 22, 2004 at 06:22:45PM -0400, Nathanael Nerode wrote: Josh Triplett wrote: First of all, even if it is the case that we can't offer a DFSG-free license for the logo without allowing it to become diluted, then that does not exempt it from being DFSG-free. I believe the suggested licenses were very clearly non-DFSG-free. Second, I'm not suggesting that we put no restrictions on the logo. I would suggest that we require people who copy, modify, or distribute the logo to acknowledge the origin of the logo, and not misrepresent it as being written by them. The point in a traditional common-law trademark is that we don't want someone to go out and start Debian Computing, Inc., use the Debian open-use logo, and proceed to run a competing organization. I'm not appreciably convinced that prohibiting this is worth the cost. -- .''`. ** Debian GNU/Linux ** | Andrew Suffield : :' : http://www.debian.org/ | `. `' | `- -- | signature.asc Description: Digital signature
Re: Open Software License v2.1
On Wed, Sep 22, 2004 at 07:09:21AM -0400, Glenn Maynard wrote: On Wed, Sep 22, 2004 at 11:44:13AM +0100, Andrew Suffield wrote: Bob creates Emacs, under a claim patent infringement in this work and lose your license to it license, which includes GIF decoding. Joe derives XEmacs from that work. This inherits, among many other things, GIF decoding. Bill sues Joe, claiming that XEmacs infringes his GIF patent. Does and should Bill lose his license to Emacs, in addition to XEmacs? I think the answer to both is yes. The copyright and patent holder has no need for a license. Bill is not a copyright holder at all in this scenario. Not a very interesting scenario, then. You can construct a scenario where any license seems reasonable, including a proprietary one. The mark of free licenses is that you can't construct any where it's unreasonable. -- .''`. ** Debian GNU/Linux ** | Andrew Suffield : :' : http://www.debian.org/ | `. `' | `- -- | signature.asc Description: Digital signature
Re: GFDL and Debian Logo
Edmund GRIMLEY EVANS wrote: Perhaps I'm being thick here, but what legal difference does the language make? Doesn't the German Wikipedia use the same licence as the English Wikipedia, and aren't they both accessible in Germany? Hosting location and intended audience, I assume. Being accessible from a country doesn't normally require that you obey their laws, because that would cause no end of trouble and ridiculous results; there have finally been a few court rulings confirming this. Operating a project from a particular country, or operating it primarily for people from a particular country, on the other hand, is more likely to subject you to their laws. -- This space intentionally left blank.
Re: Bug#265352: grub: Debian splash images for Grub
Josh Triplett wrote: Nathanael Nerode wrote: Josh Triplett wrote: Both of these licenses seem clearly non-free to me, since they restrict the uses of unmodified or insufficiently different versions. Only to the extent of prohibiting misrepresentation of other works, projects, and organizations as belonging to/being endorsed by/being part of Debian. That's a standard, acceptable class of restrictions, isn't it? This really *is* about misrepresentation and nothing more. If that is truly the case, then you don't need to say that at all in the license. See http://lists.debian.org/debian-legal/2004/05/msg00540.html and the resulting thread. Claiming endorsement by someone else without their permission is already illegal. That's for a *person*, whose name has its own rules. Organizations, particularly unincorporated organizations (corporations are often considered people by the law), are sometimes different. We also want to prevent accidental misrepresentation where someone else just happens to use the same name (Debian) or logo (swirl) for their proprietary software company. That's not actually claiming endorsement; they're just attempting to confuse people without actually making a claim; but it is common-law trademark infringement. This is a branding issue vaguely akin to requiring name changes for forks. Under US trademark law, if some unrelated company sets up as Debian Computing and proceeds to operate as such, with Debian's knowledge, and Debian does not try to stop it, Debian will lose its trademark in the name Debian, and the company will be able to go on using its name forever. Is this desirable? I assumed not. Is preventing this Free? I assumed yes. Now, trademark law nowadays allows the regulation of trademark use pretty heavily. Debian probably wants to allow anything *except* for deliberate attempts to confuse people. Exactly to what extent we want to allow others to use the Debian name and logo to confuse people is another matter. :-) Therefore, using the Debian logo to claim endorsement by Debian without permission is already illegal, so the Debian logo license should not mention it. If I have accidentally made the restrictions broader than that in my proposed licenses, then it's a mistake which should be fixed. I belive that is what happened, yes. OK. :-) That's a good argument for separate trademark and copyright licenses; it prevents accidental contamination by sloppy writing such as mine. Well, since we're agreed on the copyright licensing terms, how about we go ahead and do that. Just put a This copyright license does not grant a trademark license disclaimer after your choice of standard license, and I think we're set, right? -- This space intentionally left blank.
Re: Clarifying non-free parts of the GNU FDL
Roger Leigh wrote: During discussion with gimp-print upstream about the potential problems with the GNU FDL and the possibility of relicensing it, a number of issues have cropped up, which I'd be grateful if you could assist with. I have pointed to Manoj's draft position statement as a summary of the issues with the FDL found to date, which we have been discussing. If the documentation was to remain GFDL licenced, would be possible to add a clarification to the licence in order to counter the main problems which would affect this work? The work is written in Docbook/SGML, and contains no invariant sections. Specifically, would it be possible to 1) Allow storage/transmission on encrypted filesystems/links to counter the DRM restriction? Should be; can we come up with a good clause? In addition, despite and contrary to the requirements in section 2 of the GFDL, the copyright holders grant you the right to use technical measures to obstruct or control the reading or further copying of (1) copies you make and do not distribute or (2) copies you distribute, provided that you also make an unobstructed, uncontrolled copy available to the same recipients. 2) Not require forcing distribution of transparent copies with bulk opaque copies? Should be; can we come up with a good clause? ;-) In addition, the copyright holders exempt you from the requirements in section 3 of the GFDL. If these clarifications were to be made, would the licence be considered DFSG-free? Um... I think so. Were there any other problem clauses? Are there any other possible amendments that could be made to make the licence DFSG-free? Lastly, are there any alternative licences available? The author (and copyright holder) of the work would prefer a licence suited to documentation rather than programs (which I don't disagree with). You do realize that GFDL'ed documentation with a GPL'ed program means that moving stuff between the documentation and the program is possible *only* for the copyright holder. This is tedious in the extreme when the program or documentation has multiple copyright holders. Note that there are multiple copyright holders whenever more than one person contributes creative material of more than about 15 lines, unless one of them is employing all the others (work-for-hire), or they signed written, paper copyright assignments. GFDL/GPL dual-licensing would be a good thing for that reason. -- This space intentionally left blank.
Re: Open Software License v2.1
On Wed, Sep 22, 2004 at 11:36:54PM +0100, Andrew Suffield wrote: On Wed, Sep 22, 2004 at 07:09:21AM -0400, Glenn Maynard wrote: On Wed, Sep 22, 2004 at 11:44:13AM +0100, Andrew Suffield wrote: Bob creates Emacs, under a claim patent infringement in this work and lose your license to it license, which includes GIF decoding. Joe derives XEmacs from that work. This inherits, among many other things, GIF decoding. Bill sues Joe, claiming that XEmacs infringes his GIF patent. Does and should Bill lose his license to Emacs, in addition to XEmacs? I think the answer to both is yes. The copyright and patent holder has no need for a license. Bill is not a copyright holder at all in this scenario. Not a very interesting scenario, then. You can construct a scenario where any license seems reasonable, including a proprietary one. The mark of free licenses is that you can't construct any where it's unreasonable. This scenario is constructed to respond to a specific message of Nathanael's, to argue that 1: for patent defense clauses to be useful, this type of license loss must occur, and 2: that this behavior is very similar to things we consider free. It is not a standalone example to argue that patent defense clauses are free. Please review the thread. -- Glenn Maynard
Re: Open Software License v2.1
On Wed, Sep 22, 2004 at 05:22:21PM -0400, Nathanael Nerode wrote: However, if he distributed under a non-patent-defense license, it would *still* be non-free. So I'm not clear on how the so-called patent-defence clause makes any difference here. It wouldn't be: Right but if the patent-defence clause is absent, then either: (1) GIFEnc is still free, because GIFCorp's patents will be defeated (In which case what's the need for the countersuit?) Or (2) GIFEnc isn't free We're back at do held patents make software non-free? issue again: we don't consider a work non-free simply because patents which can be enforced against it exist, unless they're actually being enforced. This fact does make the issue murky and much harder to figure out. So what difference does the patent-defence clause make? It's essentially applying the don't allege patent violation for use of this software part of the anti-patent clause to the copyright holder, as well, by a different means (a license grant). It's usually much cheaper, by my understanding, to countersue and settle a patent suit than to defend against the patent in court--even if you know you'll win because the patent is bogus. I don't think it's free to require that others give up their ability to countersue (eg. to take the above course of action) unless you guarantee that you won't make it necessary--this is true even if you don't hold any relevant patents, or if your patents wouldn't actually stand up in court. -- Glenn Maynard
Re: Open Software License v2.1
On Wed, Sep 22, 2004 at 10:27:11AM -0400, Michael Poole wrote: Before Debian considers software free, we require proper licenses for actively enforced patents; any claim of infringement would make the software non-DFSG, even before a lawsuit is resolved. This isn't established. If Microsoft started patent infringement suits against gcc, which it knew it would lose but started anyway for, say, weird PR reasons, we wouldn't consider gcc non-free. We'd only do that if they actually won the suit, and then only extend it to other software that would infringe the same patent if they actually won it on the grounds of the patent (eg. not in the odd case of the FSF throwing a no-contest, which wouldn't actually confirm the patent, I assume). We don't have much prior practice on this, but I don't think we'd be doing anyone any good to pull out large chunks of the archive because a company started a bogus patent campaign against major free software projects. -- Glenn Maynard
Re: Open Software License v2.1
On Wed, Sep 22, 2004 at 04:06:18PM -0400, Brian Thomas Sniffen wrote: I'm not sure it is, in this example. Well. It's important because this is all part of a crusade against software patents taken too far into a crusade against patents which happen to apply to software. I understand what you're saying, but I just don't yet think this is a crusade taken too far. Can you find anything in Debian's devotion to its users and free software, however, which enjoins the project to join in this crusade, not merely by lobbying governments but also by permitting restrictions on the behavior of licensees of allegedly free software? This seems obvious to me: these clauses are intended to protect users and free software. If Debian is to reject them, it should be because they're damaging to users or free software, by restricting freedoms that we consider important. I'm ready and willing to be convinced that there's some freedom being restricted here that shouldn't be, but I havn't been pointed to it yet. This is similar to the point of the desert island re-debate where people were asking what freedom does this protect? The answer was freedom to make private modifications. From that we could move on with the discussion of whether that was important, and if we disagreed, we could understand where. What freedom does this protect? It think it's free to terminate a public license completely and universally as soon as anybody brings and wins any suit against any party that claims that the work using some patented technology. This is fairly useless: again by my poor understanding, most patent suits are be settled out-of-court, since it's very expensive to do so in court. -- Glenn Maynard
Re: Open Software License v2.1
On Wed, Sep 22, 2004 at 10:02:10PM +0100, Henning Makholm wrote: It's not a complete defense, by the way - a smart patent owner would just try to sue everybody else *but* the copyright holder who uses the code instead. The OSL 2.1's clause causes termination if you allege patent violation in the work against any licensee. -- Glenn Maynard
Re: CeCILL again...
On Wed, Sep 22, 2004 at 11:46:17AM -0700, Josh Triplett wrote: Your requirement that you be able to understand the license terms is perfectly reasonable. That does not mean Debian should have that requirement. Consider this: it is also a perfectly reasonable Debian should not require that Debian be able to understand license terms? I just don't believe that having one or two trusted people fluent in a language is a substitute for the debian-legal list, where we have an entire list of such people. Debian is not capable of exercising the level of licensing caution it attempts to maintain, with licenses written in languages other than English. requirement that you be able to read the source of all the software you use, to ensure that it does not take undesirable actions. This might well cause you to avoid using software in programming languages you don't understand, or with comments and variable names in languages you don't understand. Nevertheless, such software is still Free. If Debian can't deterine with confidence that a work is legal and DFSG-free to distribute, it doesn't distribute. Debian doesn't have a similar policy with source: if a work is written in a strange programming language that only the upstream author and the package maintainer can comprehend, it's not going to be removed from the archive (as long as it really is source). This isn't the same, and isn't treated the same, as licensing. I also fear the general case, where a notable percentage of software in the archive has licenses which are only binding in a random language, and it becomes unreasonable to avoid that software. Currently, a fluent English speaker with some background in reading licenses should be able to understand /usr/share/doc/*/copyright (if d-legal can't make sense of an English license, it probably doesn't belong in the archive, either); that would break down. -- Glenn Maynard
Re: Bug#265352: grub: Debian splash images for Grub
On 2004-09-22 23:22:45 +0100 Nathanael Nerode [EMAIL PROTECTED] wrote: A trademark license *has* to prohibit such things. Prohibiting misrepresenting the origin of the *logo* doesn't suffice. We have to require that the logo, and anything confusingly similar, is not used to identify things which aren't Debian. Aren't the debian trademarks restricted to a specified scope? As I understand English law: if debian is a computing-related trademark, its licence should not forbid use of the debian logo for a distributor of replacement body parts, for example, because that is far beyond what the trademark gives. I think forgetting that is the root of trying to use copyright law to create supertrademarks. -- MJR/slefMy Opinion Only and not of any group I know Creative copyleft computing - http://www.ttllp.co.uk/ LinuxExpo.org.uk village 6+7 Oct http://www.affs.org.uk
Re: Open Software License v2.1
Brian Thomas Sniffen writes: Can you find anything in Debian's devotion to its users and free software, however, which enjoins the project to join in this crusade, not merely by lobbying governments but also by permitting restrictions on the behavior of licensees of allegedly free software? I think the benefit to free software is obvious: Someone who uses free software with the kind of termination clause in OSLv2.1 cannot restrict its use to himself by making patent infringement claims over it. PS You know, I just thought of something. If these clauses cancelled the copyright license to *everybody* as soon as *anybody* *wins* a patent lawsuit over the software, I wouldn't mind them so much. It's the cancellation of the license for even seeking impartial justice that bothers me. Why is the type of the withheld license important? If you are the patent holder, you would not have a copyright license on the software that infringes your patent. If you are anyone else, you would not have a patent license on the software. It would not be legal for anyone to distribute (or use, except perhaps for the patent holder) the software after a successful patent claim -- unless the patent holder granted a free license. Before Debian considers software free, we require proper licenses for actively enforced patents; any claim of infringement would make the software non-DFSG, even before a lawsuit is resolved. Michael Poole
Re: Open Software License v2.1
Raul Miller [EMAIL PROTECTED] writes: Raul Miller [EMAIL PROTECTED] writes: The claim that copyleft software isn't free is nonsense. On Tue, Sep 21, 2004 at 09:44:47PM -0400, Brian Thomas Sniffen wrote: Yes, but only you've made that claim. I certainly haven't, and I invite you to quote where you think I've done so. That isn't my claim. Instead, I said that, as phrased, the following seemed to include that claim, and I asked for a phrasing which was not so inclusive: On Tue, Sep 21, 2004 at 03:35:16PM -0400, Brian Thomas Sniffen wrote: Because it's a copyright license. If I give away all these freedoms with respect to my work, then I should really be giving them away. If I'm only giving them away contingent on others with rights to the work giving theirs, I should negotiate that in an appropriately smoky back room -- and until all those show up freely, the software isn't free. If you have a patent which applies to the program, and you want to distribute, you have to grant a license to that patent to all third parties. Having just read GPL 7 again, to make sure I'm doing this right, I think I can just issue a license to use the patent for a given purpose, and that's enough. I'm not completely certain of that. GPL 7 isn't relevant here. GPL 7 is for cases where someone else holds the patent. [Note the uses of the phrase imposed on you and the phrase do not excuse you.] Try GPL 5 and 6, instead. Right. So I set up a little corporation whose purpose is to hold the patents, and I deal in the copyrights. Now we're back in GPL 7 land. I'm sure I only have to grant a license to that patent to all parties who receive a copy of the work. You can't restrict that grant to only those parties -- they must be able to further distribute the program without any further restrictions. Why must they be able to do that? I can distribute in compliance with GPL 6, and then they get stopped by GPL 7 through no fault of mine. The scenario you seem to propose looks to me like this: First, B files for a patent and publishes an invention. For the sake of argument, say it's something legitimately innovative. If the software is not free, regardless of the copyright license, then the reason it's not free is not the copyright license. Thus, this scenario has no bearing on the freeness of the license. I don't think that's true. Certainly, I see no reason it should be accepted as obviously true. A writes some software, and GPLs it. B claims that the software is patent restricted, and sues A. B wins, and now only B can distribute the software -- A can't [and no one else can] without buying a license from B. You seem to be claiming that a license which prevents this scenario is not good, and that the reason it's not good is that it prevents this scenario. It might be good. It isn't free. If it is free, why isn't the following free: : A writes some software, and GPLs it. B claims that the software is : on his hard drive, and sues A for that drive. B wins, and now only : B can distribute the software -- A can't [and no one else can] : without getting a license from B. This scenario of yours has nothing to do with the freeness of the license. It's just your patent scenario from above, rephrased to deal with physical property. Are you claiming that a license which prevents this scenario, by terminating the copyright license to anyone who sues regarding physical media containing the software, is free? -- Brian Sniffen [EMAIL PROTECTED]
Re: Open Software License v2.1
GPL 7 isn't relevant here. GPL 7 is for cases where someone else holds the patent. [Note the uses of the phrase imposed on you and the phrase do not excuse you.] Try GPL 5 and 6, instead. On Wed, Sep 22, 2004 at 10:39:38AM -0400, Brian Thomas Sniffen wrote: Right. So I set up a little corporation whose purpose is to hold the patents, and I deal in the copyrights. Now we're back in GPL 7 land. Only if you've given up control of the corporation, to the degree you have no influence over the corporation. Even there, if it can be shown that you set up the corporation for the purpose of making this happen, it's not likely that any superficial or temporary lack of control would be relevant. That said, if this tactic worked for the GPL, it would work just as well for the OSL -- for all I know, your tactic might even work better in the context of the OSL. Oh, look, it's not me that's lost rights to distribute your program that I'm claiming is illegal -- it's this corporation that owns my patents that's suing you. I'm sure I only have to grant a license to that patent to all parties who receive a copy of the work. You can't restrict that grant to only those parties -- they must be able to further distribute the program without any further restrictions. Why must they be able to do that? I can distribute in compliance with GPL 6, and then they get stopped by GPL 7 through no fault of mine. It's not through no fault of yours if it's your restriction that causes them to get stopped by GPL 7. And, if it's not your restriction that causes them to get stopped by GPL section 7, how could you possibly sue the copyright holder for patent violations if the license were osl-2.1 instead of the gpl? The scenario you seem to propose looks to me like this: First, B files for a patent and publishes an invention. For the sake of argument, say it's something legitimately innovative. If the software is not free, regardless of the copyright license, then the reason it's not free is not the copyright license. Thus, this scenario has no bearing on the freeness of the license. I don't think that's true. Certainly, I see no reason it should be accepted as obviously true. Where, specifically, do you disagree? [Let's take it for granted that if you disagree with an antecedent that you feel that the consequent is illogical.] A writes some software, and GPLs it. B claims that the software is patent restricted, and sues A. B wins, and now only B can distribute the software -- A can't [and no one else can] without buying a license from B. You seem to be claiming that a license which prevents this scenario is not good, and that the reason it's not good is that it prevents this scenario. It might be good. It isn't free. If it is free, why isn't the following free: : A writes some software, and GPLs it. B claims that : the software is on his hard drive, and sues A for : that drive. B wins, and now only B can distribute : the software -- A can't [and no one else can] : without getting a license from B. This scenario of yours has nothing to do with the freeness of the license. It's just your patent scenario from above, rephrased to deal with physical property. That's not my patent scenario, that's my paraphrase of your patent scenario. I'm claiming that the scenario is invalid, your rephrasing of it didn't make it valid. -- Raul
Re: Open Software License v2.1
Andrew Suffield writes: On Tue, Sep 21, 2004 at 07:29:10PM -0400, Michael Poole wrote: If you distribute a program under the GPL, you lose most or all of your ground to claim damages in court on the basis that the program infringes your patent or copy rights. (GPL sections 5, 6, et al.) Word games. If you license something then you lose the ability to sue people for acting in the manner you licensed them to do. Don't waste my time; you know full well that's irrelevant. You want to protect someone's right to privately use and modify software (that someone else wrote and tried to release for the world) after they file a lawsuit that -- if successful -- prohibits everyone else in the world from using that software? On the basis that to do otherwise is not free? Even if your answer to both is yes (my answer to the first is no), I am still not sure how you distinguish a patent lawsuit that claims the software infringes a patent from sublicensing except as expressly provided under this [General Public] License. Michael Poole
Re: Open Software License v2.1
On Wed, Sep 22, 2004 at 12:02:49PM +0100, Andrew Suffield wrote: Word games. If you license something then you lose the ability to sue people for acting in the manner you licensed them to do. Don't waste my time; you know full well that's irrelevant. How is that irrelevant? If agreement not to sue represents a subset of the consequences of granting a license, how can agreement not to sue be non-free when granting a license is free? [I'll grant that there might be some reason why agreement not to sue is non-free, but I've not seen those reasons described yet, and I see nothing to make me believe that what you're claiming is word games is at all irrelevant.] By the way, word games which are misleading can be dealt with by pointing out how they're misleading. And, word games which are illogical can be dealt with by pointing out how they are illogical. That leaves word games which are neither misleading nor ilogical... but... Claiming that a relevant point is irrelevant is an example of misleading. Indicating that a relevant point is a waste of your time seems illogical -- it would be better to just drop out of the discussion if that were really the case. -- Raul
Re: Open Software License v2.1
Nathanael Nerode wrote: Consider the Malicious Software Corporation (MSC). Consider work X by author Joe. MSC holds patent A covering X and patent B covering something else. Valiant Defender (VD) holds patent C covering X. Normally, MSC can sue any user of X for infringing patent A. With the narrow patent-lawsuit-termination clause, doing so will cause MSC to lose its copyright license for X. If MSC doesn't copy/distribute/modify X, it won't care. If MSC does, then this might hurt it a bit. Suppose MSC sues VD. VD could try to fight back with patent C; this would only work if MSC uses X, of course. With the narrow patent-lawsuit-termination clause, this fighting back would cause VD to lose its copyright license for X as well. Would a counterclaim regarding patent C be a compulsory counterclaim in that case? That is, if VD does not bring patent C into the lawsuit, would VD be barred from bringing a future suit against MSC regarding patent C? If so, then VD has two options: give up their copyright license for X, or essentially grant MSC a free license for patent C. (Of course, Joe could forgive the copyright license violation and allow VD to continue working with X) I suppose some people might consider that to be bad for free software. (I don't.) Now consider the case where MSC buys Joe's copyright. What changes? With the narrow patent-lawsuit-termination clause, MSC will retain its ability to copy/distribute/modify X. This will not change its ability to sue anyone over patents A or B. This will not change VD's ability to fight back with patent C (or not), which depends on whether MSC uses X. This will not change whether VD's copyright license for X is terminated or not. (Essentially, by buying the copyright, they would have gotten themselves a special license to avoid the patent-termination clause, and that's all.) Here, If a counterclaim regarding patent C is compulsory, VD can not expect the copyright holder (Joe in the above example, now MSC) to forgive the copyright license violation, so that violation means no more X (ever). Or VD could leave patent C out of the suit, which essentially grants MSC a free license to C. --Joe
missing licenses in gnulib
[ putting debian-legal on CC ] Hi! I'm trying to prepare a Debian package of gnulib, but there seems to be some legal problems we should sort out first. According to the COPYING file, we can't assume GPL for any of the files in the source tree. This is a problem for files that are not explicitly licensed. I've made an exhaustive list in the lib and tests directories: lib/alloca.c lib/atanl.c lib/diacrit.c lib/dirfd.h lib/getpagesize.h lib/lbrkprop.h lib/logl.c tests/test-stpncpy.c The legal status of these is questionable and in some cases they're being distributed illegaly (since copyright holder is not FSF). I have omitted from the check those files that were 15 lines in size since these could be considered not significant for copyright purposes. The worst problem, however, is in the m4 and modules directories, where most of the files are unlicensed. Typicaly, these macros would be implicitly licensed by a COPYING file, but there's no COPYING file covering them. This situation makes it illegal to use most of gnulib's macros in any program (even GPLed ones). There's also the problem with non-free documentation in doc directory (3 files), but I'm aware that for the FSF freedom isn't important for documentation so I'm ommiting the list here. -- .''`. Proudly running Debian GNU/kFreeBSD unstable/unreleased (on UFS2+S) : :' : `. `'http://www.debian.org/ports/kfreebsd-gnu `-
Re: Open Software License v2.1
Glenn Maynard [EMAIL PROTECTED] writes: On Mon, Sep 20, 2004 at 01:33:17PM -0400, Brian Thomas Sniffen wrote: real invention, whether implemented in software or hardware. The RSA cryptosystem is a decent example of this. So there are some legitimate patents, though they're probably a minority. But that means that those people do have a legitimate I don't believe any patent enforcement against software is legitimate; I believe that enforcement of a patent against RSA code is just as destructive and abusive as enforcement against XOR mouse cursor code. (I also concur with Michael's response.) (This may be something close to the core of where we disagree.) I disagree with you, but I suspect that your version of patent law would be better than what we've got right now. So by all means, go forth and lobby the various world governments to change patent law. Can you find anything in Debian's devotion to its users and free software, however, which enjoins the project to join in this crusade, not merely by lobbying governments but also by permitting restrictions on the behavior of licensees of allegedly free software? -Brian PS You know, I just thought of something. If these clauses cancelled the copyright license to *everybody* as soon as *anybody* *wins* a patent lawsuit over the software, I wouldn't mind them so much. It's the cancellation of the license for even seeking impartial justice that bothers me. -- Brian Sniffen [EMAIL PROTECTED]
Re: Bug#265352: grub: Debian splash images for Grub
Nathanael Nerode wrote: Josh Triplett wrote: Both of these licenses seem clearly non-free to me, since they restrict the uses of unmodified or insufficiently different versions. Only to the extent of prohibiting misrepresentation of other works, projects, and organizations as belonging to/being endorsed by/being part of Debian. That's a standard, acceptable class of restrictions, isn't it? This really *is* about misrepresentation and nothing more. If that is truly the case, then you don't need to say that at all in the license. See http://lists.debian.org/debian-legal/2004/05/msg00540.html and the resulting thread. Claiming endorsement by someone else without their permission is already illegal. Therefore, using the Debian logo to claim endorsement by Debian without permission is already illegal, so the Debian logo license should not mention it. If I have accidentally made the restrictions broader than that in my proposed licenses, then it's a mistake which should be fixed. I belive that is what happened, yes. Your proposed licenses did not permit any use of an unmodified or insufficiently modified logo, except to refer to Debian. Instead, what you probably want is something similar to the clauses from the zlib license: The origin of this software must not be misrepresented and Altered source versions must be plainly marked as such. As for not permitting others to claim endorsement by or affiliation with Debian, that is already not permitted. Anyway, as someone else said, if you don't defend your trademark against those particular uses, you don't have a trademark anymore. The original and primary purpose of a trademark is to identify a business/product distinctively. Preventing uses which cause confusion as to the identity or source of your product or business is what you have to do in order to keep your trademark at all. That's completely unrelated to whether the suggested licenses are DFSG-free or not. - Josh Triplett signature.asc Description: OpenPGP digital signature