On Wed, Jul 14, 2004 at 09:21:16AM -0400, Raul Miller wrote: > Here's a recap of one point in subthread: > > > > > > > > > This clause violates the intent of DFSG 1, in my > > > > > > > > > opinion. "The license may not require a royalty > > > > > > > > > or other fee for such sale." [...] > > > > > It's ok to say: here's the big problem, and here's some other > > > > > areas of concern that you might want to think about. > > > > > > It's misleading to say "this is a problem" when we accept > > > > > licenses as DFSG even though they have "this problem". > > Branden: > > > > Which licenses do we accept as DFSG-free even though they have > > > > "this problem"? > > Raul: > > > That's for you to say. > > Branden: > > No; it is you who has asserting that 'we accept licenses as DFSG[-free] > > even though they have "this problem"'. The burden of proof is on the > > affirmative. > > Ok if you want to focus on that aspect, I've included enough material > in this thread to show you what you originally said, and the way you > said it. All right. Which licenses to we accept as DFSG-free even though they violate DFSG 1? > You later presented your point of view in a much more coherent form. > I agree with that presentation. For reference, it's: > > The GNU GPL *does* require that you transmit something of > value -- or an offer to do so -- alongside *other* things > of value when you distribute them to another party. > Best Practical's license has an invisible hand that > reaches into your life from afar and asserts property > claims over something that would otherwise be yours. > > That doesn't mean that I agree with the original statement of your > argument. > > I'm not going to respond to the rest of your message in this fashion, > because I don't have the time or interest to show you the context of > what you're saying and how what you're saying does or doesn't relate to > that context. The Best Practical license issue has been voided; I've been in personal contact with Jesse Vincent and Andrew Stribblehill, and I've quite persuaded that the intent of Best Practical's license rider was not to violate DFSG 1, nor even to become incompatible with the GNU GPL. If one were to argue that I'm applying a hyper-aggressive interpretation of DFSG 1, this case would not be the best example, as it turns out the licensor in question isn't actually interested in failing DFSG 1 even by the lower threshold that (presumably) you have set. Thus my interest in eliciting from you your opinion as to which licenses we have accepted as DFSG-free even though they violate DFSG 1 by *my interpretation*. I don't know of any, personally. To characterize my understanding of DFSG 1 as extreme or heretical, it would be helpful for you to illustrate how it would have practical consequences on Debian today. (I wonder how, in the absence of dramatic effect, "extremism" is to be distinguished from "moderation"?) > > I reiterate: DFSG-freeness is not predicated on GPL-compatibility. > > It is in the context of GPL licensed programs. Are you talking about pure GPL, or GPL + riders? If the former, I agree -- but then that's not a useful observation in a thread about Best Practical's license, which was not pure GPL. If the latter, then I disagree. While most uses of the GPL with additional restrictions (additional permissions don't affect GPL-compatibility), result in software being unredistributable, this is not *necessarily* the case. Someone can write such a thing and simply not include or link against any pure-GPLed works. Provided the additional restriction did not fail the DFSG in and of itself, I don't see why such a license necessarily would fail the DFSG. We'd have to judge this sort of situation on a case by base basis. Unless -- we want to assert that all GPL-derived licenses used in Debian must be GPL-compatible. That is the logical consequence of your statement that "[DFSG-freeness is predicated on GPL-compatibility] in the context of GPL licensed programs." If that's the consensus view of this mailing list, I can go along with it, but I think it would come as a surprise to some people, and we should probably make an announcement about it. Heretofore I would not have had the courage to assert that using a GPL + extra restriction license is non-free Just Because. > > > Note also that in this case we're talking about a license which > > > represents itself as GPL compatible. > > > > No, it does not "represent itself as GPL compatible". It represents itself > > as being the GNU GPL version 2, with a rider tacked onto it. > > That's not a meaningful distinction. It is to me, and to the licensors who thought their licenses were GPL compatible even when they *did* tack extra-restriction riders onto them. You can't both be GPL-compatible and not GPL-compatible. Either your rider affects GPL-compatibility or it does not. If you add extra permissions, you're GPL-compatible. If your rider affects no activity within the GNU GPL's purview, you're GPL-compatible. If you add extra restrictions, you're not. Unless, of course, by "GPL", you mean, "all licenses that incorporate the GNU GPL into their text in whole or part" -- but I personally would find that usage confusing. > > We've long known that tacking riders onto the GNU GPL that aren't in the > > form of grants of additional permissions are problematic. Let me know if > > you need me to trawl the archives of this list to support this statement, > > but for the moment I'm going to assume it's common knowledge. > > This was not a grant of additional permission. I agree. It was either an additional restriction, or did not intersect with the copyright license at all. (We've since learned that the latter was intended.) I don't see how you can simultaneously assert that there is: * not a meaningful distinction between asserting GPL-compatibility and using the GPL as the basis for your license, upon which you add conditions that are not in the GPL itself; * assert that there is a distinction between granting additional permissions and imposing additional restrictions as terms of the license I'd agree that riders wouldn't matter if they couldn't make a work incompatible with the GPL. But they can. > > > Agreed. And the Best Practical license declares itself GPL compatible. > > > > No, it doesn't. Since you're harping on this, let's review. > > ... > > # This work is made available to you under the terms of Version 2 of > > # the GNU General Public License. > ... > > > I can't find any assertion of GPL compatibility in the above. > > Ok. In the future, You might try reading the whole section indicated as "LICENSE BLOCK". > > The term does not even appear. Isn't it a bit careless to put words in the > > licensor's mouth? > > You seem to be confusing spelling with meaning. I disagree that I am confused. I don't think it is intellectually challenging to understand that there is a difference between the GNU GPL and the GNU GPL plus additional terms. > > > > If I haven't expressed myself clearly, then it's quite likely you don't > > > > understand me, no? > > > > > > Exactly. > > > > Would you respond to the remainder of my message, please? > > What value would this have for debian-legal? That depends on how much of the list's time you intend to waste in the future claiming that there is an implicit assertion of GPL-compatibility in any license that tacks extra terms onto the GPL. While it may *often* be the case that a licensor intends to be GPL-compatible when imposing additional restrictions, it is not *necessarily* the case. When contacting upstream copyright holders for clarification, it is wise to prejudge the issues as little as possible, and instead ask them sincere questions about their intent. We are less likely to ask sincere questions about GPL-compatibility if we've already made up our minds that that's what people want. There are other presumptions it's good not to make, such as "the copyright holder gives a whit about Debian" or "the copyright holder will feel insulted if Debian ships a package of his or her software in the non-free section instead of main". The latter is a point of particular importance, since it is very frequently a conclusion leapt to by that portion of our membership who simultaneously feel that the non-free section is both essential to the viability of our OS distribution, and also some sort of stigma. > More specifically, which points do you consider important? I'll be happy > to focus on specific issues, but I don't want to robotically respond to > every single sentence you've written. I don't want you to robotically repond either, and I reckon if you can't work up the enthusiasm for anything more useful than that, I withdraw my request. -- G. Branden Robinson | Debian GNU/Linux | If ignorance is bliss, branden@debian.org | is omniscience hell? http://people.debian.org/~branden/ |
Attachment:
signature.asc
Description: Digital signature