ARToolKit | Mailing List Archive |
![]() |
From: | Michael Sparks <zathras@t ............> | Received: | Sep 26, 2004 |
To | artoolkit@h .................. | ||
Subject: | GPL AR Toolkit? | ||
Hi, I'm looking for something like the AR Toolkit that is GPL (or BSD, etc), rather than the license AR Toolkit is distributed under. The reason is because I'd prefer to write code in a high level scripting language under linux (I'm prepared to write the bindings). (I've seen the matlab GPL reimplementation but I don't use Windows or have Matlab :) The AR Toolkit doesn't allow redistribution, but hits precisely what I want to do hence why I'm asking the question here... NB, since I'd just be toying around I fall firmly in the non-commercial group, but that's not the issue. Thanks in advance for any comments, and apologies in advance if this an FAQ. Best Regards, Michael. (BTW, the reason I say that the AR Toolkit doesn't allow redistribution is because it's initially distributed under the clause "GPL for non-commercial use". Whilst I can recieve a copy, the GPL specifically prohibits me as a licensee not a licensor from redistributing under those terms - I'm not allowed to impose the additional restriction - meaning I cannot redistribute at all. (A licensor can do this, but as a recipient I can't)) |
From: | Michael Sparks <zathras@t ............> | Received: | Sep 27, 2004 |
To | Wayne Piekarski <wayne@c ..............> | ||
Subject: | Re: GPL AR Toolkit? | ||
On Mon, 27 Sep 2004, Wayne Piekarski wrote: > We had a discussion about licensing for ARToolkit on this mailing list a > while ago I think. I can't remember what we were discussing exactly though > but there might be a mirror of these discussions somewhere. I did find that discussion, but it doesn't really help since as you say... > If something is licensed as GPL, you cannot impose extra restrictions > because the GPL prevents you from adding any extra conditions. ... is true if I recieve a GPL'd work from someone. When I redistribute/etc the work I have to stick to every part of the GPL, for obvious good reasons. However if a copyright owner says "you may only redistribute this if you paint your nose brown", I can only redistribute under that term (or not at all - I'm not obligated to redistribute). However if it says "you may only redistribute this at a height of 10m whilst also being at a height of 20m", there is no way I can comply with that license. Back to this example, a copyright owner can (rightfully :) choose any license they wish to limit redistribution - which in this case the copyright owner has chosen GPL with an extra limitation. Since they own the copyright and a license only enables things that would be illegal otherwise, copyright law takes precedence and the *originator* is allowed to state this, and any recipient must comply. The problem is this - the license as supplied states: * You may only use this under the terms of the GPL *and* only for non-commercial purposes. Since I'm not the copyright owner I am (rightfully) obligated to pass on this license to anyone I pass it onto. The problem is this license is GPL+restriction, which since I'm not the copyright owner I'm tied to or tied to copyright law (no redistribution). However part of that license says GPL+no restriction, the full license I recieve says GPL+restriction which is the same situation as 10metres and 20 metres, therefore I can't redistribute - even non commercially. I'm not interested in changing someone else's decision regarding their software, since, well, it's theirs, but since I know there's some bright people on this list (the software is proof :-) I wondered if anyone knew of a GPL equivalent, even if it's not as good. That was why I asked the question I did. > When they say non-commercial use, they mean that you cannot distribute > binary only files and charge money for it. > The GPL prohibits you from doing > this anyways. I'm well aware of this, I'm not a lawyer and the above isn't legal advice etc, but I've read a few books on the topic to know that staying away from GPL+restriction is a bad idea, however I still think the software is cool and wondered if anyone had flattered with imitation :-) > I hope this helps. It does, thanks. Best Regards, Michael. |
From: | Wayne Piekarski <wayne@c ..............> | Received: | Sep 27, 2004 |
To | Michael Sparks <zathras@t ............>, artoolkit@h .................. | ||
Subject: | Re: GPL AR Toolkit? | ||
We had a discussion about licensing for ARToolkit on this mailing list a while ago I think. I can't remember what we were discussing exactly though but there might be a mirror of these discussions somewhere. If something is licensed as GPL, you cannot impose extra restrictions because the GPL prevents you from adding any extra conditions. When they say non-commercial use, they mean that you cannot distribute binary only files and charge money for it. The GPL prohibits you from doing this anyways. The GPL does allow you to take a GPL'd program and sell it for money, as long as you give all the source code away and keep the mods under GPL as well. However, the person you sell it to can give copies away for free if they like, and you can't stop this. So the GPL automatically prevents the forms of commercial use that the authors intended. I hope this helps. regards, Wayne --------------------------------------------------------------------------- Dr Wayne Piekarski - Researcher / Lecturer pho: +61-8-8302-5070 fax: +61-8-8302-3381 Assistant Director - Wearable Computer Lab mob: +61-407-395-889 School of Computer and Information Science ema: wayne@c .............. University of South Australia www: http://www.tinmith.net Michael Sparks wrote: > Hi, > > > I'm looking for something like the AR Toolkit that is GPL (or BSD, etc), > rather than the license AR Toolkit is distributed under. The reason is > because I'd prefer to write code in a high level scripting language under > linux (I'm prepared to write the bindings). (I've seen the matlab GPL > reimplementation but I don't use Windows or have Matlab :) > > The AR Toolkit doesn't allow redistribution, but hits precisely what I > want to do hence why I'm asking the question here... NB, since I'd just be > toying around I fall firmly in the non-commercial group, but that's not > the issue. > > Thanks in advance for any comments, and apologies in advance if this an > FAQ. > > Best Regards, > > > Michael. > > (BTW, the reason I say that the AR Toolkit doesn't allow redistribution > is because it's initially distributed under the clause "GPL for > non-commercial use". Whilst I can recieve a copy, the GPL specifically > prohibits me as a licensee not a licensor from redistributing under those > terms - I'm not allowed to impose the additional restriction - meaning I > cannot redistribute at all. (A licensor can do this, but as a recipient I > can't)) > |
From: | "Wayne Piekarski" <wayne@c ..............> | Received: | Sep 27, 2004 |
To | "Michael Sparks" <zathras@t ............> | ||
Subject: | Re: GPL AR Toolkit? | ||
> ... is true if I recieve a GPL'd work from someone. When I > redistribute/etc the work I have to stick to every part of the GPL, for > obvious good reasons. > > However if a copyright owner says "you may only redistribute this if you > paint your nose brown", I can only redistribute under that term (or not at > all - I'm not obligated to redistribute). However if it says "you may > only redistribute this at a height of 10m whilst also being at a height of > 20m", there is no way I can comply with that license. > > Back to this example, a copyright owner can (rightfully :) choose any > license they wish to limit redistribution - which in this case the > copyright owner has chosen GPL with an extra limitation. Since they own > the copyright and a license only enables things that would be illegal > otherwise, copyright law takes precedence and the *originator* is allowed > to state this, and any recipient must comply. > > The problem is this - the license as supplied states: > * You may only use this under the terms of the GPL *and* only for > non-commercial purposes. > > Since I'm not the copyright owner I am (rightfully) obligated to pass on > this license to anyone I pass it onto. The problem is this license is > GPL+restriction, which since I'm not the copyright owner I'm tied to or > tied to copyright law (no redistribution). However part of that license > says GPL+no restriction, the full license I recieve says GPL+restriction > which is the same situation as 10metres and 20 metres, therefore I can't > redistribute - even non commercially. > > I'm not interested in changing someone else's decision regarding their > software, since, well, it's theirs, but since I know there's some bright > people on this list (the software is proof :-) I wondered if anyone knew > of a GPL equivalent, even if it's not as good. > > That was why I asked the question I did. > > > When they say non-commercial use, they mean that you cannot distribute > > binary only files and charge money for it. > > The GPL prohibits you from doing > > this anyways. > > I'm well aware of this, I'm not a lawyer and the above isn't legal advice > etc, but I've read a few books on the topic to know that staying away from > GPL+restriction is a bad idea, however I still think the software is cool > and wondered if anyone had flattered with imitation :-) What you have mentioned above is not correct. There is an FAQ on the GPL from the Free Software Foundation at http://www.gnu.org/licenses/gpl-faq.html - there are quite a lot of questions here. Anyways, there is one question about integrating with the BSD with advertising clause: http://www.gnu.org/licenses/gpl-faq.html#OrigBSD The original BSD license is incompatible with GPL because it requires the application to have an "advertising" printf at the startup announcing where the program came from originally. These advertising clauses are a restriction and the GPL does not not allow this, so you can't combine GPL with Advertising-BSD licenses. However, the newer BSD license does not require advertising, and so is compatible with the GPL. Thats why Linux uses lots of BSD code all over the place. Now from the GPL, it includes the following: (see the above links) "You may not impose any further restrictions on the recipients' exercise of the rights granted herein." Even if you are the copyright owner you cannot do this. The GPL explicitly forbids extra license terms being added, and you cannot distribute software under two licenses simultaneously that are incompatible. The owners of software may distribute code under multiple licenses, ie, they can give out a BSD version, and then separately give out a GPL version, but you cannot give out GPL+modified versions. Its either one or the other but not both. Since the GPL is a copyrighted work you are not allowed to play with the wording, and the clause above prohibits extra restrictions. So therefore the GPL prevents these kinds of modifications. There was recently a massive argument over XFree86 which is a common part of Linux and BSD distros. The XFree86 people changed the BSD license in their code to add an advertising clause. All the Linux distributors had issues with this because they use so much other GPL code that there would be a conflict between the GPL and BSD+Advertising license when it was all brought together. So everyone switched to X.Org which is still under the normal BSD and now XFree86 is basically dead. So this is an example of where the GPL just cannot allow modified licenses. By distributing new versions of XFree86 the Linux distro people would be violating the GPL. When we discussed this "non-commercial use" thing on the mailing list a while ago, I think we basically came to the conclusion that the original authors (Mark, Hiro, et al) just wanted to stop people from selling binaries without source code, which is what the GPL does anyways. I think there was an email from Mark actually saying this. So ARToolkit is just plain GPL, and so whatever the GPL permits is what you can do. If you want to distribute the code commercially, you can email the original authors and pay money to acquire it under a different license, but then it will not be GPL'ed at all, and instead something totally different. They can do this because they own the copyright. I hope this helps to explain things a bit better ... regards, Wayne |
From: | Michael Sparks <zathras@t ............> | Received: | Sep 27, 2004 |
To | Wayne Piekarski <wayne@c ..............> | ||
Subject: | Re: GPL AR Toolkit? | ||
On Mon, 27 Sep 2004, Wayne Piekarski wrote: ... > What you have mentioned above is not correct. There is an FAQ on the GPL > from the Free Software Foundation at > http://www.gnu.org/licenses/gpl-faq.html - there are quite a lot of > questions here. Anyways, there is one question about integrating with the > BSD with advertising clause: > http://www.gnu.org/licenses/gpl-faq.html#OrigBSD I'm (very) well aware of the GPL FAQ. What you miss is this: * A license is a permission to do stuff you're not able to do otherwise. * The copyright owner of a work they choose to release under the GPL is NOT bound by the GPL - since if they default to copyright law they get to set all terms of redistribution. * Therefore the original owner *can* create a license impossible to comply with. (Such as GPL+restriction.) > The original BSD license is incompatible with GPL because it requires the > application to have an "advertising" printf at the startup announcing where > the program came from originally. These advertising clauses are a > restriction and the GPL does not not allow this, so you can't combine GPL > with Advertising-BSD licenses. However, the newer BSD license does not > require advertising, and so is compatible with the GPL. Thats why Linux uses > lots of BSD code all over the place. > > Now from the GPL, it includes the following: (see the above links) > > "You may not impose any further restrictions on the recipients' exercise of > the rights granted herein." > > Even if you are the copyright owner you cannot do this. The GPL explicitly > forbids extra license terms being added, and you cannot distribute software > under two licenses simultaneously that are incompatible. *I* cannot. A copyright owner isn't bound by the license however, and hence _can_ stipulate that two incompatible licenses must be adhered to. The fact that no one can comply is neither here nor there - the *owners* aren't limited. > Since the GPL is a copyrighted work you are not allowed to play with the > wording, and the clause above prohibits extra restrictions. So therefore the > GPL prevents these kinds of modifications. No, it doesn't. Copyright covers modification, redistribution, works. Simple reference is none of those, hence the phrase (as used...) of GPL+modifications isn't itself a breach of the GPL, and hence isn't a breach. The only thing is *does* prevent is me taking the GPL, changing a few words for restrictions and working from there. (A good thing) > There was recently a massive argument over XFree86 which is a common part of > Linux and BSD distros. I'm more than aware of this. I've also been involved in similar arguments in the past, and in this case I'm simply asking questions early to avoid getting burnt again. Specifically I asked if anyone knew of something for Linux which was GPL (or BSD, or...) etc. IMO simply restating points which are plain false (the fact that the AR Toolkit's license isn't possible to comply with) doens't really help. I'm not going to try and change your view (or the license). If no-one knows of something, that's fine, I'll play with something else :-) > When we discussed this "non-commercial use" thing on the mailing list a > while ago, I think we basically came to the conclusion that the original > authors (Mark, Hiro, et al) just wanted to stop people from selling binaries > without source code, which is what the GPL does anyways. I think there was > an email from Mark actually saying this. So ARToolkit is just plain GPL, I agree that plain GPL protects the uses that Mark, Hiro, et al wish to prevent, FWIW. However what they want, and what is stated are two different things... What you claim and what is stated and the case are two different things. I'm not interested in an argument about whether it is or isn't. It isn't. You won't convince me of otherwise because I've seen this several times in the past, and just wondered if anyone had encountered something even half as good that was plain/pure GPL and stated explicitly as such. I also doubt I'll convince you of otherwise either, ce le vie :-) > I hope this helps to explain things a bit better ... I'm aware of the discussion, however given the website does currently state GPL+restrictions, which as copyright owners they *are* allowed to do (simple fact of life - they own the initial copyright - simple referencing the GPL + a restriction is valid), and the download essentially says the same thing I'm in either one of two situations: * The license as stated is invalid (in which case I can't redistribute any changes). * The license is valid but self contradictory preventing it being used. I'm aware of the reasons the AR Toolkit team want to limit closed usage (which makes a lot of sense), but with the current ambiguity I think I'm best off using something else and getting out of people's hair. It's no big deal though. Since this list seems to think that the license is something it isn't (*intent* isn't the same thing) and since this discussion isn't going anyway I'll just take my leave and apologise for wasting people's time, and find something else instead. BTW, *thank you* for taking the time to answer - I really don't agree with your conclusions, probably because I've been burnt in similar situations in the past. Best regards to all, Michael. |
From: | Stephen DiVerdi <sdiverdi@c ..........> | Received: | Sep 28, 2004 |
To | Wayne Piekarski <wayne@c ..............> | ||
Subject: | Re: GPL AR Toolkit? | ||
> Thats the only reference I've really found to > how it should be distributed. So I read it as being, "this code is GPL, > which will stop you from using it commercially because of the GPL itself". just wanted to chime in here for a moment. the GPL does _not_ prohibit commercial use of the ARToolkit. i mentioned a while back that i think the problem is the ARToolkit has two personalities - that of a library and that of a tool. as a library, the GPL limits the feasibility of commercial use as applications developed with the library cannot be sold binary only. that's been elaborated upon. however, as a tool, the GPL does not in any way prohibit a commercial entity from using the ARToolkit to produce an interactive multimedia presentation and generate money with it (for an easy parallel, the GIMP, an image manipulation program released under the GPL, can be used to create images used in advertising, on commercial websites, or for touchup of photographs for sale - it can even be used to create an entire image from scratch, which the artist has full rights to). this is because of the authoring capabilities of the VRML integration that allow someone to use the ARToolkit _as an application_, with no interest in the actual code. as i understood, this was the sticking point and why the "for non-commercial use" clause was added. -steve diverdi -sdiverdi@c .......... |
From: | Wayne Piekarski <wayne@c ..............> | Received: | Sep 28, 2004 |
To | Michael Sparks <zathras@t ............> | ||
Subject: | Re: GPL AR Toolkit? | ||
> I agree that plain GPL protects the uses that Mark, Hiro, et al wish to > prevent, FWIW. However what they want, and what is stated are two > different things... What you claim and what is stated and the case are two > different things. I'm not interested in an argument about whether it is or > isn't. It isn't. You won't convince me of otherwise because I've seen this > several times in the past, and just wondered if anyone had encountered > something even half as good that was plain/pure GPL and stated explicitly > as such. I also doubt I'll convince you of otherwise either, ce le vie :-) > > I'm aware of the discussion, however given the website does currently > state GPL+restrictions, which as copyright owners they *are* allowed to do > (simple fact of life - they own the initial copyright - simple referencing > the GPL + a restriction is valid), and the download essentially says the > same thing I'm in either one of two situations: > * The license as stated is invalid (in which case I can't redistribute > any changes). > * The license is valid but self contradictory preventing it being used. > > I'm aware of the reasons the AR Toolkit team want to limit closed usage > (which makes a lot of sense), but with the current ambiguity I think I'm > best off using something else and getting out of people's hair. It's no > big deal though. I think probably the best thing to do is get this clarified by the authors of the program. I am not a lawyer of course, but when I read the wording on the home page it says its "ARToolKit is distributed free for non-commercial use under the GPL license". Thats the only reference I've really found to how it should be distributed. So I read it as being, "this code is GPL, which will stop you from using it commercially because of the GPL itself". This is why I believe that ARToolkit is plain GPL - but of course, I am not a lawyer, so we can argue about this all day and as you point out we will never agree on this. I respect the point that you have raised however because it is possibly a bit unclear exactly what is meant. And could possibly mean that we are all doing something wrong by downloading ARToolkit, even though Mark and Hiro want it to be distributed to us. Perhaps Mark or Hiro or whoever maintains ARToolkit could update the licensing page and source code downloads to be more specific as to what the license is? Could someone who maintains the latest source code respond to this point and possibly address it? > Since this list seems to think that the license is something it isn't Well, just me so far from what I can tell :) regards, Wayne |
From: | Kirk Martinez <km@e ..............> | Received: | Sep 29, 2004 |
To | Wayne Piekarski <wayne@c ..............>, Michael Sparks <zathras@t ............> | ||
Subject: | Re: GPL AR Toolkit? | ||
I agree that the distribution could do with some work ;-) there are too many versions downloadable from different sites (try a google search!) So I'd suggest HITlabNZ makes a sourceforge or other cvs/group usable distribution mechanism. I predict an exponential growth in features and possibly stability if we all behave properly ;-) Cheers, Kirk Martinez At 06:39 28/09/2004, Wayne Piekarski wrote: >>I agree that plain GPL protects the uses that Mark, Hiro, et al wish to >>prevent, FWIW. However what they want, and what is stated are two >>different things... What you claim and what is stated and the case are two >>different things. I'm not interested in an argument about whether it is or >>isn't. It isn't. You won't convince me of otherwise because I've seen this >>several times in the past, and just wondered if anyone had encountered >>something even half as good that was plain/pure GPL and stated explicitly >>as such. I also doubt I'll convince you of otherwise either, ce le vie :-) >>I'm aware of the discussion, however given the website does currently >>state GPL+restrictions, which as copyright owners they *are* allowed to do >>(simple fact of life - they own the initial copyright - simple referencing >>the GPL + a restriction is valid), and the download essentially says the >>same thing I'm in either one of two situations: >> * The license as stated is invalid (in which case I can't redistribute >> any changes). >> * The license is valid but self contradictory preventing it being used. >> >>I'm aware of the reasons the AR Toolkit team want to limit closed usage >>(which makes a lot of sense), but with the current ambiguity I think I'm >>best off using something else and getting out of people's hair. It's no >>big deal though. > >I think probably the best thing to do is get this clarified by the authors >of the program. I am not a lawyer of course, but when I read the wording >on the home page it says its "ARToolKit is distributed free for >non-commercial use under the GPL license". Thats the only reference I've >really found to how it should be distributed. So I read it as being, "this >code is GPL, which will stop you from using it commercially because of the >GPL itself". This is why I believe that ARToolkit is plain GPL - but of >course, I am not a lawyer, so we can argue about this all day and as you >point out we will never agree on this. I respect the point that you have >raised however because it is possibly a bit unclear exactly what is meant. >And could possibly mean that we are all doing something wrong by >downloading ARToolkit, even though Mark and Hiro want it to be distributed >to us. > >Perhaps Mark or Hiro or whoever maintains ARToolkit could update the >licensing page and source code downloads to be more specific as to what >the license is? > >Could someone who maintains the latest source code respond to this point >and possibly address it? > > >>Since this list seems to think that the license is something it isn't > >Well, just me so far from what I can tell :) > > >regards, >Wayne - http://www.ecs.soton.ac.uk/~km |