FOSDEM: Richard Fontana on copyleft-next [LWN.net]

Ready to give LWN a try?
With a subscription to LWN, you can stay current with what is happening in the Linux and free-software community and take advantage of subscriber-only site features. We are pleased to offer you a free trial subscription, no credit card required, so that you can see for yourself. Please, join us!

February 13, 2013

This article was contributed by Martin Michlmayr

In July 2012, Richard Fontana started the GPL.next project to experiment with modifications to version 3 of the GNU General Public License (GPLv3). The name was quickly changed to the more neutral "copyleft-next" and the license has evolved into a "radically different text" compared to the GPLv3 since project inception. Fontana gave a talk in the FOSDEM legal devroom on February 3 that presented the current status of the project and his reasons for exploring new ideas about copyleft licensing.

Fontana explained that he initially described the project as a fork of the GPLv3 but admitted that it "sounded more negative than I intended". He actually co-authored the GPLv3, LGPLv3, and AGPLv3 licenses together with Richard Stallman and Eben Moglen during his time at the Software Freedom Law Center. Fontana, who is now Red Hat's open-source licensing counsel, stressed that copyleft-next is his personal project and not related to his work for SFLC, FSF, or Red Hat, although "these experiences had a personal influence".

The complexity of the GPLv3

Allison Randal's 2007 essay "GPLv3, Clarity and Simplicity" is a powerful critique of the GPLv3 and was deeply influential on his thinking, Fontana said. The essay argued that everyone "should be enabled to comprehend the terms of the license". Based on the (then) near-finished draft of the GPLv3, Randal observed that it's unlikely that clarity and simplicity had been a priority during the drafting process.

Fontana feels that the complexity of the GPL had a side effect of creating an "atmosphere of unnecessary inscrutability and hyper-legalism" surrounding the GPL. Additionally, he perceives that legal interpretation of the license is lacking. Richard Stallman has withdrawn from active license interpretation and Brett Smith, for a long time FSF's "greatest legal authority" according to Fontana, left his position as FSF's License Compliance Engineer in May 2012. He wonders whether the complexity of the GPL, together with FSF's withdrawal from an active interpretive role, has contributed to a shift to non-copyleft licenses. He also believes that developer preference for licensing minimalism is rising.

Another reason for the creation of copyleft-next is Fontana's desire to experiment with new ideas and forms of licensing. He pointed out that every license (proprietary or free) is imperfect and could benefit from improvements. He feels strongly that license reform should not be monopolized. Due to concerns about license proliferation, the OSI has discouraged the creation of new licenses, effectively creating a monopoly for the stewards of existing OSI-approved licenses. Fontana downplayed concerns of license proliferation, partly because GPL-compatible licenses should also be compatible with copyleft-next and because copyleft-next offers one-way compatibility with the GPL. Finally, he views copyleft-next as a "gradual, painless successor to GPLv2/GPLv3".

Fontana also expressed his disappointment in the way open source licenses have historically been developed. While the drafting process for GPLv3 was very advanced and transparent compared to other efforts, it seems insufficiently transparent to him by present-day standards. He pointed to the Project Harmony contributor agreements as another example of a non-transparent process since it employed the Chatham House Rule during parts of the drafting process.

Contribution norms

Unsurprisingly, copyleft-next's development process is very different and follows the "contemporary methodology of community projects". The license is hosted on Gitorious, and there is a public mailing list and IRC channel—a bug tracker will be added in the near future. Fontana acts as the sabd(nnfl)—the self-appointed benevolent dictator (not necessarily for life).

The project has participation guidelines (informally known as the Harvey Birdman Rule, after a US cartoon series featuring lawyers). The norms reflect Fontana's intention to involve developers and other community members in the development process. They encourage transparency in license drafting and aim to "prevent the undue influence of interest groups far removed from individual software developers" (in other words, lawyers).

The guidelines disallow closed mailing lists as well as substantive private conversations about the development of the project. The latter can be remedied by posting a summary to the public mailing list. Fontana is true to his word and posted summaries of discussions he had at FOSDEM. Finally, the Harvey Birdman Rule forbids contributions in the form of word-processing documents and dictates that mailing list replies using top-posting shall be ignored.

The copyleft-next license

The copyleft-next license is a strong copyleft license. The word "strong" refers to the scope of the license. The Mozilla Public License (MPL), for example, is a weak copyleft license in this sense since its copyleft only applies to individual files. While modifications to a file are covered by MPL's copyleft provisions, code under the MPL may be distributed as part of a larger proprietary piece of software. The GPL and copyleft-next, on the other hand, have a much broader scope and make it difficult to make proprietary enhancements of free software.

Copyleft-next was initially developed by taking the GPLv3 text and removing parts from it. For each provision, Fontana asked whether the incremental complexity associated with the provision is necessary and worthwhile. For many provisions, he concluded they weren't—this includes provisions in the GPLv3 that no other open source license has needed, obscure clauses, and text that should be moved to a FAQ. The GPL has a lot of historical baggage, and Fontana believes that the reduction in complexity of copyleft-next has led to a license that developers and lawyers alike can read and understand. Those readers interested in verifying this claim can find the current draft on Gitorious.

In order to show the drastic reduction in complexity, Fontana compared the word and line counts of several popular open source licenses. The word counts were as follows:

License Words

copyleft-next 0.1.0 1423

Apache License 2.0 1581

GPLv1 2063

MPL 2.0 2435

GPLv2 2968

GPLv3 5644

License	Words
copyleft-next 0.1.0	1423
Apache License 2.0	1581
GPLv1	2063
MPL 2.0	2435
GPLv2	2968
GPLv3	5644

For comparison, the MIT license consists of 162 words and the BSD 3-clause license has 212 words.

Copyleft-next has a number of interesting features. It offers outbound compatibility with the GPLv2 (or higher) and AGPLv3 (or higher), meaning that code covered by copyleft-next can be distributed under these licenses. This allows for experimentation in copyleft-next, Fontana explained. The license also simplifies compliance: when the source code is not shipped with a physical product, distributors do not have to give a written offer to supply the source code on CD or a similar medium. They can simply point to a URL where the source code can be found for two years. Like the GPLv3, copyleft-next allows license violations to be remedied within a certain time period (although compared to GPLv3 the provision has been simplified). In contrast to GPLv3, the current draft of copyleft-next doesn't contain an anti-Tivoization clause.

The copyleft-next license also takes a stance against certain practices detested by many community members. The license includes a proprietary-relicensing "poison pill": if the copyright holders offer proprietary relicensing, the copyleft requirements evaporate—the project effectively becomes a permissively licensed one, meaning that no single entity has a monopoly on offering proprietary versions. This provision was inspired by the Qt/KDE treaty, which says that the KDE Free Qt Foundation can release Qt under a BSD-style license if Qt is no longer offered under the LGPL 2.1. Furthermore, copyleft-next has an anti-badgeware provision: it explicitly excludes logos from the requirement to preserve author attributions.

While copyleft-next started as an exercise to simplify the GPLv3, it has incorporated ideas and concepts from other licenses in the meantime. For example, several provisions, such as the one explicitly excluding trademark grants, were inspired by or directly borrowed from MPL 2.0.

Fontana made the first release of copyleft-next, 0.1.0, just before FOSDEM and released version 0.1.1 in the interim. He mentioned during the talk that he is thinking of creating an Affero flavor of copyleft-next as well. He would like to see more participation from community members. The mailing list provides a good way to get started and the commit logs explain the rationale of changes in great detail.

Index entries for this article
GuestArticles	Michlmayr, Martin
Conference	FOSDEM/2013

to post comments

FOSDEM: Richard Fontana on copyleft-next

Posted Feb 14, 2013 9:47 UTC (Thu) by nmav (guest, #34036) [Link] (36 responses)

> if the copyright holders offer proprietary relicensing, the copyleft requirements evaporate

I don't understand why is that. Why shouldn't a someone developing a proprietary product, also be able to release it under that copyleft license?

FOSDEM: Richard Fontana on copyleft-next

Posted Feb 14, 2013 10:06 UTC (Thu) by khim (subscriber, #9252) [Link] (5 responses)

Because then your copyleft-licensed version starts doing the direct opposite from what it was intended to do. Copyleft is supposed to encourage development of FOSS software but combination of relatively-obscure-copyleft and proprietary license instead encourages development of proprietary and proprietary-friendly code!

FOSDEM: Richard Fontana on copyleft-next

Posted Feb 14, 2013 16:10 UTC (Thu) by dashesy (guest, #74652) [Link] (4 responses)

Or it could mean this: If you are willing to open your source then great you can use the copyleft. If you want to keep your source closed (for whatever reason including management discomfort, or because this open source software is just a small piece in a large closed code base) we offer other licenses too. It also does not necessarily mean charging for the proprietary offerings.

FOSDEM: Richard Fontana on copyleft-next

Posted Feb 14, 2013 16:24 UTC (Thu) by rfontana (subscriber, #52677) [Link] (3 responses)

I haven't seen too many cases, or possibly any case, where there wasn't an obvious pecuniary motive (to the point where I think it might be okay to limit the poison pill to commercial contexts).

If a developer or company has no pecuniary motive for proprietary relicensing, why would it bother to use copyleft for the project to begin with? Better to allow noncopyleft use by all; what is socially gained by permitting occasional discrimination (particularly where, as your comment might imply, the discrimination is always in favor of a commercial entity)?

If there were no discrimination, then any arbitrary user could get a noncopyleft license free of charge on simple request. At that point, the original decision to use a copyleft license has no rational justification.

FOSDEM: Richard Fontana on copyleft-next

Posted Feb 14, 2013 17:27 UTC (Thu) by dashesy (guest, #74652) [Link] (1 responses)

Sorry for my ignorance, IANAL but it seems in this business everyone needs to self-educate on the subject.

So in a sense it is better to have one noncopyleft(free-to-all) license, than a copyleft+noncopyleft(free-to-all) dual license. I guess some of the potential use-cases for dual licensing can be achieved by LGPL, where the code improvements are integrated for all, but commercial entities need to also avoid static linkage (discrimination against the commercial usage). Also noncopyleft(free-to-all) means every user needs to get the software license separately, yes it is a simple request (but no request is too simple IMO), and it cannot be as easily bundled with other copyleft software. I guess part of the copyleft beauty is the ease to give software to others, and not just using it.

Another use-case (which is actually what I had in mind originally) is to have the noncopyleft license available to commercial entities but only for the final official binaries, at the same time have the source available under copyleft, with the hope that they return back the improvements to be incorporated in the future binaries for them.

FOSDEM: Richard Fontana on copyleft-next

Posted Feb 14, 2013 18:23 UTC (Thu) by rfontana (subscriber, #52677) [Link]

> So in a sense it is better to have one noncopyleft(free-to-all) license, than a copyleft+noncopyleft(free-to-all) dual license.

I don't see any point to a copyleft/noncopyleft dual license except in some odd case of perceived or actual license incompatibility.

> Also noncopyleft(free-to-all) means every user needs to get the software license separately,

I don't understand this point. A (typical) feature of both copyleft and noncopyleft FLOSS licenses is that you get the license without any particular ceremonial overhead (this is also true of many proprietary licenses).

> Another use-case (which is actually what I had in mind originally) is to have the noncopyleft license available to commercial entities but only for the final official binaries, at the same time have the source available under copyleft, with the hope that they return back the improvements to be incorporated in the future binaries for them.

Ah, if I understand you correctly that is a whole nother issue, and one I've been thinking about recently. Some weak copyleft licenses explicitly allow non-FLOSS licensing of binaries by licensees, but the source code can only be distributed under the copyleft license.

FOSDEM: Richard Fontana on copyleft-next

Posted Feb 14, 2013 22:40 UTC (Thu) by nmav (guest, #34036) [Link]

> If a developer or company has no pecuniary motive for proprietary relicensing, why would it bother to use copyleft for the project to begin with? Better to allow noncopyleft use by all; what is socially gained by permitting occasional discrimination (particularly where, as your comment might imply, the discrimination is always in favor of a commercial entity)?

It is true that several companies use copyleft (mostly GPLv2) in order to attract more attention and developers on their software and sell proprietary licenses with other terms to people who pay. I don't see any harm on that, but even if you think there is harm by someone being treated differently, I don't see why punish this company at all. They released their code as free software anyway (assuming they play fair).

What the license looks to me, is that it says, that if you have a business model that allows you to develop and release free software then we take it from you. It doesn't sound about code being free to see and modify.

You may have a point on companies abusing the copyleft terms. In that case you have to be explicit and justify those clauses, because their reasons are not apparent.

FOSDEM: Richard Fontana on copyleft-next

Posted Feb 14, 2013 11:17 UTC (Thu) by aggelos (subscriber, #41752) [Link] (29 responses)

AFAIUI, it's a level-playing field clause. The idea is to prevent a member of the community from enjoying a special status, as only they can derive proprietary versions (there is a grace period to allow smooth copylefting of code that starts as a proprietary product). Most importantly, a single copyright holder also gets to 'cure' license violations for some kind of compensation and copyleft-next wants to discourage the business model of trying to trick people into a violation or creating FUD about copyleft requirements in order to get others to "buy" "legal certainty" from you.

FOSDEM: Richard Fontana on copyleft-next

Posted Feb 14, 2013 18:49 UTC (Thu) by josh (subscriber, #17465) [Link] (28 responses)

Open Source projects *already* have a means of preventing a member of the community from enjoying that "special status": unless the project maintainer requires a copyright assignment agreement or contributor agreement handing over the right to relicense, any actively developed project will have contributions from many copyright holders, all of whom would have to agree on any relicensing (proprietary or otherwise).

Personally, I'd have found copyleft-next rather interesting and compelling if not for issues like this. An effort to produce a simpler copyleft license does not seem like an appropriate time to introduce novel opinionated additions.

FOSDEM: Richard Fontana on copyleft-next

Posted Feb 14, 2013 20:52 UTC (Thu) by rfontana (subscriber, #52677) [Link] (7 responses)

Point taken. It is certainly true that this provision is distinct from the main goal of copyleft-next.

But I do like the underlying idea behind this provision. It corrects what I consider to be an error in all previous copyleft licenses, an error that has had very troubling ethical consequences. You're right that another way to correct it is to encourage non-copyright-assignment/non-CLA copyleft project legal governance. However, I don't want this license to be tainted by that practice the way earlier/existing copyleft licenses have been. I feel quite strongly about this; you're not wrong to describe this provision as 'opinionated' in nature.

FOSDEM: Richard Fontana on copyleft-next

Posted Feb 14, 2013 21:45 UTC (Thu) by josh (subscriber, #17465) [Link] (1 responses)

I don't object to the principle you've suggested here, though I also don't see anything wrong with projects that choose to allow that practice. (I see nothing wrong with the "share your code or pay us" licensing model, and I certainly wouldn't consider it an error in all previous copyleft licenses.) However, I'd much rather have seen that as an optional addition to a license, rather than a required part of an otherwise much simpler license; as long as this provision remains in copyleft-next, I'll stick with the GPL. Given that you started this project partly in response to the many novel and controversial licensing provisions in GPLv3, such as the anti-DRM clause, it seems suboptimal to introduce your own novel and controversial licensing provisions.

The main reason I don't consider this an error in previous copyleft licenses, and the reason I brought up assignments/CLAs: the copyleft licenses themselves have no inherent asymmetry. If you write a project yourself, and hold all the copyrights, I see absolutely nothing wrong with choosing a dual-licensing model, including one that provides paid non-copyleft licenses. If you accept contributions, then unless those contributors sign an assignment or CLA, you can no longer use that dual-licensing model without the agreement of those contributors. If you want to maintain that dual-licensing model while taking third-party contributions, you have to introduce an assignment or CLA, and *that* introduces the asymmetry; the copyleft license doesn't.

I see no asymmetry here, and thus I see no bug in existing copyleft licenses here.

FOSDEM: Richard Fontana on copyleft-next

Posted Feb 18, 2013 14:23 UTC (Mon) by ian00 (guest, #55476) [Link]

"*that* introduces the asymmetry." And this excludes *that* asymmetry...

FOSDEM: Richard Fontana on copyleft-next

Posted Feb 18, 2013 14:18 UTC (Mon) by ian00 (guest, #55476) [Link]

I completely agree with you on adding this provision and why.

Ethical consequences ...

Posted Feb 21, 2013 16:21 UTC (Thu) by Wol (subscriber, #4433) [Link] (3 responses)

But if I understand this aright, your "fix" has an equally troubling ethical consequence.

Let's say I've spent loads of time and money developing my product. I want to PROTECT my customers, like Qt/KDE did. It sounds like I can't use your licence to do it. Oooppsss.

Personally, I have no problem with people trying to monetise and sell their software. But as a customer I don't want to be trapped with outdated software and no source when technology moves on, and I think developers have a duty to their customers. Pity I can't use your licence to achieve that :-(

Cheers,
Wol

Ethical consequences ...

Posted Feb 21, 2013 16:32 UTC (Thu) by rfontana (subscriber, #52677) [Link] (2 responses)

I don't understand. To make this simpler let's forget copyleft-next and suppose the license is the GPL. What is there in pure GPL that prevents you from "PROTECTING" your customers? I don't even know what you mean by "PROTECTING". It seems to be "making sure customers have up-to-date software and source code". What in the GPL prevents you from doing that?

Ethical consequences ...

Posted Feb 21, 2013 17:01 UTC (Thu) by Wol (subscriber, #4433) [Link] (1 responses)

So you ignore the point I'm trying to make!

The GPL promptly makes all my hard work in writing the software valueless (certainly in terms of being able to extract an income stream). You're imposing your "software wants to be free" ethic on me, and not respecting my "the labourer is worthy of his hire" ethic.

In my scenario *I* wrote the software, I want to *sell* the software, and I want to make sure that if I fall under a bus my customers won't be left up s**t creek.

Okay, that obviously means I don't buy in to *hard* copyleft, but your attitude is in the "GPL or die" camp. And you're using copyleft to enforce stuff that is outside the realm of copyright, namely you are (in practice) forrbidding copyright licencing agreements. That's exactly the sort of thing that alienates a lot of copyleft-friendly people who your copyleft-next is allegedly aimed at.

Just as the anti-tivoisation clauses in GPL3 alienated a lot people who saw them as stepping outside the realm of copyright and into the realm of enforcing idealism, so this clause to me smacks of exactly the same.

At the end of the day, if I am sole copyright holder (or have a licence from contributors that says I can) and distribute both proprietary and GPL versions, then those other contributors know up front what they're letting themselves in for. And it's a *contract* matter, not a *copyright* matter.

Equally to the point, I (should) know exactly what I as the copyright holder am letting myself in for. Actually, I have seen at first hand what happens when a company tries to convert their product into a dual-licence proprietary/GPL state. Firstly, they would never have touched your licence in a month of sundays. And secondly, it went rather tits-up because they didn't actually understand what they were getting themselves into. But even given that mess, they have a community. If you want to investigate, google "OpenQM". And people, if you find stuff you don't like, PLEASE don't dive in - they are nice people, really!

Cheers,
Wol

Ethical consequences ...

Posted Feb 21, 2013 17:36 UTC (Thu) by rfontana (subscriber, #52677) [Link]

> The GPL promptly makes all my hard work in writing the software
> valueless (certainly in terms of being able to extract an income
> stream). You're imposing your "software wants to be free" ethic on
> me, and not respecting my "the labourer is worthy of his hire"
> ethic.

I believe you may be confusing "free as in freedom" with "free as in
beer". After 25 or so years it is settled that businesses can extract
income streams from GPL-licensed software.

> Okay, that obviously means I don't buy in to *hard* copyleft, but
> your attitude is in the "GPL or die" camp.

No.

> And you're using copyleft
> to enforce stuff that is outside the realm of copyright, namely you
> are (in practice) forrbidding copyright licencing agreements.

No, I'm essentially saying 'if you use a strong copyleft license for
some software you write, then if you also want to use a proprietary
license for derivative works of that software, give everyone else the
right to similarly use a proprietary license for their derivative
works". It's an equality principle. Nothing is forbidden; quite the
opposite. You are saying that it is okay for a company to have the
sole right to do proprietary licensing of a GPL codebase -- that is
obviously forbidding GPL licensees from entering into proprietary
copyright licensing agreements covering *their* derivative works.

> At the end of the day, if I am sole copyright holder (or have a
> licence from contributors that says I can) and distribute both
> proprietary and GPL versions, then those other contributors know up
> front what they're letting themselves in for.

It's not the contributors I'm concerned about here, primarily, though
one problem with the model you describe is it tends to discourage
community contributions.

> Actually, I have seen at first hand what happens when a company
> tries to convert their product into a dual-licence proprietary/GPL
> state. Firstly, they would never have touched your licence in a
> month of sundays.

That's exactly the effect I wish to achieve.

FOSDEM: Richard Fontana on copyleft-next

Posted Feb 14, 2013 22:04 UTC (Thu) by aggelos (subscriber, #41752) [Link] (19 responses)

Oh, absolutely agreed that having multiple copyright holders is the healthy way to deal with the "proprietary relicensing" problem. Not just multiple in fact, also different kinds of copyright holders (e.g. not only corporations that don't care about the code itself). Given that the initial author(s) get to choose the license, if they wanted to keep their options open, they wouldn't choose copyleft-next 0.1 anyway. So the clause might be of limited use in practice (except in special conditions; no real-world examples come to mind).

FOSDEM: Richard Fontana on copyleft-next

Posted Feb 15, 2013 0:15 UTC (Fri) by josh (subscriber, #17465) [Link] (10 responses)

Exactly. And given all of that, I think that clause fails the test described in the article: "For each provision, Fontana asked whether the incremental complexity associated with the provision is necessary and worthwhile."

Given the inherent controversy involved in releasing a new copyleft license, especially one not currently endorsed by the FSF, I think copyleft-next should strive for "simple" and "uncontroversial" as primary goals.

FOSDEM: Richard Fontana on copyleft-next

Posted Feb 15, 2013 1:58 UTC (Fri) by rfontana (subscriber, #52677) [Link] (9 responses)

If that provision does get cut out, the reason will be failure to meet that standard. However, I don't see what's so controversial about the intention of the provision. Businesses experimenting with copyleft-based dual-licensing and open core models will always have other licenses to choose from.

FOSDEM: Richard Fontana on copyleft-next

Posted Feb 15, 2013 2:47 UTC (Fri) by dlang (guest, #313) [Link] (8 responses)

the problem is that you are falling into the same trap the FSF did with GPLv3.

you either thing that

A) everyone agrees with you on this topic (hint, not everyone does, a lot of people are Ok with the idea of a project selling proprietary licences if it funds the opensource code)

B) this is such a critical item that it's worth having many people not use the license, or convert from it to GPL so that they can then sell the proprietary licenses.

and if you can't make that conversion to GPL to bypass the restriction, then you are imposing an additional restriction to what the GPL includes and are not GPL compatible)

Many times projects start off under one license, and it's only years later that they decide to raise money by selling GPL exceptions. In many cases, the uses that the code then gets put to would probably be legal even under the GPL, except for creating what would effectively be a mirror of an obsolete version of the source (unmodified code compiled and shipped in an appliance for example)

So this is a provision that is controversial, and can be pretty trivially bypassed, why is it so valuable again?

FOSDEM: Richard Fontana on copyleft-next

Posted Feb 15, 2013 3:36 UTC (Fri) by josh (subscriber, #17465) [Link]

Exactly. And potential users of copyleft-next also benefit when *other* people use that license; you want to encourage widespread adoption, and license clauses not directly addressing copyleft seem counterproductive for that.

I'd really like to see the simplest possible license that provides approximately the same set of essential protections as the GPLv{2,3}; thus, I'd really like to see this license act as a strict subset of those licenses.

FOSDEM: Richard Fontana on copyleft-next

Posted Feb 15, 2013 3:57 UTC (Fri) by rfontana (subscriber, #52677) [Link] (6 responses)

> you either thing that A) everyone agrees with you on this topic
> (hint, not everyone does, a lot of people are Ok with the idea of a
> project selling proprietary licences if it funds the opensource
> code)

Of course I realize not everyone agrees with me. People disagree about
licensing policy; some prefer noncopyleft, some prefer strong
copyleft, some prefer weak copyleft. No issue with that; such
diversity is great.

But this is not a matter of selling proprietary licenses to fund open
source code, which strikes me either as a propagandistic or else a
quaint way of seeing it. Rather, it's the problem of abusing copyleft
specifically to sell proprietary licenses, with a number of specific
attendant problems, such as promoting FUD around copyleft.

> B) this is such a critical item that it's worth having many people
> not use the license, or convert from it to GPL so that they can then
> sell the proprietary licenses.

There's some evidence that these kinds of problematic business models
are in decline, so that supports the view that this may not be so
critical. But if having a provision that effectively nullifies
copyleft/proprietary dual-licensing tactics drives some people away
because they want to engage in such tactics, I call that
success. That's precisely the purpose of the provision: to drive those
people or businesses away, so that, however few or many developers use
copyleft-next, no one does so as a proprietary upsell gimmick. The
only problem is that the provision in its current form may not be as
clear as it could be in stating its objective.

> and if you can't make that conversion to GPL to bypass the
> restriction, then you are imposing an additional restriction to what
> the GPL includes and are not GPL compatible)

The license is explicitly outbound GPL-compatible. No GPL-incompatible
restrictions will survive licensing derived works under the GPL. But
no one should ever get to the point of having to convert to GPL to
bypass the provision in question. A business (or its legal advisor)
contemplating some copyleft-based open core scheme or the like will
take one look at copyleft-next and decide to use some other copyleft
license for such a purpose.

> So this is a provision that is controversial, and can be pretty
> trivially bypassed, why is it so valuable again?

I'm afraid I don't see the controversy. I see a (probably declining)
number of businesses using GPL/AGPL or other copyleft-based
dual-licensing/open core/proprietary upsell business models. I happen
to consider this cynical use of copyleft licensing to be a tremendous
ethical failure. I have mainly encountered negative views of such
business models from those who have sympathy for the core policy goals
of strong copyleft.

As for bypassing it, the point is to discourage any use of the license
for proprietary upsell purposes to begin with.

FOSDEM: Richard Fontana on copyleft-next

Posted Feb 15, 2013 5:05 UTC (Fri) by josh (subscriber, #17465) [Link] (2 responses)

I think I might understand your concern now. If you mean dual-licensing business models that follow the poorly named "open core" model and sell an extended proprietary version, then I agree with you; that seems just as problematic as selling proprietary software in the first place. On the other hand, I don't see anything wrong with selling permissively licensed versions of the same code as a revenue model.

I'd still rather see this as an optional addition to the license, though, rather than a required provision.

FOSDEM: Richard Fontana on copyleft-next

Posted Feb 15, 2013 21:19 UTC (Fri) by ballombe (subscriber, #9523) [Link] (1 responses)

To sell permissively licensed versions of the same code require you to be the sole copyright holder. So either you require copyright assignment or you refuse all external contributions, both alternatives being antisocial.

FOSDEM: Richard Fontana on copyleft-next

Posted Feb 17, 2013 12:31 UTC (Sun) by jrn (subscriber, #64214) [Link]

Or that you get a license agreement from contributors explicitly allowing you to sell permissively licensed versions of the same code.

The Chromium CLA does that, for example:
http://code.google.com/legal/individual-cla-v1.0.html

FOSDEM: Richard Fontana on copyleft-next

Posted Feb 15, 2013 5:10 UTC (Fri) by dlang (guest, #313) [Link]

> But this is not a matter of selling proprietary licenses to fund open source code, which strikes me either as a propagandistic or else a quaint way of seeing it. Rather, it's the problem of abusing copyleft specifically to sell proprietary licenses, with a number of specific attendant problems, such as promoting FUD around copyleft.

The problem is how to you block the bad behaviour (which as you note is not especially popular to start with) without blocking the acceptable versions.

your 'Anti-MySQL' clause isn't any better than the FSF's 'Anti-Tivo' clause. both are reactions to fairly rare abuse cases, and both are divisive in that there are large numbers of people in the community who don't see the behaviour as being so wrong that it's worth complicating a licence (and loosing all the non-abusive variations of the behaviour) to block the abusive behaviour.

> A business (or its legal advisor) contemplating some copyleft-based open core scheme or the like will take one look at copyleft-next and decide to use some other copyleft license for such a purpose.

that assumes that that approach is in their minds at the start of the project. It may be something that's first discussed years into the project (when it gets big enough and functional enough for people to be interested in paying for it)

It's very common for projects to start and grow, and then hit a crisis where the work to maintain them is large enough to need the core people paid to do the work, but not have a funding source in place to do this. It's times like this that the 'run it through the GPL to strip this clause' would happen.

FOSDEM: Richard Fontana on copyleft-next

Posted Feb 21, 2013 10:41 UTC (Thu) by Otus (subscriber, #67685) [Link] (1 responses)

> The license is explicitly outbound GPL-compatible. No GPL-incompatible
> restrictions will survive licensing derived works under the GPL. But
> no one should ever get to the point of having to convert to GPL to
> bypass the provision in question. A business (or its legal advisor)
> contemplating some copyleft-based open core scheme or the like will
> take one look at copyleft-next and decide to use some other copyleft
> license for such a purpose.

Suppose project A decides they are OK with the provision preventing an open core model and decide to use copyleft-next (CN) for their license. What's to stop them from rethinking later, changing the project license to GPL and developing a proprietary version with the open core model?

Yes, the lack of copyright for parts of the code would do that unless they require copyright assignment or a more permissive contributor license, but how is that any different from the situation of CN without the provision?

IOW, I don't see how the provision adds any restrictions on use of the code or project format. It just means if some project does eventually decide to go open core, their further contributions will no longer be compatible with other CN projects, only GPL ones. That seems to hurt only those who use CN "in good faith" as their license.

FOSDEM: Richard Fontana on copyleft-next

Posted Feb 21, 2013 14:39 UTC (Thu) by rfontana (subscriber, #52677) [Link]

> IOW, I don't see how the provision adds any restrictions on use of
> the code or project format.

That's right; it doesn't. It is meant to be a strong disincentive to
use copyleft-next for copyleft/proprietary dual-licensing (by
nullifying its whole basis, which is monopolization of the
proprietization right), and to confine such conduct to other licenses.

> It just means if some project does eventually decide to go open
> core, their further contributions will no longer be compatible with
> other CN projects, only GPL ones. That seems to hurt only those who
> use CN "in good faith" as their license.

That's an interesting argument. The effect of this provision would
(let us assume) be to drive such a project (or to be more precise, the
business or entrepreneur behind it) to other licenses, these days most
likely the GPL or AGPL. So you're arguing that this is just a loss for
a copyleft-next code commons because it will be deprived of some
ability to use this open core code that might otherwise be under
copyleft-next.

First, such open core code will remain available for use by
copyleft-next projects provided it is a 'Separate Work' (outside the
scope of copyleft).

Second, assuming the open core project uses GPLv2/GPLv3/AGPLv3, such
open core code will be usable *inside the scope* of copyleft-next
works, with the theoretical effect being that the larger work will be
under GPLv2 or GPLv3 or some mixture of GPLv3/AGPLv3. This need not
have the effect of taking the *copyleft-next* code out of the
copyleft-next commons, so it doesn't seem like a huge loss. After all,
this nature of code combination occurs precisely because copyleft-next
is designed to be outbound-compatible with the GPL.

So I think your argument reduces to the argument that this provision
would have the effect of keeping the universe of copyleft-next
projects smaller. To me, this is *possible*, but it seems unlikely to
be meaningful unless copyleft-next becomes significantly popular
(which would be great but admittedly is not the case at the moment
:). And if it did become popular, I would see it as all the more
important to have this provision in, to prevent the 'gaming' of
copyleft-next that we've seen with other copyleft licenses.

In short, I'd rather have the copyleft-next universe be smaller and
limited to 'good-faith' projects, with the problematic conduct
confined to existing copyleft licenses like GPL/AGPL, than have the
copyleft-next universe be larger.

FOSDEM: Richard Fontana on copyleft-next

Posted Feb 15, 2013 2:06 UTC (Fri) by rfontana (subscriber, #52677) [Link] (7 responses)

The purpose of the provision is actually to deter licensors from attempting to use copyleft-next as a mechanism for promoting proprietary license sales in the manner we have seen in the past with other copyleft licenses. I don't expect it to have some "gotcha" effect on some entrepreneur who failed to read the license. The point is to prevent such use of the license in the first place.

In this sense referring to it as a 'poison pill', while sort of appealing, may be misleading. It's effectively a "keep away from this license if that's what you want to use copyleft for" sign.

FOSDEM: Richard Fontana on copyleft-next

Posted Feb 23, 2013 19:32 UTC (Sat) by dlang (guest, #313) [Link] (6 responses)

The thing is that this anti-mssql agenda changes your license from a general purpose copyleft license to yet another niche license.

FOSDEM: Richard Fontana on copyleft-next

Posted Feb 23, 2013 21:04 UTC (Sat) by rahulsundaram (subscriber, #21946) [Link] (5 responses)

It is not a anti-mysql agenda. It is a clause that protects contributors from one sided copyright license agreements. You should be careful not to mislabel things since it is not specific to MySQL in any way.

FOSDEM: Richard Fontana on copyleft-next

Posted Feb 23, 2013 23:58 UTC (Sat) by dlang (guest, #313) [Link] (4 responses)

MySQL is the poster child for the behaviour you are against, just like Tivo was the poster child for the behaviour that the FSF was against when writing that clause in the GPLv3.

Just like the 'anti-tivo' clause in GPLv3 did a lot to cause the split in developers, this 'anti-mysql' clause in this license will split developers as well

If your purpose is to divide the community further, then go ahead.

but if you want to unify the community drop this clause

FOSDEM: Richard Fontana on copyleft-next

Posted Feb 24, 2013 2:19 UTC (Sun) by rahulsundaram (subscriber, #21946) [Link]

Using MySQL as poster child isn't accurate nor relevant. Ghotscript started the model and while MySQL adopted it, most of the current revenue for MySQL are from support subscriptions and has been for a very long time.

Besides your argument is too simplistic. BSD communities don't prefer any copyleft licenses at all so even GPLv2 was very divisive for them. So we have to ask ourselves, what is the target of the license and which communities are we helping?

FOSDEM: Richard Fontana on copyleft-next

Posted Feb 24, 2013 3:29 UTC (Sun) by rfontana (subscriber, #52677) [Link] (2 responses)

> MySQL is the poster child for the behaviour you are against, just
> like Tivo was the poster child for the behaviour that the FSF was
> against when writing that clause in the GPLv3.

There is a big difference between the copyleft-next provision and the
anti-lockdown provision of GPLv3. The copyleft-next provision works by
granting additional freedom to downstream users (freedom for everyone
to escape copyleft, giving everyone an equal right to create
proprietary derivative works rather than monopolizing that right in
the upstream licensor). The GPLv3 provision imposes an additional
requirement on downstream distributors (beyond pre-GPLv3 consensus
understanding of GPL copyleft requirements). So the comparison seems
odd to me.

> If your purpose is to divide the community further, then go ahead.
> but if you want to unify the community drop this clause

This makes it sound as though there is a sizable "community" of people
who are thrilled or enthusiastic about copyleft/proprietary
dual-licensing/open core business models. I see no evidence that such
a community exists. The effect of this provision will simply be that
those who wish to extend the life of this peculiar legacy use of
copyleft will continue to use the licenses that are being used for
this purpose today.

FOSDEM: Richard Fontana on copyleft-next

Posted Feb 25, 2013 21:40 UTC (Mon) by dlang (guest, #313) [Link] (1 responses)

the anti-tivo provision was justified by the FSF as "granting additional freedom to downstream users"

they used almost the exact same language you are using to defend that provision

FOSDEM: Richard Fontana on copyleft-next

Posted Feb 26, 2013 2:46 UTC (Tue) by rfontana (subscriber, #52677) [Link]

Okay, I do see some similarity. There's an equality principle
underlying both provisions.

GPLv3 section 6 states that the anti-lockdown provision does not apply
"if neither you nor any third party retains the ability to install
modified object code on the User Product".

In other words, the policy problem the anti-lockdown provision was
aimed at (by the time of its reformulation in public Draft 3 of GPLv3)
was the asymmetry of upstream (or some third party) being able to
install modified versions on the user's device, with the user being
technically thwarted from exercising a similar freedom to install
modified versions on the user's device.

In the copyleft-next provision, the original licensor is effectively
saying "If I get to create proprietary derivative works, then you
ought to have the same right to create proprietary derivative works".

There's also an equality principle underlying some of the GPLv3 patent
provisions that have no counterpart in copyleft-next.

So, yes, principles of equality and nondiscrimination are part of the
legacy of the GPL family (and I suppose FLOSS licenses generally), but
different licenses implement those principles in different ways. For
example, pure non-copyleft licenses say "everyone getting *this* copy
gets an equal right to create proprietary derivative works".

FOSDEM: Richard Fontana on copyleft-next

Posted Feb 14, 2013 13:41 UTC (Thu) by zack (subscriber, #7062) [Link] (1 responses)

> In order to show the drastic reduction in complexity, Fontana compared
> the word and line counts of several popular open source licenses.
> The word counts were as follows:

as a data point:

$ wc -w /usr/share/common-licenses/BSD # this is 3-clause BSD
225 /usr/share/common-licenses/BSD

Arguably, BSD-3 doesn't "do" much either, and I guess the most interesting comparison points for Richard are the most recent licenses (MPL 2.0, GPLv3).

But if the goal really is comparison with *popular* open source licenses, well, excluding BSD sounds like cheating :-) </troll>

FOSDEM: Richard Fontana on copyleft-next

Posted Feb 14, 2013 14:05 UTC (Thu) by rfontana (subscriber, #52677) [Link]

> But if the goal really is comparison with *popular* open source licenses, well, excluding BSD sounds like cheating :-) </troll>

It wasn't actually meant to be a comparison to popular open source licenses. Clearly GPLv1 is not popular as a specifically chosen license, and MPL 2.0 is (AFAICT) not used much outside of the Mozilla ecosystem and LibreOffice. I picked GPLv1 to show the historical evolution in length of the GPL, and I picked MPL 2.0 because it's the most interesting case of a post-*GPLv3, pre-copyleft-next (weak) copyleft FLOSS license.

I have easily concluded that you can't make an effective, modern copyleft license much shorter than the Apache License 2.0, hence I didn't bother to compare to the popular minimalist pre-modern BSD and MIT variants.

FOSDEM: Richard Fontana on copyleft-next

Posted Feb 14, 2013 17:58 UTC (Thu) by jra (subscriber, #55261) [Link] (2 responses)

Good summary on why creating a new license is a bad idea. It is still valid today.

http://google-opensource.blogspot.com/2008/05/standing-ag...

At this point in time inventing new licenses is just an exercise in vanity, and simply serves to fragment the already limited resources we have to create more Free Software.

nevertheless, I think that this is a good effort

Posted Feb 14, 2013 18:50 UTC (Thu) by JoeBuck (subscriber, #2330) [Link]

The FSF's effort to produce GPLv3 was impressive in many ways, but in the end I consider it a failure, because it split the developer community, by trying and failing to accomplish things with licenses that just can't be done (like stop DRM: had Linus gone along with GPLv3 the device makers would simply have forked the kernel at that point). One consequence of that failure is that more projects are choosing weak copyleft or BSD-style licenses.

Having a copyleft license that is GPL2 and GPL3-compatible, that is substantially simpler than GPL3, and that isn't encumbered with unnecessary rhetoric, would be a very good thing.

FOSDEM: Richard Fontana on copyleft-next

Posted Feb 23, 2013 14:36 UTC (Sat) by SEMW (guest, #52697) [Link]

The arguments in that blog post are about the difficulties of combining code under different and incompatible licenses. But clause 4 in copyleft-next allows relicensing to arbitrary GPL versions, which solved those problems:

49 You may additionally license the Derived Work under any version of the
50 GNU General Public License published by the Free Software Foundation
51 ("GPL"), so that the recipient may further distribute the Derived Work
52 under either this License or the GPL.

So code licensed under this license is strictly more compatible than code licensed under 'any version of the GPL' -- and *significantly* more compatible than code licensed under GPLv3, which is what that blog post recommends to use if you want a copyleft license.

FOSDEM: Richard Fontana on copyleft-next

Posted Feb 14, 2013 18:54 UTC (Thu) by josh (subscriber, #17465) [Link] (7 responses)

Interesting quirk in the "Effect of Proprietary Relicensing" clause: it allows relicensing under "a license approved by the Open Source Initiative as of 1 January 2013", which would include non-copyleft licenses that do not require distribution of source code.

FOSDEM: Richard Fontana on copyleft-next

Posted Feb 14, 2013 20:43 UTC (Thu) by rfontana (subscriber, #52677) [Link] (6 responses)

That was intentional. (BTW I consider the current way this provision is drafted to be suboptimal, as I noted in the talk.) The idea was, I wanted to zoom in on 'problematic dual-licensing'. If some project does something like what jQuery used to do (MIT|GPL), this is essentially the same as just MIT licensing.

As originally conceived, this provision is supposed to apply only to the original upstream licensor. Anyone downstream is subject to the copyleft requirement already. However, I have heard a recent report from a credible source about purported copyright holders of copylefted software who are not clearly upstream-original-licensors shaking people down for proprietary licenses, in some cases trying to give the misleading impression of affiliation with some well-respected organizations active in GPL enforcement. So I made a tweak to the provision to attempt to cover that case. That conceivably could involve a downstream licensee doing the shakedown. The added complexity of covering that case may not be worth it, because I'm seeing a common confusion among some readers about what this provision is principally aimed at doing.

FOSDEM: Richard Fontana on copyleft-next

Posted Feb 21, 2013 17:14 UTC (Thu) by Wol (subscriber, #4433) [Link]

Maybe what you need is to word the clause so it doesn't apply where the code under the proprietary licence is the same as the code under copyleft.next.

That hits open-core successfully, while not affecting projects that are truly dual-licenced.

Cheers,
Wol

this provision is supposed to apply only to the original upstream licensor

Posted Feb 26, 2013 11:04 UTC (Tue) by Wol (subscriber, #4433) [Link] (4 responses)

OUCH OUCH OUCH !!!

Just seen this!!!

Which implies a total mis-understanding of copyright.

You can NEVER NEVER NEVER apply a copyright licence clause to the original licensor. The whole point of being the licens*or* is that you are granting rights to others. They can have no claim whatsoever against you.

This is what really bugs me about my understanding of this clause. I can understand the Tivo-isation clause from GPL v3, even if I don't like it. It was meant to stop A taking B's code, and locking it down against B's wishes. It was truly fixing what the FSF saw as a bug in the GPL, in that it did not properly protect the owner of the code from having their code "stolen".

The problem with this clause is it is telling B what they are and are not allowed to do with their OWN CODE. THAT IS "THEFT" (and CANNOT be enforced by copyright law).

I honestly cannot see ANY developer choosing to use a licence like that. They'll use GPL. At the end of the day, Open Source relies on enlightened self interest. If that clause is there, this licence is a waste of time, and, basically, insulting to developers!

Cheers,
Wol

this provision is supposed to apply only to the original upstream licensor

Posted Feb 26, 2013 13:21 UTC (Tue) by nye (guest, #51576) [Link]

>OUCH OUCH OUCH !!!

> Just seen this!!!

> Which implies a total mis-understanding of copyright.

> You can NEVER NEVER NEVER apply a copyright licence clause to the original licensor. The whole point of being the licens*or* is that you are granting rights to others. They can have no claim whatsoever against you.

The way this is being discussed here is confusing and wrong, but the actual licence doesn't make that mistake. The relevant clause is as follows:

>If, more than one year after My first Distribution of My Work under this
>License, I offer to license a Covered Work in a manner that fails to
>satisfy the Open Source Definition published by the Open Source
>Initiative as of 1 January 2013, then I additionally license My Work to
>You under the Apache License (excluding any preexisting material
>contained in My Work over which I do not hold copyright).

Ie. the copyright owner is granting additional rights that become available in the specified circumstances. The licence doesn't claim to change the rights of the licensor; as you say, that would be nonsense.

this provision is supposed to apply only to the original upstream licensor

Posted Feb 26, 2013 15:42 UTC (Tue) by rfontana (subscriber, #52677) [Link] (2 responses)

You can NEVER NEVER NEVER apply a copyright licence clause to the original licensor.

It actually depends on what you mean by 'apply', and what sort of copyright license you're talking about.

The whole point of being the licens*or* is that you are granting rights to others. They can have no claim whatsoever against you.

Leave aside (as you are doing) the possibility of the clause in question applying to a downstream licensor (I haven't decided if this is a problem or not), and focus on the original licensor:

The clause is a copyright license grant from the original licensor. Here is it's wording in copyleft-next 0.2.0, the latest numbered version:

If, more than one year after My first distribution of My Work under this License, I offer to license a Covered Work in a manner that fails to satisfy the Open Source Definition published by the Open Source Initiative as of 1 January 2013, then I additionally license My Work to You under the Apache License [2.0] (excluding any preexisting material contained in My Work over which I do not hold copyright).

This is just saying that, if certain circumstances occur, in addition to granting you a license under copyleft-next, I also grant you a license under the Apache License.

I would not say this gives the downstream licensee a 'claim' against the original licensor, as worded, so much as a defense against a claim by the original licensor for copyright infringement. So, for example, if I'm the original licensor, and I engage in copyleft-next/proprietary dual licensing, and a year goes by and I don't stop this practice, by operation of the license I give you the opportunity to receive the software under an additional license, the Apache License 2.0. If you decide to create a proprietary derivative work out of my code -- much as I myself did -- and I sue you for copyright infringement arising out of your failure to follow the copyleft requirements of the license, my defense is that that very license gave me an escape route to use the Apache License, which is non-copyleft and permits creation of proprietary derivative works.

The problem with this clause is it is telling B what they are and are not allowed to do with their OWN CODE. THAT IS "THEFT" (and CANNOT be enforced by copyright law).

No, it's not theft. As noted, it's a copyright license grant, giving rise to a defense against copyright infringement. The copyright holder obviously cannot enforce this against itself, but that's not the point. It's designed as a self-imposed obstacle for the copyright holder. Much as any developer who uses the BSD or MIT or Apache license cannot complain if a downstream licensee creates a proprietary derivative work.

I honestly cannot see ANY developer choosing to use a licence like that. They'll use GPL.

Developers who wish to reserve the option of a viable business model based on monopolization of the right to create proprietary derivative works of a copyleft codebase are better off using the GPL or AGPL (or some legacy copyleft free or pseudo-free license). Developers interested in a license that discourages such conduct should look into copyleft-next, since it is the first license to attempt to do something about this issue.

this provision is supposed to apply only to the original upstream licensor

Posted Feb 26, 2013 16:22 UTC (Tue) by mpr22 (subscriber, #60784) [Link] (1 responses)

Any entity that wishes to engage in copyleft/proprietary dual licensing already has to either reject any external contribution that would be adjudged significant enough to be copyrightable, or demand a contributor agreement of some sort from everyone who wants to contribute. They also have to either start from nothing, or be derived from existing permissively-licensed code.

As such, I really can't see what this license aims to achieve. Anyone who wasn't planning to engage in the behaviour it's targeting doesn't need it; anyone who was planning to engage in such behaviour won't use it.

this provision is supposed to apply only to the original upstream licensor

Posted Feb 26, 2013 19:40 UTC (Tue) by rfontana (subscriber, #52677) [Link]

> Any entity that wishes to engage in copyleft/proprietary dual
> licensing already has to either reject any external contribution
> that would be adjudged significant enough to be copyrightable, or
> demand a contributor agreement of some sort from everyone who wants
> to contribute. They also have to either start from nothing, or be
> derived from existing permissively-licensed code.

Correct.

> As such, I really can't see what this license aims to
> achieve. Anyone who wasn't planning to engage in the behaviour it's
> targeting doesn't need it;

This is just one of many features of copyleft-next that might be
attractive to some developers. But I suppose I was thinking also that
some developers might like the fact that here, for the first time, is
a license that takes a stand against this behavior.

> anyone who was planning to engage in such
> behaviour won't use it.

That is really the main thing the license seeks to achieve through
this provision. There are benefits to keeping a copyleft license
'pure' by warding off use of it for copyleft/proprietary
dual-licensing. As an example, it will minimize the problem of
pecuniarily-motivated unreasonably-restrictive interpretations of the
license. It will also enhance the ethical reputation of the license
and the community of developers choosing to use it.

Updated URL

Posted Oct 22, 2014 12:25 UTC (Wed) by hmh (subscriber, #3838) [Link]

Just for the record: GPL.next, now named copyleft-next, is here:

https://gitorious.org/copyleft-next