[RSS] Conservancy Blog

Displaying posts by Bradley M. Kuhn

Challenges in Maintaining A Big Tent for Software Freedom

by Bradley M. Kuhn on August 30, 2018

In recent weeks, I've been involved with a complex internal discussion by a major software freedom project about a desire to take a stance on social justice issues other than software freedom. In the discussion, many different people came forward with various issues that matter to them, including vegetarianism, diversity, and speech censorship, wondering how that software freedom project should handle other social justices causes that are not software freedom. This week, (separate and fully unrelated) another project, called Lerna, publicly had a similar debate. The issues involved are challenging, and it deserves careful consideration regardless of how the issue is raised.

One of the first licensing discussions that I was ever involved in the mid 1990s was with a developer, who was a lifelong global peace activist, objecting to the GPL because it allowed the USA Department of Defense and the wider military industrial complex to incorporate software into their destructive killing machines. As a lifelong pacifist myself, I sympathized with his objection, and since then, I have regularly considered the question of “do those who perpetrate other social injustices deserve software freedom?”

I ultimately drew much of my conclusion about this from activists for free speech, who have a longer history and have therefore had longer time to consider the philosophical question. I remember in the late 1980s when I first learned of the ACLU, and hearing that they assisted the Klu-Klux Klan in their right to march. I was flabbergasted; the Klan is historically well-documented as an organization that was party to horrific murder. Why would the ACLU defend their free speech rights? Recently, many people had a similar reaction when, in defense of the freedom of association and free speech of the National Rifle Association (NRA), the ACLU filed an amicus brief in a case involving the NRA, an organization that I and many others oppose politically. Again, we're left wondering: why should we act to defend the free speech and association rights of political causes we oppose — particularly for those like the NRA and big software companies who have adequate resources to defend themselves?

A few weeks ago, I heard a good explanation of this in an interview with ACLU's Executive Director, whom I'll directly quote, as he stated succinctly the reason why ACLU has a long history of defending everyone's free speech and free association rights:

[Our decision] to give legal representation to Nazis [was controversial].… It is not for the government's role to decide who gets a permit to march based on the content of their speech. We got lots of criticism, both internally and externally. … We believe these rights are for everyone, and we truly mean it — even for people we hate and whose ideology is loathsome, disgusting, and hurtful. [The ACLU can't be] just a liberal/left advocacy group; no liberal/left advocacy group would take on these kinds of cases. … It is important for us to forge a path that talks about this being about the rights of everyone.

Ultimately, fighting for software freedom is a social justice cause similar to that of fighting for free speech and other causes that require equal rights for all. We will always find groups exploiting those freedoms for ill rather than good. We, as software freedom activists, will have to sometimes grit our teeth and defend the rights to modify and improve software for those we otherwise oppose. Indeed, they may even utilize that software for those objectionable activities. It's particularly annoying to do that for companies that otherwise produce proprietary software: after all, in another realm, they are actively working against our cause. Nevertheless, either we believe the Four Software Freedoms are universal, or we don't. If we do, even our active political opponents deserve them, too.

I think we can take a good example from the ACLU on this matter. The ACLU, by standing firm on its core principles, now has, after two generations of work, developed the power to make impact on related causes. The ACLU is the primary organization defending immigrants who have been forcibly separated from their children by the USA government. I'd posit that only an organization with a long history of principled activity can have both the gravitas and adequate resources to take on that issue.

Fortunately, software freedom is already successful enough that we can do at least a little bit of that now. For example, Conservancy already took a public position, early, in opposition of Trump's immigration policy because of its negative impact on software freedom, whose advancement depends on the free flow of movement by technologists around the world. Speaking out from our microphone built from our principled stand on software freedom, we can make an impact that denying software freedom to others never could. Specifically, rather than proprietarizing the license of projects to fight USA's Immigration and Customs Enforcement (ICE) and its software providers, I'd encourage us to figure out a specific FOSS package that we can prove is deployed for use at ICE, and use that fact as a rhetorical lever to criticize their bad behavior. For example, has anyone investigated if ICE uses Linux-based servers to host their otherwise proprietary software systems? If so, the Linux community is already large and powerful enough that if a group of Linux contributors made a public statement in political opposition to the use of Linux in ICE's activities, it would get national news attention here in the USA. We could even ally with the ACLU to assure the message is heard. No license change is needed to do that, and it will surely be more effective.

Again, this is how software freedom is so much like free speech. We give software freedom to all, which allows them to freely use and deploy the software for any purpose, just like hate groups can use the free speech microphone to share their ideas. However, like the ACLU, software freedom activists, who simultaneously defend all users equal rights in copying, sharing and modifying the software, can use their platform — already standing on the moral high ground that was generated by that long time principled support of equal rights — to speak out against those who bring harm to society in other ways.

Finally, note that the Four Software Freedoms obviously should never be the only laws and/or rules of conduct of our society. Just like you should be prevented from (proverbially) falsely yelling Fire! in a crowded movie theater, you still should be stopped when you deploy Free Software in a manner that violates some other law, or commits human rights violations. However, taking away software freedom from bad actors, while it seems like a panacea to other societal ills, will simply backfire. The simplicity and beauty of copyleft is that it takes away someone's software freedom only at the moment when they take away someone else's software freedom; copyleft ensures that is the only reason your software freedom should be lost. Simple tools work best when your social justice cause is an underdog, and we risk obscurity of our software if we seek to change the fundamental simple design of copyleft licensing to include licensing penalties for other social justice grievances (— even if we could agree on which other non-FOSS causes warrant “copyleft protection”). It means we have a big tent for software freedom, and we sometimes stand under it with people whose behavior we despise. The value we have is our ability to stand with them under the tent, and tell them: “while I respect your right to share and improve that software, I find the task you're doing with the software deplorable.”. That's the message I deliver to any ICE agent who used Free Software while forcibly separating parents from their children.

Tags: conservancy, law, licensing

Software Freedom Ensures the True Software Commons

by Bradley M. Kuhn on August 22, 2018

Update (2023-11-14): Not long after publication of the post below about the so-called “Commons Clause”, Neo4j utilized that Clause to add a “further restriction” to the Affero General Public License, Version 3 (“AGPLv3”). When John Mark Suhy tried to remove it, as permitted and encouraged by AGPLv3 itself, Neo4j sued John Mark and his small company (PureThink). That case sadly resulted in multiple ill-informed judgments forbidding such removal. As SFC's Policy Fellow, I filed an expert report on the case — explaining my first-hand knowledge about the drafting of the relevant “further restriction” removal clause, but, alas, the Court has not changed its view. The case is, however, ongoing, so please watch SFC's website for updates.

The SFC remains dismayed that the proprietary “AGPLv3-only WITH Commons-Clause” has been allowed to stand as valid since 2018. If the SFC were the copyright holder of the text of the AGPLv3, or the trademark holder of the license's name, we would have intervened in this case to clarify these matters for the Court. Since the SFC did not create the AGPLv3 (my personal involvement with AGPLv3 drafting was not as an SFC employee), filing the expert report was the only action that SFC could take to assist in this matter. However, going forward, we do encourage anyone facing a “further restrictions” issue with copyleft license to contact us for support — so that those who care about the future of copyleft can coordinate a response together.

(Original post follows:)


Proprietary software has always been about a power relationship. Copyright and other legal systems give authors the power to decide what license to choose, and usually, they choose a license that favors themselves and takes rights and permissions away from others.

The so-called “Commons Clause” purposely confuses and conflates many issues. The initiative is backed by FOSSA, a company that sells materiel in the proprietary compliance industrial complex. This clause recently made news again since other parties have now adopted this same license.

This proprietary software license, which is not Open Source and does not respect the four freedoms of Free Software, seeks to hide a power imbalance ironically behind the guise “Open Source sustainability”. Their argument, once you look past their assertion that the only way to save Open Source is to not do open source, is quite plain: If we can't make money as quickly and as easily as we'd like with this software, then we have to make sure no one else can as well.

These observations are not new. Software freedom advocates have always admitted that if your primary goal is to make money, proprietary software is a better option. It's not that you can't earn a living writing only Free Software; it's that proprietary software makes it easier because you have monopolistic power, granted to you by a legal system ill-equipped to deal with modern technology. In my view, it's a power which you don't deserve — that allows you to restrict others.

Of course, we all want software freedom to exist and survive sustainably. But the environmental movement has already taught us that unbridled commerce and conspicuous consumption is not sustainable. Yet, companies still adopt strategies like this Commons Clause to prioritize rapid growth and revenue that the proprietary software industry expects, claiming these strategies bolster the Commons (even if it is a “partial commons in name only”). The two goals are often just incompatible.

Here at Conservancy, we ask our projects to be realistic about revenue. We don't typically see Conservancy projects grow at rapid rates. They grow at slow and steady rates, but they grow better, stronger, and more diverse because they take the time to invite everyone to get involved. The software takes longer to mature, but when it does it's more robust and survives longer.

I'll take a bet with anyone who'd like. Let's pick five projects under the Affero GPL and five projects under the Commons Clause, and then let's see which ones survive longer as vibrant communities with active codebases and diverse contributors.

Finally, it's not surprising that the authors chose the name “Commons”. Sadly, “commons” has for many years been a compromised term, often used by those who want to promote licenses or organizational models that do not guarantee all four freedoms inherent in software freedom. Proprietary software is the ultimate tragedy of the software commons, and while it's clever rhetoric for our opposition to claim that they can make FOSS sustainable by proprietarizing it, such an argument is also sophistry.

Tags: conservancy, GPL, CLA, law

Update on Trademark Action (Fraud Claim Dismissed, New Filing)

by Bradley M. Kuhn on April 30, 2018

As per our commitment to transparency, here's a brief update on the trademark action we have discussed previously.

On Thursday, the Trademark Trial and Appeals Board (TTAB) dismissed SFLC's claim of fraud against Conservancy, as we predicted.

This sensible ruling is just the next step of many; the suit will proceed at the usual near-glacial pace of litigation. On Friday, we moved to the next procedural step, which is to ask the TTAB to allow us to update our answer as planned.

UPDATE: as of 2018-08-24, the Summary Judgement motion is refiled and pending before the TTAB.

We continue steadfast in our previous position: the entire matter remains a waste of resources for both organizations, and SFLC should do the honorable and right thing and simply withdraw their complaint.

Tags: conservancy

Why Scènes À Faire Should Apply to Command-Line Interfaces

by Bradley M. Kuhn on January 3, 2018

Today, Conservancy joined other amici in the Cisco v. Arista case. Specifically, the amicus brief discusses why the scènes à faire affirmative defense for copyright infringement is appropriate and actually necessary regarding imitation of command-line interfaces. I hope this blog post will convince you that software freedom contributors should care about the issue.

The easiest example to understand these issues is Unix. Most of us know the basics of Unix's user interface, which primarily consists of commands that live in /bin and /usr/bin, that each include various command-line options that we've memorized. When the GNU project started, as RMS has described in his talks, he chose to imitate this user interface. Many reasons were obvious, but the most important one was that Unix was already an industry standard and users already knew its interface.

At the time, no one would have considered that you'd be liable for copyright infringement merely writing some new programs — 100% from scratch — that happened to have the same names and the same command-line options that were found in Unix. That interface, in fact, has been reimplemented at least a hundred times — by many Unix vendors and by various software freedom projects (GNU of course, but also by Conservancy's BusyBox project and others). As developers, we'd be incredulous if told that GNU infringed Unix's original copyrights. But that's exactly the argument that Cisco made about Arista's imitation of Cisco's command-line interface.

I'm not a fan of either Cisco nor Arista; all the software in question is proprietary software. Indeed, GitHub, which is one of our joined amici here, produces much proprietary software around Git, and that's bothers me too. I don't like it when any company writes proprietary software to work along with FLOSS. However, I agree with GitHub and Arista that copyright restrictions should not extend too far; copyright should not stifle simple command-line interoperatiblity. Merely imitating a command-line interface of one program in another should not cause (by itself) a copyright infringement.

Now, the last part to discuss are the questions: What is an affirmative defense, and what is scènes à faire? So, to explain it roughly with as little legalese (IANAL) as possible, an affirmative defense is one that you must prove after you're accused, usually through a trial (which is what occurred here). The burden is on the Defendant to prove that affirmative defense. (By contrast, if Arista had shown that, in fact, their command-line interface bore no similarity to Cisco's, that would have been a “negating defense”. Such defenses are much more assured to win, as they don't place such a burden on the defense.)

So, what, specifically, is the affirmative defense of scènes à faire? It's a concept originally from fictional works that generally expresses this idea: “if you're going to tell this story at all, you need at least these elements”. In this example, the analogy works like this: if your users will give a router textual commands via the command-line, that user will expect certain commands to work. Cisco's commands are industry standard and expected by users, similar to those in Unix. The amicus brief argues that this is a reasonable application of scènes à faire, because there is great benefit to the public and users if such imitation is permitted on command-lines without copyright restriction. Remember, under the USA Constitution, copyright exists as an “exclusive Right to … Writings” only because such exclusive controls “promote the Progress of Science and useful Arts”. Copyright is not an absolute right of control over written works by the authors, and its tentacles must be shortened by the public interest.

Finally, I call on the Linux Foundation to publicly ask their platinum member, Cisco, to stop this aggressive litigation on an edge case of copyright. Such a request would be consistent with the Linux Foundation's public criticism of others for copyright enforcement. This case is one where we all should stand together in the interests of free innovation.

Tags: conservancy, law

Next page (older) » « Previous page (newer)

1 2 3 4 5 6 [7] 8 9 10 11 12 13 14 15 16 17

Connect with Conservancy on Fediverse, X, Facebook, and YouTube.

Main Page | Contact | Sponsors | Privacy Policy | RSS Feed

Our privacy policy was last updated 22 December 2020.