Blog by Sumana Harihareswara, Changeset founder

26 Mar 2021, 16:44 p.m.

Not The First Time We Tried (FSF, GNU, RMS, etc.)

Here's the open letter in which thousands of people and several organizations ask for major changes at the Free Software Foundation and GNU in light of the FSF's recent and extremely frustrating choices. I haven't signed the letter; I am in a position much like Andy Wingo's (I have some obligations to GNU Autoconf that are not yet discharged). But I agree, for instance, that FSF needs a new board, and I want to put something down here to mark my solidarity with so many who have spoken up in the last week.

Some of my peers think this is the first major effort to file a bug about Richard M. Stallman and about FSF and GNU governance, that critics went directly to asking for his removal. It's not the first bug report or request for negotiation; far from it (as former FSF board members Bradley Kuhn and Matthew Garrett discussed in 2019). Today's FSF board announcement would be more promising if it didn't follow years of one-step-forward, two-steps-back conversations about FSF and GNU governance. I have participated in past attempts to talk about these problems with FSF and GNU, with lower stakes, and I figured I'd lay that out here.

In particular, you ought to know that the FSF and GNU have repeatedly failed at fair and consistent treatment. I care about everyone having to obey the same rules and having the same freedoms and the same opportunities; the FSF and GNU have demonstrated that they do not. Boring stuff ahead: Caution that this is a somewhat long and boring post about governance, policies, and similar dry stuff. Skip to the "Fairness" section for the wrapup.


My first time shaking my head and sighing at something Stallman had done at a FLOSS conference was in 2009, if not earlier. And over the years I heard more and more. In particular, I became aware of multiple instances of inappropriate behavior over the years at the FSF's conference, LibrePlanet, such as taking over sessions through loud disruptions. And, in 2017, RMS explicitly said that, as president of the FSF, he was not subject to the rules in the Q&A of Marianne Corvellec's 2017 talk (here is a recording).

During the session submission period (in late 2018) for LibrePlanet 2019, a significant number of former speakers, including myself, jointly contacted the Free Software Foundation Board of Directors. In our message, we expressed concern to the Board over inconsistencies in how the LibrePlanet Safe Space Policy is applied to members of the Board itself.

During discussion with the Board over a few weeks, the group expressed the critical need for LibrePlanet's Safe Space Policy to apply to all participants, including all the members of the Board, which included Stallman, FSF Board President. During the discussion, the Board did not address the following specific actions we requested:

  • That the Board explicitly clarify that if RMS violates the Safe Space Policy again organizers will step up and impartially apply the rules to him.
  • That the Board appoint a disinterested ombudsman or committee with the power to oversee and handle safe space policy complaints, and enforcement, including for Safe Space Policy complaints against members of the Board.

On November 8, 2018, the Board told us they were working on a response but that it would take time. In early February 2019, many of us got individual email replies from one of the FSF Board of Directors, representing the Board. The response from the Board said FSF was working with a third-party consultant to improve safety procedures. I hoped to see a public announcement of the name of the consultant, or a Code of Conduct transparency report after the con (example). We have seen neither.

The board also said that safe space policy complaints against FSF staff, board members, and officers would be handled by other members of the board, rather than appointing a disinterested ombudsman or committee. But in the last few years, outgoing and former FSF board members including Bradley Kuhn, Benjamin Mako Hill, and Kat Walsh (further re: Walsh) have all tried to use their Board and Voting Member seats to appropriately limit Stallman's behavior and authority, and were evidently unable to form a majority to do so.


In May 2018, during the discussion of a controversial joke in glibc's documentation, I emailed a few FSF staff and board members as well as the GNU Advisory Committee. I shared my assessment of the relevant policy documents:

The governance question currently affecting glibc (context: hinges on whether Richard has the authority he claims to have and whether he is a responsible user of that authority. I've reviewed the conversation and relevant policy documents.*

In my assessment, while Richard has a tenuous claim on his privilege as Chief GNUisance to prohibit the removal of a joke he wrote from the glibc documentation, his choice raises questions about his fitness for the role of Chief GNUisance, especially as the joke contravenes GNU documentation standards ("Make sure your manual is clear to a reader who knows nothing about the topic and reads it straight through.") and "Information for Maintainers of GNU Software" ("Don’t feel obligated to include every change that someone asks you to include. You must judge which changes are improvements—partly based on what you think the users will like, and partly based on your own judgment of what is better. If you think a change is not good, you should reject it."). In my experience as a free software community leader, Richard's choice is also apt to cause attrition among existing GNU maintainers and contributors, which will slow work towards GNU's goals.


* for reference: 0.

I asked:

What would the criteria be for re-evaluating Richard M. Stallman's position as Chief GNUisance?

If the answer is that there are and can be no such criteria, and there is no change in circumstance or aspect of Richard's behavior that would cause GNU to demote Richard and promote someone else to Chief GNUisance, then I would like that explicitly and publicly stated.

And if the answer is that there exist such criteria, or that you would like to develop them but have not yet done so, then I would like that explicitly and publicly stated.

The GNU Advisory Committee as a whole did not make a formal reply; a few individual members replied with criticism of my "bullshit". I did not pursue the question further.

A year and a half later, in late 2019, Stallman was not on the Board of Directors of the FSF, but still claimed leadership of the GNU Project. A collective of GNU maintainers signed a statement saying, "We think it is now time for GNU maintainers to collectively decide about the organization of the project." I know of no public response to this statement by Stallman or the GNU Advisory Committee. "The FSF never officially helped or even replied to our requests to formulate an open and welcoming working relationship with us as GNU volunteers."

In February 2020, the FSF published a post saying that "The Free Software Foundation and the GNU Project leadership are defining how these two separate groups cooperate." They asked for input from the public. I wrote to them on February 12th to ask:

  • There are many standards by which to judge whether an institution's work is moving the mission forward, and by which to judge whether a person in a position of leadership is acting as a good steward of that institution. For FSF and GNU, those standards might include fundraising and money management, effective public speaking and writing, keeping principles like the Franklin Street Statement in the public consciousness, providing compelling new visions and articulations for the movement when necessary, growing the number and types of volunteers who contribute to crucial free software projects, making strategic choices to improve free software infrastructure, recruiting and mentoring successor leaders, offering useful architectural guidance or code review, and project management. Which standards are appropriate to judge FSF, and which for GNU, and which for both? And how can we ensure that each organization regularly checks that the other is staying on course?

  • Similarly: let us look forward to what the FSF and GNU can do in the next several years. Can we make strides on things free software developers care about -- including platform support to help developers do their work better, financial infrastructure, protecting freedom in concrete ways, and recruiting new people into the free software movement and retaining them? Which organizations' leaders are genuinely interested in and excited about those goals?

  • How shall the board members, and the voting members, for the FSF be chosen? How can you ensure that those lists are transparent (I believe right now the list of voting members is not public) and that FSF members like me have a voice in those appointments and elections?

  • Similarly, for GNU: how is the leadership chosen, what voice do GNU volunteer contributors have in choosing leadership and policies, and is that process in accordance with free software values?

  • As of 2019, FSF was working with a third-party consultant to improve safety procedures at events such as LibrePlanet. What were the outcomes of that consulting, and did it result in FSF adopting or strengthening any principles or procedures that should be included in a partnership framework with GNU?

I received an acknowledgment that my message had been received, but nothing further, and neither GNU nor the FSF ever made an announcement regarding the cooperation framework. The only subsequent announcement was an email announcement by Stallman, later that month, documenting the existing GNU governance structure.


Fairness is at the heart of free software values. We set policies -- the GPL, for instance -- that apply to everyone equally.

To remix another phrasing, no one is indispensable and no one is disposable.

I contributed, in good faith, to efforts to address the unfair, inconsistent treatment of Richard Stallman regarding LibrePlanet safety standards and in the GNU maintainer procedures and documentation policies. Others did similarly in several other areas. I joined my fellow free software advocates in doing this not just because of the individual incidents that free software advocates have been reporting for many years. I did this because of the meta-behavior of Stallman's claim that he is not subject to the same rules as everyone else. To borrow a phrase, I was tired of treating Stallman like a missing stair. So we filed bugs about the situations caused by his behavior, and by FSF's and GNU's unwillingness to consistently apply their policies.

The grief of having all those efforts marked as WONTFIX causes some ache -- as does the dismissive attitude I see from some peers in the tech industry, as though our care and work were foolish and useless. As though this setback means we should scrap the whole endeavor.

To me the history of free software is partly the history of us making it better, on a social and infrastructural level, the same way that the history of the United States (I'm an American) is partly the history of us making it live up to our ideals. With words and actions (codes of conduct, The Carpentries, Software Freedom Conservancy, Python's Steering Council model, so much more) we've been steadily working to improve, and -- with or without the FSF and GNU -- we will keep working.

I've been using GNU/Linux for more than twenty years, free software is my profession, freeing people using free software is core to my values, and I ain't stopping.