halting problem :: codes of conduct

:: ~4 min read

you would think that, in 2014, implementing a code of conduct for conferences or conventions would not be a controversial topic. sadly, you'd also be mistaken. there are various contrarian positions about implementing anti-harassment policies; most, if not all of those positions are wrong.

a discussion on guadec-list about adopting a code of conduct for GUADEC prompted me to write down some thoughts about the issue.

GNOME, as a community, was if not the first, one of the first high profile free software foundations to define and implement a code of conduct.

Photo credit: Jonathan Thorne, CC by-nc-2.0

to be perfectly, honest I never thought an anti-harassment policy would be a controversial issue at all in 2014, after the rate of adoption of codes of conduct, and of anti-harrasment policies, at convention and conferences all over the world. there have been high profile cases, and speakers as well as attendees have finally started to stand up, and publicly state that they won’t attend a convention or a conference (even if sponsored) if the organizers do not put in place these kind of documents.

GNOME, as a community, was if not the first, one of the first high profile free software foundations to define and implement a code of conduct. I’ll actually come back to that later, but that was a point of pride for me.

yet, I have to admit that my heart sank a bit for every email in the discussion on guadec-list, especially because they were from members of our own community.

I do understand that we like our conferences like we like our software: free-as-in-speech, and interesting to work on.

as I said, GNOME already has a code of conduct, pitiful and neutered as it may be, and it applies to every venue of communication we have: mailing list, our web servers, and also our conferences. to have and implement a code of conduct is not a per-GUADEC-edition, local-team-only decision to take — and how do I know that? because it’s the board that approved the code of conduct for the mailing lists, IRC, and web servers, and it was not the moderation team, or the IRC operators, or the system administrators that took this decision.

I do understand that we like our conferences like we like our software: free-as-in-speech, and interesting to work on. that does not imply that we should just assume bad stuff won’t happen, or that people will automatically find the right person to help them because somebody else decided to be a jerk. to be fair, our software is full of well-defined rules for redistribution, and our conferences should be equally well-defined when it comes to acceptable behaviour and responsible people to contact. why we do that? because it helps in having a clear set of rules and people responsible to avoid abuse, if that happens.

I am lucky enough, and privileged enough, that I have not been discriminated for who I am, what I like, who I like, what I do, or how I do it.

I honestly have zero patience for the people saying that «everyone can be offended by something, thus we shouldn’t do anything». the usual, trite, argument is that having an anti-harassment policy will provide a “chilling effect” on attendees; it should also not be necessary to have these policies, because we trust our community to be composed of good people, and these policies automatically assume that everyone will be misbehaving. those are both ridiculous positions, even if they are sadly fairly common in the free and open source community at large. they are based on a fairly obvious misunderstanding; the code of conduct is not a sword for preventing people from misbehaving: it’s a shield for people being the object of discrimination and harassment.

Photo credit: Jenn and Tony Bot, CC by-nc-2.0

I am lucky enough, and privileged enough, that I have not been discriminated for who I am, what I like, who I like, what I do, or how I do it. others are not in such position, and they attend GUADEC. we want them to attend GUADEC, because they are our next contributor, our next user, our next tester, our next designer, our next bugsquad member, our next person submitting a documentation patch. I don’t want them to be placed in a position where they balk at the idea of participating at GUADEC because they don’t feel safe enough, because they aren’t part of our community yet. you want to talk about a chilling effect? that is the chilling effect. I have no concerns for members of our community: I know that most of them can actually behave like actual human beings in a social context. that knowledge comes from 10+ years in this community. I don’t expect, and I’d be foolish to do so, that new people that have not been at GUADEC yet, or have been newly introduced to our community, also posses that knowledge.

Photo credit: diffendale, CC by-nc-sa 2.0

a final thought: I actually want a better code of conduct for GNOME’s online services as well, one that is clear on responsibility and consequences, because our current one is a defanged travesty, which was implemented to be the lowest common denominator possible. it does not require responsibility for enforcing it, and it does not provide accountability for actually respecting it. it is, for all intents and purposes, like not having one. it probably was thought as a good compromise eight years ago, but it clearly is not enough any more, and it makes GNOME look bad. changing the code of conduct is a topic for the new board, one that I expect will be handled this year; I’ll make sure to prod them. ;-)


I’d like to thank to Marina and Karen for reviewing the draft of this article and for their suggestions.

more information

  1. How will our Code of Conduct improve our harassment handling?
  2. Code of Conduct
  3. Codes of Conduct 101 + FAQ
  4. Conference anti-harassment
  5. My New Convention Harassment Policy
  6. Convention Harassment Policy Follow-Up

essays free-software communities

Follow me on Mastodon