The dire state of NixOS's moderation culture

1 year ago (github.com)

I coincidentally started lurking and responding to some support threads the NixOS discourse a few weeks ago, because I appreciate the project and I wanted to get more involved in it. In the past couple days I've seen this train wreck play out in real time, both on discourse and on github discussions.

From what I could gather, the main factor that led the community to this point is Anduril's involvement in the community. Anduril is a miltech company and a contractor of the USA DoD. It looks like they use Nix extensively to train ML models and deploy combat drones, and several of their employees have very close ties to the Nix community; Jon Ringer (an Anduril employee, who is mentioned several times in this document) has recently been a release manager for three versions of NixOS, and was nominated to become a release manager for the upcoming 24.05 release.

Anduril wanted to sponsor NixCon last year because of this deep involvement in the community. I was not there to see it, but from what I've seen lately, many members of the community were not happy about it. The sponsorship was eventually pulled, because the on-site video team for the conference (https://c3voc.de/) threatened to pull out on their own over this sponsorship.

From my understanding there has been no official decision taken over Anduril's sponsorship since last year's debacle. Many people in the community think that this is because Eelco Dolstra, the creator and de-facto BDFL of Nix, is currently employed by Determinate Systems, which is rumored to have a contract with Anduril.

Edit: I just want to make it clear that I don't agree with what is presented in this document. I just wanted to present an overview of what, from my perspective, happened up to this point.

  • As someone with decision making power in a small Canadian robotics firm that occasionally does military work (mostly environmental monitoring), this does make nixos a lot less attractive. The reproducible builds make it very attractive for reliable robotics, but if the nix community is that against me taking money from the military, well that's going to present a whole other set of challenges. Even for the 90% of our work that doesn't involve any potential military use.

    • > if the nix community is that against me taking money from the military

      As far as I understand it, the objection would be if

        - you were simultaneously a key person in Nix, like a release manager
        - your company officially sponsored events that associated Nix with military
        - several of your employees were active contributors to core parts of Nix
      

      I think the last one might not even be a big issue, or at least it might not blow up.

      Surely, the usage of Nix in any domain is a free choice. You can't really police that anyways.

      2 replies →

    • The main problem is

      * taking non-anonymous donations, making Nix seem like the extended arm of the military industrial complex

      * DetSys being the de-facto nix consultancy due to employing the BDFL, not moving work done / guarantees given (installer, flake stability) back into the foundation and taking money from the same military industrial complex

      FOSS is open to everyone and one can't stop people from using it; that doesn't mean the foundation (and it's de-facto corporate arm) need to openly become dependent on military contractors either.

    • > this does make nixos a lot less attractive

      Why should they care? You try to make yourself sound like some kind of important user they should cater to, which is absurd. They don't want you to use it for military stuff, so it's not a big gotcha that someone with "decision making power" (wow!) doesn't want to use it.

      Like, most FOSS projects just gets abuse and entitlement from their users, but actually little value from their users. You would be no different, and frankly sound entitled just from the get-go.

    • I personally hope some sort of reasonable policy will come from this whole situation and that the community will be stronger for it. I've seen some discussions over this on Discourse, but the community doesn't seem to be in any state to agree on something like this currently.

  • Eelco (alongside his co-founder) is Determinate Systems, and their active refusal to answer if they are involved in the military contract likely means they are.

  • > The sponsorship was eventually pulled, because the on-site video team for the conference (https://c3voc.de/) threatened to pull out on their own over this sponsorship.

    This isn't accurate.

    Per a summary at the time, https://discourse.nixos.org/t/nixcon-2023-sponsorship-situat...

    - c3voc resolved to not redistribute talks from Anduril (or with Anduril branding), but would otherwise cover the event.

    - Anduril was withdrawn as a sponsor, since the venue had a policy against military funding (https://en.wikipedia.org/wiki/Civil_clause), and time constraints didn't allow for clarity about whether the event could be held with Anduril as a sponsor.

    FWIW, at the time of writing this post, Anduril is a sponsor of Nixcon 2024. https://2024-na.nixcon.org/

This submission link concretely and factually documents the various malfeasances of the NixOS moderation team in the last few months, and is linked from NixOS RFC 175 as supporting evidence: https://github.com/NixOS/rfcs/pull/175

The two authors of this RFC, shortly after they opened the PR, got banned by the NixOS moderators.

You may join the following Matrix room to discuss the RFC openly, https://app.element.io/#/room/#rfc-175-all-together:matrix.o...

  • Good.

    I can't believe some people are upset because other people don't want to take money from a company that specializes in killing people.

    • That's who sponsors safe programming systems. It's why we have Ada.

      America is very naturally safe because we're on another continent than everyone else, but eventually if you don't build weapons someone else will and fire them at you.

From this document, it appears something obviously wrong is happening in this community, but as an outsider new to all this, it's very hard to understand what's going on.

The document is very long and I don't really get the structure. It's very hard to make anything of it really. It somewhat makes me suspicious, even. I would write exactly like this if I knew how to do this and had to sidestep an issue. Long, hard to follow stuff filled with words.

Is there a clear summary, an easy-to-follow timeline? (of course these are always going to be presented from one side)

Also,

> Simultaneously, this group, still upset about the failure of RFC 98, is using the myth of fascism combined with an abusive extension of the paradox of tolerance

- What is the myth of fascism?

- What is the abusive extension of the paradox of tolerance?

I get the paradox of tolerance is that you can't really tolerate intolerance or else you'll be eaten. [1]

> The paradox of tolerance states that if a society's practice of tolerance is inclusive of the intolerant, intolerance will ultimately dominate, eliminating the tolerant and the practice of tolerance with them

[1] https://en.wikipedia.org/wiki/Paradox_of_tolerance

  • If you want a summary I can give you a really quick one. I've watched this bias, manipulative and corrosive moderating action escalate from the moment the team was created.

    I was there arguing against the moderation team when it was first founded because I believe in principles of consent and civility, not compulsion and coersion, which were already starting to manifest themselves in argument for the moderation team to begin with. I was very vocal at the time, but not radical, when I lost I didn't fight it. I just consented and watched as things got worse.

    I've watched a lot of people, some who were quite prolific contributors walk away, I've watched extremely well reasoned positions, including some of my own going completely ignored or silenced, and I've watched technical problems that could and should have been addressed by some of these very same people (some of which were actively being worked on previously) go completely stale.

    I understand the document is "a lot of words", which I am not necessarily a big fan of either. But we have tried several times to give specific examples of corruption and been shot down, so we felt the only way to show proper cause at this point, that would be impossible to dismiss, would be to compile a much longer and more thorough narrative.

    And as it states at the top, it is by no means exhaustive. Only a few months worth of activity in the 4 years or so I've been watching this trainwreck. Also, the last few days have shown us almost and equal if not moreso amount of strife.

  • It appears the bulk of the actual requested changes are here: https://github.com/NixOS/rfcs/pull/175

    When there's stuff like this going on, I don't think it's helpful to try and do a full review of who did what wrong, or what the hypothetical motivations of any of these groups of people are, but instead to just review the actual proposed policy changes.

    • Yes and no.

      It's often useful to know the motivation / reason behind a change. Especially a policy change. Maybe the change can subtly enable harmful stuff you don't see coming if you don't have the context around the change.

      You also want to know the intents being a change to know if the change actually works towards the intended goal while reviewing. Or else how do you check if the change help?

      It would be like a code change that doesn't say what it fixes or add, how do you check it does the right things?

  • The Nix moderation team is woke and they're purging non-woke people from Nix.

The entire thing seems like a takeover attempt partially because of dislike towards Eelco partially due to some peeps being power hungry. First RFC98, now the open letter. Kinda reminds me the ffmpeg/libav situation.

OP is also the one that was banned from nix communities. How did such a childish ban appeal could reach top HN ???

  • I have no idea. Reddit seems to mostly be on this dude's side too, while most other places I frequent related to Nix do not.

    There's certainly a lot to break down, but I'll start with Jon calling the behavior of associating him with Anduril "doxing", when it's literally on his linkedin and his buddies are doing full-name callouts in this supplementary article to RFC 175, or how he starts a reddit thread that he says isn't meant to stir the pot, but then goes on in the comments about being a brave fascist-labeled throught-criminal, when he mostly got banned for being extremely abrasive / concern-trolling alongside injecting himself into every single conversation, including approving the PR that removed person with the 4. most commits leaving due to his employer.

I’ve been wanting to try NixOS because I really like the declarative approach, but all this drama is turning me away from the project.

  • I use Nix and NixOS and I didn't even know about this. Unless you want to become a maintainer I don't think it affects you.

I wonder what will happen if one starts developing, say, a RTOS for cruise missiles in Rust. Will this be controversial enough to destroy the Rust community?

  • Rust has had at times multiple members of its leadership team employed by Palantir and not imploded, so probably not.

    • Sheesh. And you can't write a reproducible compiler for Rust either, because every version of rustc was compiled with the previous version. What are the chances there isn't a backdoor?

      1 reply →

These documents always seem so desperate, as if the authors are fighting to defend their families.

I wonder how much of it is because they have over invested in this community versus their local ones.

It's always good to see such statements with a poisoned well from the start. I have no investment in any of this, but this looks really like the authors of this are the root issue here.

And I also agree with - what they call - the opponent, that politics should not be forced upon an open-source project. Proclaiming these are fascists, identitarians or otherwise using the Nazi stamp doesn't really help the author's position either.

I'd say this document doesn't belong on HN.

  • >Proclaiming these are fascists, identitarians or otherwise using the Nazi stamp doesn't really help the author's position either.

    The document isn't the one saying "fascists", it's criticizing the people throwing that term around.

    • I must assume that the quotes presented are the best they have on them and the word fascist comes from the author(s). It's just very involved and the author(s) look therefore very bad to be honest.

      1 reply →

People for whom gender is part of their identity are covertly hostile towards people who don't care. The people that are expelled are never those who have pronouns in their bio's. Respecting (=affirming) gender identity is part of every CoC.

Just a feeling. I have no proof.

  • While I believe one should not put their gender into their identity, the vast majority does, and is also not hostile towards those who don't care, and you can't really ignore gender-related issues and discrimination of current society right now because, while I think it ought not be, gender is a big deal and you can't ignore this.

    And, It should go without saying, respecting one's identity is the least we can do, including the gender since it's in there. Except, of course, for horrible stuff, but surely gender is no such thing.

    Actually, while I don't care about my gender, if you are not ready to respect my fellow human beings and their identity, I don't really want to have to interact with you.

    When you say "not caring", do you mean "Can't be bothered to use pronouns people wish to be referred to" or "don't mind being called with whatever pronoun?" Because the phrasing is ambiguous and two of the possible meanings are radically different things.

    • Actually they are hostile if you are non-conforming (not all but many enough that its a problem). This results in the discrimination you speak of.

      At this point im not sure if socially constructing your identity is that common outside of your teenage years.

      7 replies →

  • I agree. Having this added to CoCs is a method of trying to enforce compliance to a controversial ideology.

    Those of us who reject the idea that women and men are to be redefined in terms of "gender identity", and consider this to be a sexist belief system, either have to pretend to hold this view or be excluded from participating.

  • Bunching several concerns together and calling everyone who cares about them the 2024 equivalent of "blue-haired SJW" with some extra transphobia sprinkled in is the opposite of intellectual curiosity.

    You could've just not commented.

  • These people want to steer away projects from being involved in the military but these people are also always at war with whoever that does not agree with whatever they currently think.

    • >These people want to steer away projects from being involved in the military

      Then they don't understand what free software is.

      Not surprising considering these people are primarily activists and software is merely means of gaining personal power in another space.

I love how in these current years you can shutdown anything with just simply shouting fascism/nazis. Don’t like something? They are fascists444

  • On the same note, its hilarious that right now twitter of all the websites on earth has the best fake news detection mechanism in community notes but it's also the single website everyone wants down. I remember before Elon took over, there were daily hitjobs on facebook but just yesterday people were cheering threads having 150M MAU.

This is exactly what the opponents of CoC-culture warned about: it's a clever rhetorical smokescreen to allow leftist freaks to do power grabs like this. Words like "harm", "inclusive", "community" etc have dual-meanings, one for the clique insiders and another for the outsiders.

And lmao wow, one of the mods involved in this is also a mod at lobste.rs. No wonder the "discussion" there looked so uncanny.

  • wrt. Lobsters, check your facts.

    - Moderation actions are public on lobsters: https://lobste.rs/moderations

    - Irene wasn’t active in that discussion.

    - Some of the most upvoted comments were against the moderation clique.

    • Irene also wasn't involved in me removing srid's attempted brigading or it being the last straw for his ban.

      The author of this article doesn't seem particularly interested in checking facts, they implied some nasty stuff about my actions and motivation without bothering to read the publicly available info or contact me.

      10 replies →

  • without commenting on this specific case because TLDR: the same argument could be used in a situation where moderators allow harassment against certain groups in the name of freedom of speech, while moderating harassment against others.

    moderation is prone to power plays no matter which angle you approach it with.

There’s always Guix.

  • I'd love to try it, but they're super-hostile to running on a Mac.

    • Not hostile. We just can't do it. There is no free toolchain for Mac. So we'd have to arbitrarily cut the trust graph and graft it on top of a huge blob of a proprietary toolchain and set of system libraries. That's like building a completely new distribution. Since I'm not using macos and have no interest in committing my own money to keep paying for freedom restricting software to provide a service that would earn the label "supported" there's no way I'm going to make that effort.

      4 replies →

"far-left identitarian" - what's that even supposed to mean, https://en.wikipedia.org/wiki/Identitarian_movement is defined as "far-right ethno-nationalism".

"left-wing ethno-nationalism" is an oxymoron.

> This is a GitHub web page for the repository "rfc-evidence" owned by user "nrdxp". The page contains the file "rfc_evidences_experiences.md" which is part of the repository's master branch. The page includes various HTML, CSS, and JavaScript files that are used to render the GitHub user interface. The page provides metadata about the repository, such as the description, analytics, and social media sharing information. Overall, this appears to be a standard GitHub repository page for a project called "rfc-evidence".

Time saved: 171 minutes.

Thanks Kagi-search summarizer....

  • Did you notice your AI didn't speak at all about the content and only provided obvious info that you get at a glance when loading the page? For the tiny relevant part of the summary anyway.

    But I suspect no AI will be able to provide correct understanding of this, we are sorely lacking context here.

    Also, you guys providing LLM generated stuff in HN comment keep convincing me I should not invest time in this stuff.

    "this page provides HTML and CSS to render GitHub's UI", no joke.

    (edit: fixed IA -> AI)

    • The punchline to that was "178 minutes saved", and that the "IA" wasn't able to provide any useful information out of it.

      If you're not familiar with these tools, that is very abnormal output for one of them. Normally they'd be able to provide a reasonable summary for all of this.

      So taken in that context, the subtext is "is this worth any of our time"?

      I did link directly to the RFC in another comment, which I think is more worth people's time.

      1 reply →