Should organizations relating to the defense sector being able to sponsor NixOS?

[aou]

Now that the thread has a finite lifetime, I feel comfortable sharing my read and suggestion.

The Foundation and Board are not set up in order to dictate the direction of the project, but to empower the community to make decisions on the direction of the project, while handling administrative, legal, and financial issues in that role. (I’m taking this from my read of this .) However, there is an unavoidable interface, sometimes blurred, between empowerment and leadership; for example, organizing a collective direction of a diverse community necessarily involves imposing some limits and guidelines in order to reach an objective, such as creating deadlines, processes, and policies.

This thread has provided a lot of information and context on what the posters care about and why, which is useful and important. However, it doesn’t necessarily point to how the board / foundation should use that information to create policy, because that needs to be a judgment call that will probably not satisfy everyone. Which is ok, and unavoidable.

My suggestion to the board is probably unnecessary, I believe they are cognizant of their role and will be going about it as SCALE winds down and maybe they recover some energy. I would say, draft a policy and put it out there, see how the community responds. (Like I said, unnecessary, but here for completeness.)

My suggestion to the community is to respect the board and foundation’s roles, and what their roles entail, and do not entail. It may seem through silence or lack of full-throated support of a deeply-held belief, that they are in opposition to said belief – that is the way deeply held beliefs usually feel. But I believe they are acting as the role requires, to empower the community, a diverse community of deeply held beliefs, to move forward.

[delroth]

In the end this is how I view this situation re: Anduril sponsorship: even if you’re personally ethically ok with this sponsorship, you can still hold the stance that Anduril shouldn’t be a sponsor because you empathize with the community members that aren’t ok with it. Not being sponsored by Anduril mostly doesn’t impact you if you’d be ok with that sponsorship, at least not at a $5K one-time amount. Heck, some individual community members have donated more money than this to the foundation out of their own pockets!

People have nitpicked the numbers of how many community members have supported the open letter against MIC sponsorship. For the same reason, I think that the exact numbers don’t matter. This isn’t a democratic majority vote. The people signing the letter are more impacted by the decision than the people who don’t care. One “sure, why not” doesn’t equate a “no, absolutely not” when it comes to finding consensus. In the end, what matters with the letter and its signatories is to show that a non-trivial portion of contributors are uncomfortable with the decision. The current numbers unmistakably show this.

And of course, there is a limit to be found somewhere: community members hold very different ethical and moral opinions, and that’s generally good. But we don’t have to define the exact limit if we listen to people’s complaints when they happen. It should be pretty obvious for example that even though people on this thread have said they don’t like Google, there was no sign of a large outrage against Google like there was against Anduril when both sponsored NixCon EU 2023. If you think there’s a slippery slope here: please consider that your fellow community members are reasonable people and are able to live with minor ethical misalignments. For many, Google sponsoring Nix is fine, or is a small misalignment they’re OK with. That’s not true of an Anduril sponsorship. Let’s figure out how to deal with the fine details of the grey area when we’ve already established that we can deal with the stuff that’s clearly outside said grey area.

[artemist]

I wasn’t originally planning to post, but I’ve been regularly seeing people expressing on social media and in chats that they will not use nix, will stop using nix, or will stop contributing to nix due to military contractor sponsorships.

We do have to draw “the line” somewhere, and deciding not to reject sponsorships on moral or ethical grounds does not sidestep the problem, it just places the line somewhere that is not acceptable to a non-trivial portion of the community.

There’s no need to rigorously define a policy and make every choice now. The question here is just “Should Anduril be allowed to sponsor and advertise at NixCon”. We can figure out any other questions as we come to them.

[billksun]

Seeing the community having such strong reactions, I have mixed feeling about this. As with most things in this world, I believe black and white issues are few and far between.

While Anduril is a startup and currently is razor focused. Many companies in the defense industry (in the U.S. at least) have a rather wide range of products and services. That is to say they also contribute directly to our livelihoods in addition to providing life-taking (or protecting) technologies.

DuPont started off as the largest gunpowder supplier to the U.S. Military, but now many of it’s products are used in everyday materials, and Kevlar, is one of the world’s most famous life-protecting material. Boeing opened up commercial flight to the everyday people. Raytheon famously discovered and commercialized the microwave oven. The U.S. Department of Defense (which I hold joint-responsibility of all U.S. military related casualties) sponsored and contributed to the development of the Internet, the very fabric of our modern society and what enables us to have this discourse online. The Mother of All Demos, which shaped the computer industry for the next 50 years and even now, was also sponsored by the U.S. Department of Defense.

And lets not forget that the first Turing-complete, electronic computer was created to calculate ballistic trajectories for the U.S. Army. Many cryptographic algorithms that we now use to secure our privacy and freedom also saw their first wide-spread application in military use.

Many of these technologies have such high barrier of entry that only the lucrative defense contracts and government sponsorships can get it off the ground. And there’s no denying that war technologies frequently have much deeper and long-running effects in advancing the technologies we use in our daily lives.

I guess what I am trying to point out is, the military and technology sector have a symbiotic, if contentious relationship. And I think it is a bit naive to use such broad stroke policy as to bar any and all defense sector organizations from sponsoring or participating in our community.

Perhaps we should examine this from the lens of societal reputation, and prohibit organizations with bad reputations to sponsor us (military or not). But this itself is a can of worms, as reputation can change with time and circumstances, and an organization can have a superposition of both good and bad reputation resulting from differing view points and core values from different groups.

Perhaps we can deny sponsorship from “controversial” organizations (as I think, in general, that’s what could drag the Nix community’s reputation down), until said organization has quelled the controversy and has showed good faith in restoring their status as well-intentioned world citizens.

Maybe moving in this direction can lead us to a fairer policy that doesn’t resulting in discriminating against a certain group of organizations?

[oxij]

I just want to point out that

1. Any discussions of aye-nay numbers for the support of anti-MIC petition w.r.t. contributions to Nixpkgs is ridiculous when most Nixpkgs contributors have no idea this is happening. I learned about Anduril sponsorship from a thread on Mastodon, not from any “official” NixOS resource. Most Nixpkgs contributors will learn that this thing happened after its gets merged into master after they see Anduril logo on a slide in a video from NixCon.

2. The whole framing of “only 7% (29%, whatever) of contributors are against it so it does not matter” is ridiculous. Even if those numbers were true (which I find to be unlikely), I would think things like this should be opt-in, not opt-out. How many Nixpkgs contributors would actually vote for including Anduril if they were explicitly asked? What if the voting was done via https://cryptopoll.org/ (which uses LSAG code borrowed from Monero for anonymous voting) or a similar tool so that you could express your opinion honestly, without your vote influencing your future employment prospects? (Which, note, works both ways. I find it surprising nobody mentioned this issue here before.)

3. Now, back to point 1, why would all those Nixpkgs contributors not learn about this issue from an official NixOS resource? Because the community has none!

I usually check recently merged non-r-ryantm PRs and updates to release notes to get my Nixpkgs news.

Where do I go for relevant community news? NixOS Discourse? This thread is another good demonstration how heavy-handed moderation kills any productive discussion here.

Remember the aptly named “Censorship Issues” thread here where people complained about that in 2018 (https://discourse.nixos.org/t/censorship-issues/1218)? Ah, but it’s 404 now, must have never existed in the first place, right?

Is it a good time for the Nix community to adopt GitHub - nostr-protocol/nostr: a truly censorship-resistant alternative to Twitter that has a chance of working protocol yet? It’s a Social Networking Protocol that I, having read all of its specs and played around with it, dislike the least:

• it’s completely egalitarian, no federation;
• you can easily archive and later re-post Nostr messages to unrelated relays, and it will actually work and keep certified authorship, since everything is properly cryptographically signed;
• so, if a relay starts censoring somebody, including by deleting their older posts, you can just re-post other people’s stuff to other relays without involving the authors of the messages at all; liked something, but the moderators tried to censor it? they get no say! other users will not even notice it happened!
• it’s much less awful than ActivityPub protocol-wise;
• at least one of the clients GitHub - unclebob/more-speech: A Nostr browser in Clojure. has web-of-trust moderation, nothing as powerful as what Freenet Messaging System has, but Nostr is not Freenet, you need to subscribe to stuff first for it to start being shoved at you, so it needs not be as good as what FMS has;
• Edward Snowden is a Nostr user, famously.

So, to summarize, my predictions: this thread will gets locked, the issue will persist, most Nixpkgs contributors will learn about this stuff from NixCon videos, another chunk of good contributors will start contributing less or abandon Nixpkgs altogether because NixOS foundation did another thing nobody asked for.

Personally, I’m working on Own Data Privateer · GitHub currently, but I’m planning to un-hiatus and publish SLNOS ([Nix-dev] The Church of Suckless NixOS is looking for followers) eventually. I’m aware I suck at sufficiently bland and unoffensive communication in general, and so I suck community management in particular, obviously, but if you agree with SLNOS’s Suckless mission and you are interested in building an anti-fragile community built on top of anonymous shit-posting with web-of-trust moderation and anonymous voting using the above-mentioned tools, drop me a message (https://oxij.org/).

I imagine, we’ll actually need to start with extending OpenPGP to include LSAGs and/or blind signatures (how does one go about doing that properly? RFC?), then make a NIP to gossip OpenPGP keychains over Nostr (because one Nostr key is not enough for us), then build tooling to do LSAG/blind signature voting over Nostr (our use case it might be slightly harder than what https://cryptopoll.org/ does, so blind signatures might be a better solution, maybe), then write a NIP and build tooling for gossiping git commits and do general software forging over Nostr, … In short, doing it properly is hard, and it will probably need funding.

But will this message get censored, I wonder?

[kiara]

@oxij
this idea might warrant a separate thread. i don’t understand enough to judge the idea myself, but imagine the idea wouldn’t be addressed before this threads gets locked.

[arijoon]

As a somewhat new member I normally don’t voice an opinion on such matters but will make an exception in this case. I would like to see nix’s mass adoption and not remain a hobbyist tool so we can more easily use it professionally and recommend them to companies. For this the foundation requires funding to grow (most major tools have strong backing), especially from those companies that use nix (they are enjoying the benefits of free labour, why not give something in return)

Additionally I see a sponsorship as the company endorsing nix and not vice versa. Getting an endorsement that this tool is used in literal military grade products is a testament to its reliability. From a tech standpoint, this endorsement is positive. E.g. if I see “Nix is used in F-35”, it has the wow factor, it must be really good then.

The cons are of course from the potential damaging association due to public perception. I think it should be judged on a case by case basis depending on the impact and reach the sponsorship can bring and not a blanket defence contractor rule. In the case of Anduril I don’t know much to judge, but if BAE Systems offered a sponsorship I would be in favour as I believe it would be a good endorsement