LWN: Comments on "NixOS moderation team resigns"

Where moderation steps in

bauermann — Fri, 03 Oct 2025 22:56:34 +0000

> I suppose the issue being raised is that the moderation comment is visible to readers as a reply to the latest (at the time it was posted) comment in the thread. This could give the impression that the latest post was primarily responsible for the moderation comment.

If this is the issue, perhaps it can be improved with a UI change? E.g.:

Instead of the moderation comment being a regular comment in the thread that went into the weeds, the UI could instead collapse that thread, and display the moderation comment on top. The user could still be able to click through the overlay and expand the thread to see it.

Of course, a reasonable argument can be made that there are better things to do with the limited resources available to develop the website UI.

Much information is missing here, sorry to say

rbranco — Fri, 03 Oct 2025 20:48:45 +0000

> And it was apparently possible to impose this suspension in spite of the moderation team resignation. This is the sort of "essential capability" that needed to be maintained, apparently: to silence someone who objects to this sort of rhetoric, because the rhetoric was aimed at all the acceptable targets.

This is not moderation. This is a Kangaroo CoCourt.

Much information is missing here, sorry to say

mpg — Fri, 03 Oct 2025 20:33:06 +0000

> [...] a harmful tool wielded by people in specific identity groups, who are thus being described as inherently bad. [...] denigrate several identity groups

If I understand correctly, you're saying that you read Chen's message as denigrating the identity groups she names, and describing them as inherently bad?

I've read her message in full, and I'm absolutely not seeing that in her message.

To take a comparison (imperfect as all comparisons), imagine I, not a native English speaker, am having a debate with a few native speakers, and at some point I say that I feel at a disadvantage compared to them as it's harder for me to formulate my points as precisely as I would in my own native language. I don't think saying that would imply they're doing anything bad here, even less that the "identity group" of native English speakers is inherently bad. Just pointing out the language we're all using is native to them but not to me - as a statement of fact, devoid of any judgement.

I think what Chen is saying here is that "etiquette" that some may see as "universal" is really a cultural norm that is more "native" to members of certain identity groups (that she proceeds to name) than to others. I don't see any kind of judgment passed on (members of) those groups here, any more that in the situation above.

Now, where the two situations might diverge, is that I think most people would agree a common language is needed to communicate efficiently. While the overall point Chen is making is that we should dispense with the idea of a common etiquette as the basis for moderation, because it will never be as "neutral" or "universal" as we might wish it to be (and perhaps sincerely think it is), and base moderation on something else entirely (which she proceeds to describe).

I'm not saying I agree with this overall point, and am not really interested in discussing that now (even less whether I think specific actions of the moderation team were justified). I just wanted to react to the idea that her message was in any way denigrating certain groups, because I really don't think that's what she's saying here.

Just ban non-technical discussions?

zahlman — Fri, 03 Oct 2025 18:31:41 +0000

Thanks for your perspective.

> There are several online venues where I regularly discuss all of these topics, including sometimes very intense and fiery discussions, that are resolved constructively with a mutually increased understanding virtually every time. The crucial things that make it work are a basic set of shared moral values, a mutual interest in constructive discussion (as opposed to competitive 'debate'), and effective moderation to weed out people who deliberately use abusive tactics to incite conflict (mostly just relevant in the public rooms).

I once moderated for a community that I would consider very much like what you describe. It seemingly defied odds to survive for a while on a platform where a large fraction of the views expressed would otherwise have been verboten. (Generally, things that are likely outside of the Overton window of much of the developed world, but which would not rise to the level of "hate speech" in those countries — such language and rhetoric was and is definitely rejected.) Eventually it moved on to its own hosting (I had left well before that point, so I can't comment as to the reasons) and still seems to flourish — although, perhaps unsurprisingly, the "otherwise verboten" views are now dominant.

So, yes, these things are absolutely possible. I'm not even convinced that "shared moral values" are strictly necessary, as long as people are able to express those moral values clearly and distance the argument from the personal emotions it invokes.

But, the important part, I *absolutely would not want any of it to get anywhere near* any FOSS I'm involved in. It's blatantly irrelevant, and requires upholding a very particular conduct that most people with strong political opinions are simply uninterested in even attempting.

Much information is missing here, sorry to say

zahlman — Fri, 03 Oct 2025 18:10:38 +0000

There are a few things that, in my opinion, people ought to know when evaluating the moderation team's actions. Of course I introduce a bias this way; everyone is welcome to bring their own relevant observations as well.

First: from the post itself: "Measures are in place to ensure essential capabilities are maintained." There is a strong suggestion here that the "resignation" doesn't actually entail any relinquishment of power.

Second: the structure of NixOS governance is such that the SC is elected, but the moderation team is not — they appoint their successors, and appear to be accountable to no one. The SC has only attempted to persuade them to fire or hire certain members, and it seems to have been a long, drawn-out process. Certainly the SC could not act unilaterally to fix what they apparently have seen as a serious problem for a long time. And in response to this interference, it comes across that the moderation team seeks power over the SC, to establish that the elected group should instead be accountable to the appointed one.

Both from personal experience and from prior observation of other communities, this governance model is bad news. It works fine when a project's BDFL *directly* appoints people to positions of power (to moderate or lead discussion, make crucial decisions etc.) within the project, or at least has oversight of such appointments. A project that loses its BDFL really must elect everyone (and they should be pre-committed to a contingency plan established ahead of time, per the BDFL's will, lest they be endlessly mired in meta-discussion).

Third: the moderation team has a long established history of expressing a very clear political tendency — one which some subset of their users is undoubtedly very happy about, but also one which would naturally tend to exclude a lot of reasonable, well-meaning people and make them feel unwelcome. For reference, this is the same project where the founder was forced out largely over disagreement with that political tendency.

There is an example from the other day — after the publication of this article, but I think it is very instructive about the current attitudes within the NixOS community. As it happens, at the same time as this moderation team resignation, the SC is up for re-election. Shahar Or — a long-time contributor, and self-described "author of the Full Time Nix podcast" — was suspended for a month from the NixOS Discourse (https://discourse.nixos.org/u/mightyiam/summary), which is to say, until after the SC election. The apparent cause of action is Or's critique (https://nitter.net/pic/orig/media%2FG2JEt_uXUAA1FVK.png) of a comment (https://github.com/NixOS/SC-election-2025/issues/428#issu...) made by an SC candidate, Leah Amelia Chen (https://github.com/pluiedev):

> One common stated goal for moderation is to preserve "etiquette", but that word is intrinsically linked to culture, and in the context of the Global North, it depends on a set of codified social norms based on cisgender, male, White, European standards of politeness, which does not always universally apply and in fact have been frequently used to silence minorities from other cultures and backgrounds that do not construe expected conduct in the same manner.

Which is to say: the idea of expecting people to follow "the practices and forms prescribed by social convention or by authority" (American Heritage Dictionary definition) — i.e., to follow *a Code of Conduct* along with generally trying to fit in with the community — is somehow a harmful tool wielded by people in specific identity groups, who are thus being described as inherently bad. The objection is apparently to a word, which is used as an excuse to denigrate several identity groups, even though there is nothing wrong with the meaning of the word in context. Chen proposed no concrete examples of how people from other cultures would have difficulties with expectations set by supposed oppressors on a web forum, nor gave any coherent reason why "preserving etiquette" would entail upholding oppressive standards.

(Quite frankly, while I can imagine how sex, ethnicity etc. influence socialization and thus social norms, I cannot fathom that these lead to meaningful statistical differences in what is seen as "polite" or not *while communicating in English text over the Internet*.)

Or, apparently, proposed in a chat room that this was bigoted. Forum moderators responded by suspending Or, instead, for "bigotry" (https://nitter.net/pic/orig/media%2FG2JEt_uXQAADSCQ.jpg). While Or's specific wording didn't make the case very well, I can understand fairly easily the argument that Chen's claim tended towards bigotry. But I cannot, on the other hand, fathom any argument that makes Or's response bigoted. It quite simply did not assert anything pejorative about any identity group. Rather, it sought to defend identity groups.

And it was apparently possible to impose this suspension in spite of the moderation team resignation. This is the sort of "essential capability" that needed to be maintained, apparently: to silence someone who objects to this sort of rhetoric, because the rhetoric was aimed at all the acceptable targets.

I'm just sad

zahlman — Fri, 03 Oct 2025 17:41:04 +0000

> As a European NixOS user, this makes me worry that running the OS is quickly becoming a liability.

Personally, I can't see a reason why disagreements between other users in "the community" (i.e. the subset of other users who take an interest in posting publicly about their use of the software) should cause a problem for you, no matter what side you take (or whether you care at all about the topic of disagreement). The worst that happens, practically speaking, is that important contributors are sidetracked from development. It's not as if the project will attempt to insert malware targeting people based on their beliefs (as gross a violation of FOSS principles as I can imagine, FWIW) or anything like that. Forks of the project would also struggle to see more active development or maintenance than the original.

Unless, of course, you are dependent on that "community" for technical support, and can no longer feel welcome or heard in its forums....

Weaponization of negative descriptions

zahlman — Fri, 03 Oct 2025 17:34:09 +0000

Well said.

It is, in fact, impolite to call people "bigots". First off, because it makes a serious accusation that commonly brings real-world consequences, while not necessarily being supported by the facts. To say that someone is being "described" as a bigot begs the question: it assumes that the target actually expressed bigotry. Second, because bigotry is a characteristic of conduct, not of identity. There is a large difference, for example, between calling someone a "liar" and claiming a statement to be "a lie" (which in turn is a stronger claim than simply saying that it is "untrue").

Of course it is a problem for people to express bigotry. (Here I understand that we use the term in the general sense of expressing unjustified prejudice, particularly towards groups of people defined by identity characteristics, rather than in the dictionary sense of general narrow-mindedness or obstinacy.) However, this must be understood to be true of all forms of bigotry, not just those that a moderation team finds sympathetic.

The kinds of "call-outs" that people defend in the name of holding back supposed "bigots" is commonly demeaning, grossly inaccurate and frankly beyond the pale. It is frankly wrong to suppose that "framing" this as impolite is "a tool that people sometimes use to position such behavior as *not* unacceptable". This is an entirely unsubstantiated claim. The plain reading of such accusations of impoliteness is quite literal and transparent. People should simply not throw around such language in public, and moderation teams had better be able to make a fully reasoned argument that would be convincing to any reasonable person, not just to people with aligned political views.

Where moderation steps in

paulj — Fri, 03 Oct 2025 10:02:51 +0000

I'm from a country with a PR-STV political system, where our politicians generally have to do their best to appeal not just to some "base", but also - at a minimum - still appear reasonable to the "bases" of at least a few other political parties. This system selects for politicians who can maintain broad-appeal in at least a few political dimensions. They still get it wrong sometimes, but when they do the system also allows the electorate to correct (cause you can /always/ give your #1 vote to whatever minority politician who is on the right side of some issue; without worrying that you will then inadvertently allow a disfavoured politician in; cause you also have #2, #3, etc. preference votes).

From my vantage point, no one in that cartoon looks good, at all, and it's a damning indictment of the political system that fostered it.

Where moderation steps in

paulj — Fri, 03 Oct 2025 09:55:20 +0000

Maybe.

OTOH, it's a fairly small and long-standing community here in the LWN comment threads. I suspect the vast majority of us are well aware of what a "Gentlemen, ladies, time to stop." style LWN-mod-comment means and does not mean.

And the breadth of injury that may result from a few misunderstanding appears to be very very limited and not worrying about.

Finally, LWN's resources are very limited - however much we may think there is some better system of moderation that LWN could apply, there is also the question of whether working towards that is in any way worth the effort and cost. And the marginal benefits may be tiny compared to the high cost in precious LWN editor time.

The system currently largely works. If the LWN editors see easy ways to improve it, sure, great. However, I see no great pressure for them to spend any significant amount of their precious time on that either. I'd rather see them doing the stuff we value them for - producing the excellent content of the site.

I'm just sad

koverstreet — Fri, 03 Oct 2025 00:36:46 +0000

It's quite indisputable that bigots (and others of even more dangerous persuasions; i.e. fascists) do in fact exist in real life; that the words may be used as "name calling" doesn't mean that they're no longer subjects of serious conversation.

We might wish to be free from hearing "impolite" discourse, but that can lead in very dangerous directions when taken too far. Martin Luther King Jr. had a lot to write on the subject awhile back.

Unusual for this stuff to come up in tech, but I think there's enough going on in the NixOS world re: Arduril that looks perhaps not entirely proper that we should be careful not to dismiss out of hand.

Stop here

prokoudine — Fri, 03 Oct 2025 00:11:48 +0000

I did not quite expect to see this much of a snowball effect following my remark. Either way, I completely understand the kind of stress you have to deal with and trust that you exercise your best judgment.

I'm just sad

prokoudine — Fri, 03 Oct 2025 00:02:00 +0000

> "bigot", for instance, is a descriptive term. It's not "impolite" to describe someone that way; it's a problem for people to act bigoted.

Let's not muck about. There's a whole range of words, "bigot" included, that are casually used today for name-calling, well outside their original meaning.

This little game of calling the other party bigoted or nazi or other interesting words to establish one's own moral superiority has become quite tiresome. There's a saying where I come from: in an asylum, whoever is first to grab a white coat gets to be the doctor.

I wish it didn't feel like we are all in an asylum. Alas...

Stop here

paulj — Thu, 02 Oct 2025 10:36:13 +0000

> Us vs Them dialectic that only benefits those in power and where we all lose.

This is the essence of it. The elites - at whatever level of combination of conscious planning and emergent behaviour arising out of the class incentives - use all kinds of delineated rhetoric to divide and conquer the masses, on lines of ethnicity, religion, culture. But never, never along the lines of class.

And we lap it up. And on occasion, when the curtain briefly splits for a second, and we see how the elites supposedly representing different factions - "Dem" and "Republican", "Labour" and "Tory", "Fianna Fail" and "Fianna Gael", "Al Qaeda" and "CIA" - hobnob together, laughing and slapping each other's backs, few seem to see it.

Too distracted by fighting other little people about whatever little line, whatever flag, they've been cultivated to obsess over, to notice.

Stop here

rbranco — Thu, 02 Oct 2025 09:49:24 +0000

It's a pity we missed the chance for people to get educated on the Paradox of Tolerance beyond their wikipedic knowledge. We got a meta-discussion instead justifying the use of the term "bigot" to make sure we stick to the Us vs Them dialectic that only benefits those in power and where we all lose.

Where moderation steps in

raven667 — Thu, 02 Oct 2025 04:26:39 +0000

Once you identify the pattern of behavior, of someone skating on the line like that, you can recognize it as the toxic behavior it is and bounce the person out of the community. As the owner of the platform you don't need to put up with people who troll other users that way, and _that_ keeps the temperature down, conversation civil.

Where moderation steps in

amacater — Wed, 01 Oct 2025 22:27:17 +0000

Some people are Olympic class figure skaters - *just* over the line with something then quickly pulling back, then *just* over the line as a pattern of behaviour. That makes moderation and community building hard.
Many communities exist on a barely articulated common understanding that lasts until someone questions it.

Where moderation steps in

madscientist — Wed, 01 Oct 2025 18:05:38 +0000

I suppose the issue being raised is that the moderation comment is visible to readers as a reply to the latest (at the time it was posted) comment in the thread. This could give the impression that the latest post was primarily responsible for the moderation comment.

I don't believe that this is how the LWN mods intend for it to be perceived: they are talking about the entirety of the thread not specifically about the post they're replying to.

I suppose the people concerned about this perception would prefer that the LWN mods should back-track the thread to find the post that appeared to be primarily accountable for the problematic digression, and reply to that rather than the "latest post".

I'm not sure how feasible that is; it is assigning blame much more directly and thus, could cause more arguments than it prevents. Is it enough for us all to understand that the reply is not intended to indict the direct parent post? Or maybe there needs to be some standard disclaimer language in moderation comments?

Where moderation steps in

ferringb — Wed, 01 Oct 2025 18:02:02 +0000

> One common problem in moderation in general is that many people are very good at "calmly" inciting in a fashion that garners a well-needed negative response and/or call-out.

You pretty much nailed it on the head, in regards to the most dangerous version of behavior.

It's absolutely the hardest to keep in line w/ CoC- it never crosses the line blatantly, but the responders all get nailed as problematic or crossing the line in the sand. Eventually you get additions to CoC and moderation to try and address the lawyering, etc. Said additions to the rules just makes things worse, and harder to bring back to the spirit of the communities original intent.

This sort of thing is what I now watch for in communities; if it's left unchecked I just find somewhere else to go, assuming I have any choice in the matter. If I have to consume their code, sure, but even bug reporting is something I'd prefer *not* to do since I just don't want that crap in my life.

I'm just sad

jzb — Wed, 01 Oct 2025 16:50:47 +0000

Daroc wrote up an article about Nix alternatives / spinoffs last July, here: "Nix alternatives and spinoffs".

Not sure how they've fared in the interim, it might be time to revisit them to see how they've developed.

I'm just sad

Baughn — Wed, 01 Oct 2025 16:40:35 +0000

As a European NixOS user, this makes me worry that running the OS is quickly becoming a liability.

What’s your recommendation for an alternate OS?

Where moderation steps in

paulj — Wed, 01 Oct 2025 14:36:06 +0000

Hmm.. if the LWN mods have come into a sub-thread in primary school mode then it is very safe to assume that the sub-thread has gone long past any kind of constructive criticism.

I'm not sure what you're advocating for, that the LWN mods then start to "constructively criticise" people in the thread? That may well just make things worse - it will make LWN worse. LWN is not mired in the US-culture-war caustic shite that other parts of the media/net are, to any significant extent. LWN commenters seem largely sensible and able to take hints to shut-down chains that have stopped being useful. Advocating for moderation solutions here that assume it is may be counter-productive, and cause it to MOVE TOWARDS that undesirable state of toxicity, as much as anything else.

Or ??

Where moderation steps in

josh — Wed, 01 Oct 2025 11:51:34 +0000

+1 to this.

Where moderation steps in

pizza — Wed, 01 Oct 2025 11:50:37 +0000

> That comes across as the problem being exclusively the heat, not the toxic statement, or other toxic comments like people attacking the call-out for being "impolite" (where the implied "polite" would be "tolerant of intolerant/toxic people").

This highly-relevant comic came across my feed just yesterday:

https://leftycartoons.com/2025/09/26/doin-discourse-with-...

Where moderation steps in

hkario — Wed, 01 Oct 2025 11:47:51 +0000

> allows everyone to walk away.

that only upholds the status-quo and is not conducive to constructive criticism

and yes, calling out toxic behaviour is constructive, even if it's not "polite"

I'm just sad

josh — Wed, 01 Oct 2025 11:45:11 +0000

There are a whole raft of meta-level problems that arise.

For clarity, it is indeed the case that it is important to be accurate, and not use a term that doesn't apply.

But also, people who wish to be bigoted (or wish to not care whether they are or not), and who also don't want to experience any negative consequences themselves, have a vested interest in applying a wide variety of tactics: trying to lower or abolish standards, attacking callouts as "impolite", trying to act as if the person calling them out is bringing politics into things but the person being bigoted is not, trying to position themselves as the "default" and equating "politics" with "non-default politics", and a hundred other things.

Sometimes these are intentional tactics, sometimes they're picked up by osmosis, and sometimes they're just the natural consequence of trying to defend and normalize one's behavior by any available means rather than reflecting on it or changing it. In general, people who are doing something wrong and don't want it to be wrong have many different tactics, and the better they are at it, the closer it can get to attacking epistemology and the connections between words and reality.

In short: it's important to be right when calling out bad behavior. It's also important to recognize the common rhetorical pattern of not refuting an accusation of bad behavior but instead attacking and impugning and narrowing the very concepts and words that allow calling out bad behavior.

Where moderation steps in

josh — Wed, 01 Oct 2025 11:26:49 +0000

One common problem in moderation in general is that many people are very good at "calmly" inciting in a fashion that garners a well-needed negative response and/or call-out.

The net result can often be:

A) "Dispassionate" but toxic statement

B) Well-deserved call-out

Mod) This is getting heated, please stop

That comes across as the problem being exclusively the heat, not the toxic statement, or other toxic comments like people attacking the call-out for being "impolite" (where the implied "polite" would be "tolerant of intolerant/toxic people").

This is one of many patterns where it's important to flag the underlying problem, and not doing do lets people get away with trolling and incitement, repeatedly.

Weaponization of negative descriptions

farnz — Wed, 01 Oct 2025 11:11:43 +0000

That rhetorical technique works for most negative descriptions; if you describe someone as "X" (where X is something negative, like bigoted, foolish, stupid, dumb, arrogant, whatever), then, when someone pushes back on that, you can summarily dismiss the pushback on the grounds of "if you defend being X, then you are part of the problem, because being X is indefensible by definition".

The trick to it is that you skip over whether the accused is X, assuming that it's true and therefore whether it's OK to be X is what's being questioned, not whether the accused is X.

Moderating a distributed community requires the moderators to spot the leap from "I'm not bad" to "defending being bad is not possible", and call it out - if someone's been accused of being bigoted, but claims they're not, the question is not whether bigotry is defensible or not, but rather whether the accused's words and deeds rise to the level of bigotry.

Where moderation steps in

paulj — Wed, 01 Oct 2025 10:53:10 +0000

I think the primary school teacher method works well. No one is singled out (mostly), and the lack of specific blame allows everyone to walk away.

I'm just sad

intelfx — Wed, 01 Oct 2025 10:34:48 +0000

There is also a problem where accusations of "bigotry" are used simply to attack anything you disagree with. And then, when someone pushes back on your accusation, you can summarily dismiss that pushback on the grounds of "if you defend bigotry, then you are part of the problem, because bigotry is indefensible by definition" (i.e. roughly what you are saying here).

It is a very convenient political weapon.

Where moderation steps in

corbet — Wed, 01 Oct 2025 07:55:46 +0000

Moderation is not something anybody here at LWN wants to be doing; perhaps it is not surprising that we do not do it as well as some would like. When a thread is going off the rails, our first objective is to get it to stop; that seems rather more important than determining who should be decreed guilty for starting it. We also lack a ready strike team ready to react within milliseconds of the first bad post, sorry; we can only react after we see a problem.

When we do take action against a specific user, such as putting them on permanent moderation, we do not make public proclamations about it. You just don't see unpleasant stuff from that person anymore and never notice.

We are doing the best we can; if you see us as an "awful primary school teacher" I can only apologize. Time for recess!

Where moderation steps in

Wol — Tue, 30 Sep 2025 22:52:52 +0000

As someone who gets flagged a lot, I do feel sometimes that it's just clamping down on "off topic". That said, if things drift off topic that can let in bad behaviour.

In the old days when it was just Jon, there wasn't much calling out of off-topic (to be honest there wasn't much calling out), but LWN has grown and I think there's also more people looking for places they can stir up trouble (I gather there's quite a lot of Russian "espionage/subversion" going on), so I can understand the team clamping down.

But I do feel it doesn't seem there's much attempt to clamp down on the real bad boys - although the problem there is one person's bad boy is another person's "agent provocateur" is another person's "just asking questions".

I think PJ's rule of "if I wouldn't have it in my living room, I won't have it on my site" was a good one, but it relies on getting to know your posters ...

Cheers,
Wol

I'm just sad

josh — Tue, 30 Sep 2025 22:52:36 +0000

Thank you, that's a good way of putting it.

"bigot", for instance, is a descriptive term. It's not "impolite" to describe someone that way; it's a problem for people to act bigoted. It is not "polite" to refrain from calling out bad behavior, and to state unequivocally that it isn't acceptable behavior. Attempting to frame that as "impolite" is a tool that people sometimes use to position such behavior as *not* unacceptable.

Where moderation steps in

josh — Tue, 30 Sep 2025 22:32:49 +0000

I would in general say that it often seems like there's a pattern of calling attention to the comment that calls someone out for bad behavior (e.g. flagging the point where people start arguing) rather than calling out the incitement of bad behavior.

Or to put it another way, LWN moderation sometimes comes across like an awful primary school teacher saying "I don't care who started it" and acting like everyone involved was *equally wrong*.

Just ban non-technical discussions?

joepie91 — Tue, 30 Sep 2025 21:21:20 +0000

> Discussing politics, religion, money and power in general does not work on the Internet. It hardly works elsewhere but there's basically zero chance online. We've known that since the Internet was invented. Did nixOS people not get that memo yet? What am I missing?

There are several online venues where I regularly discuss all of these topics, including sometimes very intense and fiery discussions, that are resolved constructively with a mutually increased understanding virtually every time. The crucial things that make it work are a basic set of shared moral values, a mutual interest in constructive discussion (as opposed to competitive 'debate'), and effective moderation to weed out people who deliberately use abusive tactics to incite conflict (mostly just relevant in the public rooms).

The idea that "discussing politics online is hopeless" is a belief that IME is mostly trotted out when people try to avoid solving the hard problems by declaring them non-problems or unsolvable. But it's never actually been *true*, and that becomes apparent pretty quickly once you start doing serious community moderation and learn to detect the patterns of conflict. And if I'm being honest, with 'internet debate culture' now leaking into mainstream politics, it's never been clearer that none of this was ever really specific to the internet.

(We could have a long conversation about why things failed specifically in the context of NixOS, but it would just be a re-run of the last 5 years of debates and I don't think anyone would gain anything from it if "it can't work anyway" is the starting point for the conversation.)

Dazed and confused

joepie91 — Tue, 30 Sep 2025 21:21:17 +0000

> Nix is far from the first community to have had moderation issues, but given the topic, it's not always clear how the wider community can learn something from these incidents. There is so much that cannot be aired publicly that we only see fragments of one side of the story, but not a very clear picture of why the internal pressure kept building until it had to explode very publicly.

Speaking here as someone who has been involved in trying to get the governance issues resolved for the past 5 or so years (it's been ongoing for much longer than it might seem publicly), the fundamental problem at the root of all this is that project governance failed to take warnings and complaints from the community seriously. The default response to any and all reports of issues with governance/communication/coordination/etc. within the project was to either ignore the issue or trivialize it, with the tone of "I'm sure it'll all work out".

Notably this was originally just a problem with things like documentation, accessibility for new contributors, glacially slow reviews, and so on. It only started revolving around moderation and, eventually, the involvement of arms dealers much later, once the popularity of the project spiked and a large influx of new people started. But because the root problem of laissez-faire governance was never addressed, the project never had any infrastructure in place to deal with these much more complicated problems once they became relevant.

There are a lot of details around specific bad decisions by specific people, specific problematic personalities, and so on; but ultimately those details don't really matter, the refusal of (the various forms of) governance to make decisions in response to problems and commit to them has always been at the root of all of these issues. And eventually all the built-up frustrations around "nothing is happening, everything feels immovable" explode publicly. To a point where the specific frustrations and policy views almost don't matter anymore.

Just ban non-technical discussions?

Wol — Tue, 30 Sep 2025 19:45:05 +0000

> Why can nixOS not just completely censor all discussions unrelated to nixOS?

Because I get the impression it's NOT unrelated to nixOS. What they're shutting down is things that are *controversial*. You know - those awkward things like governance. Internal Politics. Etc etc.

(Personally, I feel shutting down "unrelated" stuff can be a very bad thing. I've seen too many discussion fora become "business only" places, and very rapidly the only thing left is tumbleweed ... It's hard work to keep things under control if you allow wider discussion - as our esteemed editors know - but it brings in engagement. The problem is that if you shut down "unrelated" stuff you end up with an echo chamber, if you allow unregulated stuff things can turn nasty very quick. And if you're really "unlucky" - Facebook anyone? - you end up with vicious echo chambers ...)

Cheers,
Wol

Just ban non-technical discussions?

GNUtoo — Tue, 30 Sep 2025 19:11:54 +0000

> Why can nixOS not just completely censor all discussions unrelated to nixOS?

The problem is also that many of the things you cite can be related to a given FLOSS project. For instance projects like Guix and Debian sometime have to discuss "extremely political topics" because they do impact Guix or Debian.

For instance what software not to package will bring "political discussions", and sometimes people can decide not to package a given software just to avoid these discussions. How to make contributors welcome is also a very "political" discussion. Also not all distributions have the same policies with regard to laws and jurisdictions. For instance libdvdcss is legal in France, it might not be everywhere. The inclusion or not of nonfree software is also political. And if you accept that "political" has a wide definition, then everything is political.

Though projects like Debian have a long history of having discussions and so they also have insights on how to get things done while limiting infighting between people that are basically on the same side.

Though not everybody agree on political views and/or are on the same side and this is how things are. There are FLOSS projects with completely different goals. For instance some enable privacy (the Tor project) while other probably do the opposite (tracking). And there might actually be more than one field where there are cat and mice games between FLOSS projects and/or people or organizations that have completely opposite goals, though I don't pretend to know every FLOSS project or communities out there.

Just ban non-technical discussions?

daroc — Tue, 30 Sep 2025 18:44:37 +0000

I think, as with any difficult social problem, it's easy to dismiss the situation as easily resolved from the outside. I don't have a full picture either, but from my research it seems like many of the problems boiling over this time are the same problems that I wrote about in May of last year, which had, at that time, already been ongoing for several years. Anything dealing with people and community organization is tough.

Not least of all, at one point the community did try to make a decision about sponsorship in the way you suggest, and it nearly didn't stick. The NixCon organizers removed Anduril as a sponsor in 2023 due to feedback from the Nix community — and then Anduril attempted to become a sponsor again the next year. The resulting discussion involved a lot of debate about how, exactly, sponsors should be selected.

Just ban non-technical discussions?

marcH — Tue, 30 Sep 2025 18:30:27 +0000

Why can nixOS not just completely censor all discussions unrelated to nixOS? This should benefit everybody. Does not matter whether they are offensive or not, which is always subjective: just censor them all. Would that not be enough? This happens naturally in most open-source projects! Without censoring or even asking. Unsurprisingly, most people don't like to mix work and politics. It's not like there is a shortage of places to whine on the Internet, just go and find a more appropriate one. That's what most people do; so why not nixOS?

Rare exception: when a company wants to be a sponsor. Exception restricted in space and time. Decide one way or the other and move on. If the decision is unacceptable for you, leave the project or fork. Either way, move on.

Discussing politics, religion, money and power in general does not work on the Internet. It hardly works elsewhere but there's basically zero chance online. We've known that since the Internet was invented. Did nixOS people not get that memo yet? What am I missing?

Stop here

daroc — Tue, 30 Sep 2025 17:54:17 +0000

Unfortunately, there are only four of us. So when there are several comments in quick succession, we may only get to the thread afterward.

... but I agree that we could probably do a better job of indicating _why_ a particular subthread is asked to stop.