[e1ven]

I had written a toy social network a few years back the past which had these features - It worked very much like an encrypted version of usenet.

This is probably my bias as an engineer showing, but the technology doesn't seem like the hard part-

I always understood that having an resilient network means people will use it to post some bad things, but I don't know if I really internalized the scope of that until later.

I had originally envisioned it might be useful in oppressive countries, where people needed a way to communicate - Recent events have shown how dangerous that can feel when you're in the midst of people who feel like that describes them.

As another HN post pointed out, there are two natural audiences for such networks - Idealists, and those who can't get away with stuff on other networks.. And the second is going to be far more common. That will influence the culture, and help to drive other "good" people away from the service, amplifying the effect.

Even if you have user-selectable moderators (Which I had, similar to the request the author makes), without a huge war-chest to hire a large team of default moderators, you'll never be able to keep up. The default experience for the average user will be terrible.

Over and over, I ran into issues like that - It's relatively easy to built the technological network, but managing the social network aspect is an unsolved problem.

[at_a_remove]

If you have an old-ish head unit in your car, it may receive an RDS (Radio Data System) feed that could tell you what station it is or what song is playing. However, many stations around here are using it to advertise Club Fitness and Golden Oak lending. If you cast further back into the mists of time, anyone could send you a webcam or chatroom invite ... this was naturally exploited by spammers approximately ten milliseconds after its invention.

I have since formulated the concept that any communications channel, any at all, where it does not cost to transmit per message will eventually be colonized by the advertising fungal organism. Even low-cost messaging can be colonized, but the lower the cost, the faster it comes.

Similarly, like FreeNet, any communications channel that can be used to post Things You Do Not Like will be used to do so. And that once you implement some kind of wide-scale filter against that, absolutely nothing can be done to stop someone from attempting to take over, to add and subtract to that filter, for their own purposes and their own ideology.

I have no solutions for this.

[nelsondev]

The author (OP) proposes community generated and shared block lists.

I’m imagining an opt in block list, where with enough downvotes, if you are a member of the block list, the content is hidden.

If the block list starts blocking content you want, you can fork it, keep the parts you want, remove the parts you don’t.

[new_guy]

This is a genius idea, I'm adding it to my sites today!

Everyone knows one another, so if one person blocks a bad actor, it makes sense that everyone else will block them sooner or later.

[intrasight]

Made me laugh and weep ;)

[grahamburger]

IMO this is why federation is an important aspect of decentralized networks, and is commonly listed as a reason for use Mastodon / the fediverse. Each instance can set their own moderation policies and decide what other instances they want to federate with. Notably mastodon.social and the instances related to it haven't become cess pools of hate speech, because they do have strong moderation policies, but for users who want to post that stuff there are other instances they can find.

[convolvatron]

more importantly, moderation is an overlay. instead of worrying about what and what isn't acceptable speech, let subcommunities form with their own policies and they can curate their own worldview

that doesn't at all address bubbleism, but trying to decide which set of statements is 'ok' for everyone seems like a lost cause

[breuleux]

It could still address bubbles in a small way, by making them more accessible: if you can switch the overlay easily, you can get a peek at what other people see and understand their point of view a little better. You can also see what your favorite overlay is censoring and decide whether you're okay with that.

Also, one avenue to radicalization is a feeling that your views are being censored. If you get to choose your bubble, that argument is undermined, so extreme bubbles might have a harder time growing.

[prox]

Bubbles by themselves aren’t necessarily the problem. I could live in art related bubble with fellow artists, and there is no problem there.

Insulated bubbles are the problem I think, these echo chambers are a problem, especially unmoderated or those thriving by hate, aggrevation and/or exclusion.

[commandlinefan]

I like the default HN approach (Slashdot used to do this too): users can upvote/downvote, and by default you don't see content that's downvoted far enough, but if you want to see everything, you have the option. That seems to work well enough against spammers.

[fuzxi]

That's the same as reddit's system. HN hasn't fallen into the pit of lowest common denominator jokes is the strong moderation, but the system still has inherent issues with high-quality posts losing visibility because they're less appealing.

[commandlinefan]

Well, no, reddit "supplements" it by banning pretty much every forum whose regulars don't match the owner's political ideology. There's no "uncheck this box to see all the stuff I didn't like" option there. I remember Reddit before they got ban-happy, and it was a much friendlier place then - and actually less overtaken by extremists.

[virgil_disgr4ce]

You just invented Reddit. The same thing will always happen: you say "communities should moderate themselves" and then you get "communities" that scream that 'libtards' need to be murdered, etc etc etc.. So then what?

Wash, rinse, repeat.

[grahamburger]

The difference is that each instance is a completely separate entity. They live or die on their own. They have to secure their own hosting, figure out their own revenue stream, and deal with law enforcement on their own. It's reddit minus reddit.

[necrotic_comp]

The main difference with mastodon is that there isn't just one server hosting all of the material ; you can create your own mastodon instance, and likewise, if a hateful/illegal mastodon instance is found, people can avoid it and/or law enforcement can find it and squash it.

It's more analogous to email and shutting down an email server than shutting down reddit.

[sneak]

> Each instance can set their own moderation policies and decide what other instances they want to federate with.

Yeah, this is why the fediverse is terrible, too. Your site admin shouldn't be the one deciding who you can DM, or which people can follow your feed.

Imagine if email worked that way.

[jonfw]

The good part is that you can be your own site admin, or you can choose one who reflects your ideals.

Moderation is great, the trouble is when a single moderation standard is used across a monopolized market.

[sneak]

I am my own site admin. The third largest instance's admin blocked my site (because I disagree with him about censorship), preventing 17,000 people from being able to follow me, or send or receive DMs from me. I also can't follow or read any of those users from my own server, despite their messages/profiles being public and available to the whole web.

[grahamburger]

This seems strictly better than what we have, though, where Twitter or Facebook decide the same. At least you can move instances, or start your own.

EDIT: also to note, email does kind of work this way with administrator applied spam lists. I would not want to use an email service that couldn't or wouldn't filter spam.

[berkes]

But email does work that way.

On my mailserver, I make the rules. I decide how my spamfilters, greylisting and blacklisting is tuned. I decide who is allowed to have an account. And therefore, I decide what is allowed to be sent and received.

[vorpalhex]

That's a feature many users want. Plenty of users never want to see things they disagree with or take objection to, and want to have someone else enforce that pattern.

As long as it's optional - and in the fediverse design it is - then it doesn't seem entirely unreasonable. Whether or not those instances actually survive and have any users or just immediately collapse into internal bickering is beyond the strict scope.

[ars]

> Plenty of users never want to see things they disagree with or take objection to, and want to have someone else enforce that pattern.

Am I unusual? When there is hate speech directed at me, I don't really care if I see it or not.

But I do want to prevent other people from seeing that hate speech (I don't want the hate to grow).

[sneak]

It's supremely rude to decide that you can be permitted to read something, but other people can't.

[zozbot234]

> Your site admin shouldn't be the one deciding who you can DM, or which people can follow your feed.

That's a necessary affordance, given the technology. A mail server could indeed refuse to forward some of your emails, and this could even make some sense e.g. as part of spam prevention policy.

[dane-pgp]

There is certainly an argument for preventing unsolicited messages being received from unknown users, but rarely do people value a spam filter which prevents them from sending messages to people. (I suppose there are some corporate filters which try to prevent accidental sending of sensitive information to unauthorised recipients, but that's not the "feature" we are talking about here).

It is perfectly reasonable for a user to want to DM someone on an instance that has different moderation policies from their own, and it is equally reasonable to want to receive replies to those DMs.

If the specifications or implementations don't allow that, then I suppose it has to be justified by saying that the DMs could be used for sharing copyright infringing material (or worse), and admins don't want to run the legal risk of hosting that on their servers. Legally, though, that doesn't seem any different to operating a mail server, which don't typically have Content ID matching systems on them. Perhaps the implementation of end-to-end encrypted DMs would assuage some of these concerns a little.

[notJim]

It doesn't quite have to work like that, necessarily. Your instance could provide higher-quality tools for users to moderate their own feeds. Or there could be different degrees of opt-in to the site-wide moderation. Or the moderation could be community driven.

I think there is actually a lot of room for experimentation around how we interact online, and moderation is one of the most important areas. We need to think harder, rather than retreating to one of the two default positions.

[folkrav]

Doesn't this basically boil down to the model we currently have, but decentralized, or am I misunderstanding?

[notJim]

Any idea why Parler took off rather than a right-wing instance of Mastodon? Is it about framing?

[zozbot234]

Isn't Gab exactly that? They seem to have "taken off" quite successfully among their own audience. Parler simply made a different choice wrt. whether to support interop and federation.

[deft]

They were for a bit but they've defederated and the software is a huge mess. There are right wing free speech instances but they don't have any of the media attention.

[at-fates-hands]

Timing.

It was right around the time congressional hearings were taking place about Twitter and FB suppressing conservative voices. A lot of conservative talk show hosts were talking about how they had moved to the platform and urging their listeners to do the same if they thought what Twitter and FB were doing was wrong.

When you have heavy weight talk show hosts pimping your product without paying them and they're touting your platform as one that doesn't censor speech, you're going to have a huge increase in followers.

[multjoy]

Because Parler was basically a Twitter clone. These are not ‘first adopters’ by any stretch.

[hinkley]

Before I decided that 'game developer' was not in my future prospects, I discovered the concept of reckless entitlement, where people will do anything the system allows, and a persistent subset of those people will rationalize that if you didn't want me to do that, then you should have written the code differently.

This is tantamount to "Stop hitting yourself."

Like the old line about academia, "the fighting is so vicious because the stakes are so low," the stakes for gaming and socializing with internet strangers are both pretty low.

Additionally, computers make you an efficient asshole. You can make a pretty big mess before you have time to think about whether you really should be doing what you just did. To err is human. To really foul things up requires the aid of a computer.

[mawise]

There's another dimension that nobody seems to talk about, and that's what happens with access to content is restricted. There is basically no censorship and no abuse from "those who can't get away with stuff on other networks" in the realm of SMS (or WhatsApp/WeChat/whatever your country uses). You can have a very healthy network if the authors are responsible for access control to their content. That's the idea behind my side-project of an easy-to-use, easy-to-selfhost private blog.[1]

When people want to use the platform to "become a thought leader" or "expand their network" then you're in the realm of public publishing which is where all the problems you cite become issues. The web makes privacy possible, and I don't understand why so few people are interested in that angle. For me I want to be able to share photos of my daughter with family and friends. I don't care if someone else wants to privately rant about the government to their friends--privacy enables both of these.

[1] https://github.com/mawise/simpleblog