Honestly, I say this all the time but I think people just need to treat federation like it's email. The difference is that the "email threads" are public to all, rather than in your inbox.

Seriously, the concern trolling around here regarding federation is almost comical. Apparently when major private corporation admins or volunteer mods arbitrarily intervene in the functioning of a platform it's totally fine, but when local server admins do it by moderating locally or de-federating from problem instances (whether temporarily, to deal with a flashpoint problem, or permanently in the case of problem servers), suddenly it's a huge problem and an indication of "drama".

I clicked, excited. I closed the tab when it required me to select a server. I don't want to read up on what rights server admins have over my account, why I should choose one versus another, which servers are de-federating which others, et cetera.

This is a real and recurring hurdle to the adoption of these technologies.

Philosophically, no. Practically, yes. Because it actually delivers the product. Creating an ideological and technical filter at the mouth of the funnel is absurd.

So then just pick a large, popular instance and be done with it. You've already made it clear you don't care what the platform's policies are, so why are you pretending this is a barrier?

Federation might run 100x better if instances were suggested based on geographic proximity rather than semantics, a concept which makes intuitive sense to people. 'Pick from a random and inconsistent list of servers in no particular order' is like demanding that people who are considering taking a holiday decide where to eat lunch after they arrive before they buy the plane ticket.

I'm not. Absolutely the onboarding experience can be improved.

The problem is you're losing the plot, here.

The original commenter complained about having to "read up on what rights server admins have over my account, why I should choose one versus another, which servers are de-federating which others, et cetera."

But you don't have to do that if you don't want to. If you're already willing to blindly join Reddit, you can blindly join mastodon.social.

And the app is already now driving people to do that if they really don't care (which is what the OP claims).

So this is already being improved (and yes, can absolutely be improved further).

Forced choices drive away users. People don't like making decisions without context because they feel like scams. That's why uptake is slow almost a decade into federation. The UX model is bad.

Ah, cool, so you just wanna complain. Got it. Carry on!

.world is having technical issues. (I could sign up. But first login spawns an infinite spinny. The only reason I know that's one of the larger servers is because of this thread.)

I'm doing it. But it's tedious, and the hacker in me sees an opening for a competitor to scoop out the 90% of users who don't care about federation, they just want it to work.

> you don't care what the platform's policies are, so why are you pretending this is a barrier

You really don't see the barrier?

It will end up like Reddit. But right now it isn't, and that's good enough to make a play for the users. Given a choice between that and choosing a server, signing up, finding its log-in unresponsive, looking for another server, signing up... (I haven't gotten further than this) who do you think wins?

By the way, we agree. I want a federated system to work. But simple sign-up fuck-ups, where even someone who's curious for curiosity's sake has to spend half an hour figuring out which servers even work at the moment,

I think that you're right in that the sign up UX is not great - there are also serious performance concerns, and in many ways the platform isn't ready yet. But I don't think that's going to persist for too long. I think some kind of "I don't care" instance (probably lemmy.world) will emerge, and the UX will improve. Perhaps it won't be in time, though.

Power users' power is users. We've frequently seen the celebs-first gambit by new social media entrants, most notably Clubhouse, and while it can generate hype for a bit, it's far from a proven strategy. It's frustrating to watch a re-play of Mastodon's fumbles, particularly since this time the protest is actually semi-organized.

Blindly clicking through a ToS while chuckling that 'nobody reads that shit lol' is in fact a better user experience for almost everyone. If your approach were so great, there would be physical cities whose population consisted entirely of architects and engineers.

the analog to a digital federation in the physical world, is literally just that, an actual federation. People making choices about what community they participate in isn't technical, it's social. Everyone who lives in a democratic society does in fact participate in how their city is run and governed or understands how to move from one state to another.

You do not need to understand the technical details of federated systems, but it is absolutely infantilizing to pretend that people are unable to choose or build the communities they want, and take some responsibility in maintaining them. This 'consumer' mentality needs to die, people need to learn to be proper citizens on the internet. We ask it from people in the real world, so we can do it online. Why are we pretending the online version of some 19th century company town is inevitable?

Not only is it infantilizing, it's wrong.

People create communities online all the time. Whether it's Facebook groups or subreddits or Discourse forums or Discord groups and on and on and on. That's literally how the internet has always worked.

Going back to the Reddit example, it's easy to browse Reddit for [whatever]; when a person decides to create and account, they become part of the u/* supercommunity, and from there can start posting/voting in the particular subreddits that interest them. The fediverse does this backwards, equivalent to making people choose a default subreddit as a condition of becoming a user. It forces people to pick a home community which will shape their whole experience of the fediverse and how they will be perceived by other members of the fediverse before they have a chance to explore the system in a noncommital way.

it is absolutely infantilizing to pretend that people are unable to choose or build the communities they want, and take some responsibility in maintaining them

It takes time for people to figure out what communities they want to inhabit within a new protocol, which is why the perception of locking them into a single originating community is an antipattern. People don't necessarily want to be judged by their originating instance, because it reduces them to a one-dimensional caricature of themselves. You could join a general-purpose instance, but there are multiple general-purpose isntances, what makes one better that or differnt from another? There's no way to tell.

Critics of the fediverse model have been pointing out this rather obvious stumbling block since it was established, and fediverse stans just keep sticking their fingers in their ears going 'la la la can't hear you.'

This 'consumer' mentality needs to die, people need to learn to be proper citizens on the internet.

It has nothing to do wish consumption vs citizenship. You're demanding people tie themselves to a point of origin as a condition of existing in the fediverse, which is one of the worst aspects of meatspace. If you join a general server and then start participating in some niche topic (idk, fursuits), people in the niche instance are likely to write you off as a tourist. Conversely, if you decide to join a furry instance but then participate in discussions about sports, a lot of your interactions are going to consist of 'go away, furry weirdo'. You could create multiple acconts for your different interests, but now you have the headache of maintaining multiple accounts.

There just isn't a good reason to lock people into a particular instance in order to sign up. It's just reproducing nationality, a concept that many of us would like to dispense with altogether.

Federation is confusing and alienating to non-technical users, and nerds who love federation tend to have a mediocre grasp of social dynamics and gloss over the inevitable abuses. Federation stans need to grasp the fact that nobody who is not a full time nerd cares about how federation works at the technical level, they just want a place to socialize with the assurance that they won't be overrun by assholes. All the Federation stans go into it the idea that 'you can just defederate' whereas non-technical users go into it with the idea that they don't want to get raided in the first place.

The existing model of federation is not working. Users don't want to know about the infrastructure any more than people going into a coffee shop want to looka t the architectural blueprints of the building, and federation is clearly unable to pre-empt raiding behavior automatically.

Neither does Reddit. Have you not heard of brigading?

When I show a site like Lemmy to my non-technical friends, their eyes glaze over and they are confused by the necessity to choose an instance and the inability to make any kind of meaningful comparison between them. This is the exact same reason that Mastodon has never really taken off.

That's quite understandable actually, the Lemmy UX sucks. Try showing them fedibb.ml instead. It's Lemmy under the hood, and federation works perfectly - but you'd never guess that!

Doesn't this also describe Reddit mods?

https://arxiv.org/abs/1803.03697

If enough creators move over then the lurkers will follow. But attracting lurkers at this stage is pointless if there isn’t any content for them to look at.

Governance is up to individual instances/communities rather than one faceless megacorp.

It’s a feature not a bug.

So my appeal to join lemmy is not aimed at Joe Blow, it’s aimed at the actual contributors who make communities possible. If enough of them switch over then lemmy will gain critical mass, and the rest is history. Joe Blow can go wherever he wants. I doubt he’ll stick around at Reddit when the content stream disappears.

Contributors generally create content because they want it to be seen by more than a handful of nerds and don't want it to disappear within weeks because their instance has ran out of money.

Those contributors will either stop contributing for good, go back to Reddit (because it's still a better experience than any of the "fediverse" bullshit) or jump ship the second an actually competent company builds a centralized alternative that just works.

So no, I reject the premise.

People absolutely do care about governance. They care if they're getting spammed by crypto scammers or getting targeted by abusive trolls. They care if their political views are being censored or things they find offensive are being promoted.

The difference is, on a traditional, privately owned platform, the users have limited choice and no say, and they've gotten used to that as the status quo.

And if you're a user who really doesn't care, cool, just join mastodon.social or lemmy.ml and move on. Problem solved.

As for complaints about onboarding, the official Mastodon app already drives people to mastodon.social (much to the chagrin of some folks in the community), so I have no doubt those issues will smooth out with time.

If the fediverse is like email, then filtering should happen on the client. It's not like I expect gmail to "defederate" an email server that's used by people I don't like. I expect it to facilitate my emails with anyone else, regardless of which mail server they use. And I want gmail to do some spam filtering for me, but I expect that I can see the filtered messages and show the filter what it missed or wrongly labeled as spam.

If we leave the federation up to the servers, we'll just get a "fragmentverse" of walled gardens instead of a single big one.

I'm pretty sure gmail defeds email servers all the time. IP blocks are the oldest tool in the toolbox

With forums there are arbitrarily many recipients of any message. A lot of people may not like furries or communists or what have you, but demanding that every other instance ban them as a prerequisite to interoperability is how you kill the network by suppressing everything but the lowest common denominator.

This problem is created by tying accounts to instances. Because then other instances, instead of banning the accounts they don't want from only their own instance, can try to get them banned from everywhere by threatening any instance that allows them with disconnection. Which is poison.

That's a huge issue. My identity should not be tied to anyone but me. This has to be one of the biggest things that scares (admittedly more knowledgeable than average) people off early when they're required to pick a server. Mastodon's "don't worry you can move servers later" is better but still not frictionless, and does not work unless your source server is still online.

Boom, you have just reinvented usenet.

The instance I use hasn't defederated them yet, but I've already had to block most of lemmy.world's communities from appearing in my feed.

Nah, just kidding, it's a good point. But it's a good starter server, and once people get used to the idea, I think they'll naturally gravitate to other servers.

Let developers create each "source" of content or comments, packaged into a module ("NYT article fetcher," or "Reddit comment fetcher"), and let me choose which modules to install. Then delegate the infrastructure for executing scrapers or curating feeds to a set of federated servers that I can opt into. Or better yet, execute the scrapers in a local sandbox on my own device, so that the NYT can't block me because my requests are my own. You could argue it's just another form of a web browser.

I think what we are seeing is a taste of what could be, as the technology improves and the UIs of the clients become more friendly and add features to leverage the tech.

We are watching it all happen in real time. Problem is everyone wants a polished experience day one. It’s going to turn a lot of people off, but I’m not sure there’s anyway around it. Once refined, this is the kind of tech could be key to the people controlling the future of social media.

> I can log into Mastodon and search and find the user I created on lemmy. I can also search, follow and look at the posts @[email protected]. All from my Mastodon app/account.

So is your user registered at lemmy? Or at lemmy.world? Is there a difference?

And you can view the posts at @[email protected] - but do you have to have a separate beehaw account to post?

Well, you could, but Beehaw basically blocked Lemmy.world. But otherwise the metaphor holds true. If you make an account on some other instance, you can subscribe to communities on Beehaw, post on Beehaw, and reply to users on Beehaw posts. When doing so your username, instead of being e.g. "mahogany", changes to "mahagony@[LEMMYSERVER]". That's how you can tell who is a user of that particular server.

The fediverse is set up kind of the same way. You can interact with people if you know how to find their account, the address of which is formatted similarly to an email address. You and whoever you're interacting with don't have to have accounts on the same server to communicate with each other.

It's like if you opened up a reddit thread while logged into your hacker news account, and decided to respond to a comment in the reddit thread. Federation would give you that ability. Your username, in that thread, would be something like "[email protected]" or something like that, to show that you were posting from a different server.

You have an account on one server eg. [email protected]. People on example.com smallreddit will see you're posting from bar.com (ie. your username will look something like [email protected]). The bar.com server's UI may expose stuff from other servers to bar.com users, and the reverse.

Defederation means severing those visitation rights and other interop between two small twitters or two small reddits, something Mastodon servers use rather liberally.

In practice, what ends up happening is that at least Mastodon has hideous degrees of ideological conformity via defedding and general woke modding, and there's essentially standalone dissident "witches' covens" and an archipelago of servers whose maintainers tend to think different politics are de facto evil. I could hardly use the same account to see what progressives, gendercrits and various strands of the right are talking about the way I can on Twitter, for example, because of the defederation moats that have been erected.

And yet, every single one of these fediverse discussions is full of people basically saying "This doesn't spoonfeed me with zero effort, I turned it off after ten seconds!"

I have had my Mastodon account since 2017, I posted once and then never really looked again, and then after Twitter imploded and people started using it, I came back. I was an active Twitter user at the time as well, and I just stuck with it for a while, but eventually I just quit using it because there was just no quality interaction on it at all, with people I agree with or otherwise.

I also had kind of a boring Masto feed at first, until I figured out to subscribe to hashtags, and now I have several quality conversations a day about things I'm interested in, with congenial strangers who are also just there to talk about cool stuff. You have to change your habits a bit, it's fine and it's better.

I have also identified a strong tendency in this forum to do very well-outlined explanations of what is difficult or inefficient about federation, but invariably, the problem they're describing exists in a much worse form on the silo'd alternative they are implying we should stick with.

The fediverse (and decentralized social media in general) breaks the mold of extreme ease of use and therefore subverts the cultural expectations, thereby violating one of the ways a user would establish if the application is "good", and thereby it looks "bad".

It's like, technically X algorithm is a superior algo to use. But I can import Algo Joe's library that's 50 years old and already integrated into every language I know, so I'll do Joe.Sort().

Anyways, I see zero evidence that Reddit will come back from this conflagration in any sort of good shape, and I see even less evidence that the VC business model is going to do any better with anything else.

Selective low effort is what those professors exhibit. But their ideas are valuable and worth publicizing.

Twitter is an example of many low-effort scientists writing some good content.

Twitter is also the current textbook example of the problem with low-effort platforms that are controlled by capital. Like, do I even need to explain this to anyone? Are there still any Musk fanboys in the room who think he's playing 4d chess?

But while Twitter circles the drain and the world scrambles to figure out what to use instead, and Mastodon steps up as a competitor which cannot be seized and ruined by capital, ever, Reddit stands firm as the one true platform that got it right! Everyone gazes in awe at the one last hope for the capitalist internet, which seems to be doing just fine.

And Reddit looked over at Twitter, and then Reddit looked down like Rorschach, and said... "Hold my Over 8000 beers..."

Also, reddit’s value is exactly that you have very different demographics on each subreddit, and “hearing the voice” of people is the whole point.

The fediverse is in its early days. Nobody would argue that Reddit.com was "easy to use" in 2007.

As a matter of fact, there's already a Lemmy/Fediverse UX that works just the same as an old-school forum/BB; the whole federation facet is exposed as an extra feature on top of that: https://github.com/LemmyNet/lemmyBB We know that UX works, because it was quite popular even among novice users.

And instead of relying on servers to federate with each other (which basically just shifts the problem, replacing one centralized walled garden with a patchwork of smaller gardens), why not let the client decide which servers to subscribe to? In an ideal world, the client could even merge comment threads when the same story is posted to multiple servers that the client subscribes to.

Without the centralization, why bother? Without the monolithic environments it's all private gardens. There's no point going and standing in someone's private garden while they're away.

It's seeking of the public square that is generating this situation, over and over and over again. This mess IS the territory. Either there's a way to have best of both worlds, or some kind of 'both worlds, in a compromised way', or this will always happen and this, too, is the territory: all public squares will be bombed for one reason or another until they're gone.

'Documentation? Just learn programming and it will be obvious.'

This is what old reddit did so well. You just open reddit.com and you see a dense wall of stories, along with vote count, number of comments on each story, and which subreddit that particular story came from.

All in ZERO clicks. No images or graphics at all, aside from little thumbnails you can easily ignore. Minimal, dense, and clutter-free.

I joined the former today and can't log in. (That's fine. Scaling difficulties are a champagne problem.)

But first I had to spend about half an hour reading comments here to identify lemmy.world, as well as confirm I wasn't joining e.g. the Nazi server. I still don't understand what power I'm giving the admins of that server. I started looking into another server to join, but after looking at more posts, I'm not seeing a discussion that's pulling me in right now.

Another feature that would help with bootstrapping would be something like a "Reddit bridge," i.e. a Lemmy instance that proxies requests to Reddit and allows you to authenticate your Reddit account so you can read it and post comments to it like any other Lemmy instance.

Thing is I _still_ closed the tab and will likely forget about it. Turns out that thousands of people posting nothing but complaints about "REDDIT EVIL", isn't that interesting to read.

In terms of discovery, there is a centralised community browser of sorts... but you'll see it's got quite a way to go: https://browse.feddit.de

However, as someone willing to put the time in to learn something new I've found my efforts rewarded. It does take longer than 10 seconds to get used to Lemmy and the fediverse, but I'd say no longer than 15-20 minutes really. I enjoy being part of something new and I hope it lasts. Perhaps it won't hang on to many of the new users it has gained, but I expect this won't be the last time Reddit does something its users don't like and next time around more will stick, hopefully presenting a very good alternative.

Lemmy - A link aggregator for the fediverse.

Join a Server - Run a Server - Follow communities anywhere in the world

Lemmy is a selfhosted social link aggregation and discussion platform.

No, no. You lead with the benefits to the user. This isn't Github.

It suffers from the same problem as Mastodon - "Join a server". But which one?

These federated things need to be set up so that there's "no wrong door". Enter anywhere, see the same stuff regardless of where it's hosted. Discovery and hosting may be distributed behind the scenes, but they have to have a unified user entry point.

USENET had that problem solved. It was federated, but it didn't look federated to the end user.

Seriously. Have a default server (or set of defaults) newbies dump into after completing a simple sign-up flow. Then let them move it as they benefit from the product and gain an incentive to learn about it.

Reddit is up in arms about API pricing and mods. Do you think the average Reddit user would have signed up if they had to first jump through hoops to prove their respect for community-based architecture? No! An infintessimal fraction of users care about philosophy or architecture, even when those values and decisions directly cause effects they follow. Users came for their cat pics. Give them their cat pics. Pitch them on the back end afterwards.

- Servers choose what they want to host and what their personal bandwidth limit is

- A user visiting a specific subforum automatically downloads from whatever servers are currently available to serve it

Then you’d never have to manually choose a server.

Similarly, you need to join a Lemmy server to have an identity there, or anywhere on Lemmy. Unlike private torrent trackers, you only need 1 Lemmy identity, and you can subscribe to other servers' communities from your original server.

You could just bruteforce IPv4 until you hit a peer that will give you a set of known peers.

most every client comes with a list of 'bootstrap peers' for dht though.

This has to be parody.

> An infintessimal fraction of users care about philosophy or architecture [...] Users came for their cat pics. Give them their cat pics. Pitch them on the back end afterwards.

I'm afraid you don't understand the concept of the fediverse. The philosophy is the point. Making reddit2 is completely useless. This is not a product, it's a vision of how we manage communications as a society. It implies involvement, because in a democracy you are part of the decision process, and you act as a community. Your egocentric, utilitarian, self-obsessed mindset is the complete opposite.

I can't be really surprised to see this on hn, but oh well.

Nothing about cleaner onboarding requires anyone to compromise principles. And nothing about a guiding philosophy requires you constantly evangelize. Let the merits of the system speak for itself.

> implies involvement, because in a democracy you are part of the decision process, and you act as a community

Then the sign-up page is a poll test. Democratic ambition strives to make participation easy. That doesn't mean the act of participating is easy. Just that we identify needless barriers and ideological tests as undemocratic. As a result, democracies--time and again--outcompete their centralized peers.

What I'm seeing is not competitive, and it is not democratic. It's closer to theological. I'm hopeful that someone will embrace these principles more productively so we have middle ground between Huffman-Musk and crypto/web3.

What happens here is that we have the occasion to take control of our community gatherings. How do you take control if you don't even know what you're talking about ? How can you discuss the form of society we're building if you don't want to discuss and be told how to think ?

Now, I understand that what I'm asking for is out of the norms, because we're not used to decide collectively and run our own lives, the entirety of our education and our lives happens with someone else taking decisions for us. It's comfortable, we don't need to think too much, especially when we agree with the decisions. Here is the occasion to do it differently. Why not do it ? And if picking an instance is too hard, why not pick a random one ? We want to have the best thing for us individually, but we want to be told what it is. We want to not be subjects to the whims of the owner of a platform, but we don't want to understand where we're setting foot and prefer a generic instance that will be for everyone. Hust lite reddit, but different. That's not building, that's consuming, and if you're only there to consume, do you really want a democracy ? It sounds like a benevolent dictatorship is better.

In short, I don't believe actual democracy can happen without understanding the situation.

So if I buy the philosophy (which I do), how do I (practically) get started? Let's say I want to use the internet to actually interact with people in my community. Is that even possible anymore? Do I just download Mastadon and start wandering the dark? I've wanted to penetrate this "intellectual" space but I just do not know how, and I am a software architect of 20 years.

Then when you want to participate, that's your starting point. Lots of people are lurk-centric, but you are already in the comments section here, so moving your posting probably will not be the issue so much as contextualizing it, which the different software and instances do help with.

A lot of "community" is in having some ritual, some events to show up for, or some events you organize yourself. It doesn't take a lot of those to fill up a calendar, but it definitionally needs to be non-zero.

This isn’t how actual democracies form. Solon and Cleisthenes didn’t gingerly nudge their ways towards democracy; that would have been a gift to Athens’ elite to stymie them. They put forward swooping reforms that changed who held power and brought everyone* into the tent at once. They told people why their ideas were better, because they were. (There is actually a great analogy here in the tribes. Tribes were assigned. You could later change them, but there wasn’t a giant before we start being useful we have to put on these hats process.)

Providing a simple path isn’t the same as taking away the difficult one. You can deal with a newbie pool from which one must “graduate” in some amount of time, for instance, thereby giving nobody a default bias advantage in users.

Systems compete. Ideas compete. I think the collective ideas the internet were founded on were strong. And not every one of them need be a cultural force—it’s fine for e-mail to be practical even if it’s in practice controlled by a few companies. But this actually has that opportunity to create a new culture, and it’s throwing it away by telling the curious it’s too good for them.

I think there are going to be incredibly valuable lessons-learned from these federated models, and interesting things will happen there. However, all of that will be capitalized by a nimble, forward-thinking, idealistic-yet-centralized model that gets rid of all the ideological cruft and polishes a real, useful, and engaging product.

The "product" mentality is the cancer that led us to where we are. We must get rid of it.

> I think there are going to be incredibly valuable lessons-learned from these federated models, and interesting things will happen there.

I totally agree with that. The control an admin has over an instance is still problematic, and someone shared something about confederated protocols (https://nexus.blacksky.network/zine/00000001/confederal-prot...) and that is an interesting way forward.

Lemmy existed before the Reddit exodus. It will exist when they all go back.

IMO, this will be the future of high quality interaction. Either small communities that are federated to prevent the same issue forums had, or places that charge a monthly fee for entry.

> Most people only care about the end product, the end experience.

Because that's what we keep telling them: don't think, let me take care of this, CEO knows better. That is how we arrive at this crooked situation where sharholders come first, business partners second, and users last.

USENET also had the "chose the wrong server" problem.

If I chose my ISP (e.g. Verizon) as my USENET entry point, then I would miss out on many newsgroups that Verizon didn't have. Or, Verizon had the newsgroup but had extremely short retention.

So, I paid $12.99/month for GigaNews to get all the newsgroups with months-to-years retention.

Lemmy/Mastodon/etc replicate the same federation issues as USENET because running servers costs money and different people have different thresholds of spending. This independence/freedom of the admin running the server the particular way they want is touted as a positive but it also has inherent disadvantages.

The contemporary situation of Beehaw instance defederating lemmy.world -- is replaying the same tradeoffs that USENET went through. (E.g. similar to an ISP's USENET service choosing not to carry alt.binaries.* or whatever.) Beehaw explains they have limited resources and can't handle the influx from lemmy.world. Yep. Very understandable.

It creates the possibility for a business model where subscribers pay for access to a big news spool and lots of groups. Just like USENet.

There will be some growing pains, but there will eventually be a continuum of "free" lemmy servers, for pay premium ones, and lemmy servers where the front end is only an app on your phone.

And they'll all see the same messages.

I think you should re-evaluate that assumption:

https://www.google.com/search?q=mastodon+instances+blocks+ga...

https://www.google.com/search?q=mastodon+blocklists

https://www.google.com/search?q=mastodon+blacklists

https://www.reddit.com/r/Mastodon/comments/kw8jht/how_do_i_s...

Also, expand the ">Moderated servers" heading and scroll down through the instances they black:

https://mastodon.social/about#unavailable-content

Why do different Mastodon instances decide on not federating some of the other instances?!? It's because the owners pay money for running their particular server which makes them feel entitled to run it the way they want.

The "pick any server, it doesn't make any difference, you'll see _all_ messages anyway" -- is not realistic given that each fediverse node administrator can exercise their freedom to choose what messages their server accepts.

That is true, but most Mastodon apps make multi-accounting a breeze, and I imagine the Lemmy ecosystem will provide the same functionality.

It also helps that Lemmy is trying to court the userbase of Reddit, who are already a pseudo-anonymous lot and in my experience don't seem to have as much of an attachment to their old content as Twitter users, deleting their old posts and accounts on a regular basis.

Originally, Google was pure discovery. There was no Google content, just links. That's pure discovery. A standalone blog is pure content. Together, you have a social system.

Separating those is a step to freedom from oppressive service providers.

The nostr community right now mostly want to replace twitter rather than reddit, but in principle it would be possible.

Regardless, if I was able to capture 1% of the net, that's not a failure by any stretch.

How did IRC succeed here? Have any hard evidence as opposed to intuition or speculation?

IRC, like e-mail, is grandfathered in. There wasn't a centralized alternative that worked when they were born.

I think there were many options at the time.

Maybe the DCC transfer and easy scriptabilty thanks to Khaled Mardam-Bey's mIRC might be a valid claim.

Perhaps also the impermanence of the history of the channels afforded certain types of interactions that you wouldn't do with you know, theoretically forever scrollbacks.

I guess this brings us back to classical marketing about how you can't have a sustainable differentiated product based on negatives. (as in, this is not the bad guy). You need to have something in the affirmative

1988 versus 1996 and 1997, respectively. AOL proper was an IRC competitor, as was CompuServe, but their definitions of working weren't different from IRC's.

That same argument didn't save WAIS, Fido, or gopher. Nor did it keep Tymnet or Bitnet around and it also didn't give Compuserve and The Well a seat at the winners table.

Magnavox putting out the first home gaming console in 1972 hasn't made them a gaming juggernaut nor does Xerox run the desktop and Palm, Go and IBM do not control the smartphone market nor is my laptop by Grid Compass my microcomputer by MITS my spreadsheet by VisiOn or my pants from Arnold Constable.*

It also wasn't a saving grace for Friendster, Myspace, or LiveJournal.

Citing an early creation date is a survivorship fallacy here.

The real question here is why didn't it die like everything else. Why is it one of the few legacy survivors?

---

* If that cadence sounds familiar, I'm borrowing from Ogilvy's Confession of an advertising man bottom of page 61 (https://archive.org/details/confessionsofadv0000ogil_r2a4/pa...)

Separate problems: getting off the ground versus surviving.

IRC got off the ground because it didn't have centralized competition with a capability advantage. Why it persisted is a deeper story. Lemmy is still trying to get off the ground. It, unlike IRC, does have such competition. As such, the old playbook is obsolete.

I'm going to disagree. The old playbook is to empower users in unique ways that at the time feel almost forbidden and magical and to competently execute that.

You could claim all those avenues have been explored but I disagree there as well. The surrounding context and possibilities are always on the move so the underlying potential is always changing.

That's why say, YouTube, the 40th or so on-demand video company, which happened to launch when digital cameras and broadband internet were becoming widely used, was the first successful execution or why smartphones didn't take off until the rollout of 3G or, looking into the future, VR might finally take off at attempt 100 after some related thing changes.

Text only, transient/short life data, is a lot cheaper to process and serve than images, permanent posts, etc.

It won the initial buy-in of us geeks/nerds/hackers/whateverthephraseofthedayis who gave it a rather solid base.

It’s a very personal type of communication. Real-time, immediate, and to a lot smaller audience (more intimate) compared to web forums, Reddit etc.

And finally, I would posit that it did actually die. What remains now is small, compared to how popular the likes of Reddit Twitter etc are, vs how popular IRC was in its heyday.

This is an interesting question. Wikipedia claims 230,000 users at peak times which is still quite a bit more than say, gopher. A 98% drop is real but you'll still see IRC occasionally for software purposes (like say, Debian)

Maybe it was a coalition of people there for different purposes and some of those groups have fallen away for different places.

For instance, people were doing dating and sextalk on irc back in the day. That application has been superceded by many other places.

It didn't help that the protocol evolved at an absolutely glacial pace with very uneven support across various networks, and doing anything but the most trivial sorts of channel management had to be done through a bot. Matrix seems like the obvious IRC successor due to its seamless two-way interoperability with IRC networks, but I suspect that it won't hit critical mass until Discord starts causing problems for its userbase.

The mega pun is that if you are hard working, intelligent and reasonable this is going to be incredibly hard for you. You are going to need help from someone who is non of those things.

Now that I've been on Mastodon and Lemmy, it makes more sense, but this is a HUGE barrier to adoption.

I would argue they are. Any of the servers will work fine. There needs to be a paradigm shift that only things that are centrally managed and have a profit motive are good.

Is this actually an issue?

If you're looking for a straight up reddit alternative in all its scope, I can see it being an issue, but if we simply want a better iteration, then isn't one more catered to technical users, perhaps be a good thing?

To me this feels similar to windows vs linux. The pains you describe in learning to use lemmy reminds me of windows users who complain about linux not being exactly like windows, UI wise.

But for users like me, after having un-learned windows and customized my linux experience, the result is just downright better from any angle. Yes, it was a time investment learning the ropes, but it was worth it.

Perhaps the modern approach is to stop trying to be like the most populist standard, and embrace platforms that are midway between the social giants and obscure niches.

Well, to be honest, that part does sound exactly like reddit.

But people fall on reddit by links to a public page somewhere, avoiding the empty page problem. After a short time on their site, I still don't know if I can link to a comment in Lemmy or if I can browse a community.

My current list of grievances:

- No standards of any kind on sharing the same identity, e.g. between Lemmy and Mastodon.

- No support for any distributed multi-master database (I'm currently trying to see if I can hack Lemmy to work with CockroachDB, but I'm probably going to give up).

OT but what pains did you experience with matrix / what would you have done differently? this'd make a good blog post/HN article at this moment imo.

The issue with Synapse is the same as I currently have with Lemmy - lack of HA. One particular machine goes down and the system is 100% unavailable. And most of my machines are hardware I own, located at homes (mine, parents, etc) in different countries, so they're slightly more prone to random power or uplink outages than AWS' us-east-1. Hobbyist-grade geo-distributed cluster, hah :)

I've solved this for my email (two mailhosts on different continents, Dovecot dsync replication for mailboxes) and my storage (Minio works, and I'm slowly experimenting with Tahoe-LAFS), but I haven't solved it for Matrix.

Dendrite supports distributed deployments and reportedly works with CockroachDB (albeit not officially supported). Not sure what Conduit supports (last time I've checked it was all about embedded databases like sqlite, rocksdb, sled and persy), but it has pluggable database backend architecture so when I'll finally decide I want to fix my stuff, I'd certainly check it out as well.

We’re also working on account portability (MSC4014) that would eventually support active/active at the matrix layer, but it’s not ready tet.

PostgreSQL replication is hard for me. I know how to set up "classic" replication, but automatic failover and recovery that can solve split-brain issues is, sadly, not something I've figured out. And if it's not the machine, but the network that goes down (which happens few times a year on consumer-grade connections), it's a problem that happens in practice. I have learned how to repair nodes by hand, but I don't want to do this.

So far, I've managed to avoid thinking much of it by using distributed databases (CockroachDB and Consul, in particular) that already have all the magic built-in, so I just have to be careful with the settings.

Maybe I just need to learn how to deal with PostgreSQL...

(if you have a third-party commercial PG extension with clustering, like EnterpriseDB or CitusDB, ask them, they'll tell you what they want you doing.)

there is no inherent clustering support in postgres. You can't scatter tables across servers/etc. It's not supported. Always one master server - if you need more storage you can use SAS/fibrechannel/etc of course but there is one postgres instance that's in charge.

All of the postgres replication idioms rely on that model: the WAL is reliable up to the point it's been received/verified. If you've faithfully played out the WAL throughout your existence, everyone's storage instance is at the exact same state and the only thing that matters is who's got the most recent timestamp from the host.

This also includes bringup strategies: if you have an atomically checksummed snapshot of the datasets (all storage, and pg_wal, at the same instant) then you can also spawn new nodes. ZFS can do that. And with an incremental copy-on-write you can replicate the base state, then replicate only the changes (which is much faster) and then fail over quickly. That's popular for other applications that don't have sophisticated ACID stuff.

Anyway short version is, I'd think about a simple blue-green system (or round-robin system) where you can set an environment flag, and blue frontend nodes can simply choose to direct traffic to green backend's pgbouncer or similar when they see that. Then you flip web traffic over. Then you shut down the blue pg instance, and stuff fails over gracefully. The "healthy", "progressing", and "migrated" states, if that makes sense. And there is always one healthy node, even if you need to bring it up from backups. Yes, you have to look whether the other side is still healthy but that is now a k8s/pod problem and not an app one.

If you can live with an instance in the US and one in Europe or similar, a simple blue/green failover is probably easiest.