To me "Web3" frequently feels more centralized than just the regular web.
I love that on the regular old internet I can stand up different websites, host them myself, and have anyone with a web connection be able to pull it up on their computer and see whatever it is I'm putting out there. There is some level of centralization if I want to buy a domain name or do other certain things, but it's pretty minimal and I feel mostly in control, at least for the kind of things I want to do.
That kind of standing up my own little server and running it myself while being accessible over the broader web feels decentralized. I can put whatever I want, you can put whatever you want, anyone can access it. Buying into some more distributed web3 type hosting feels more centralized at least in the way that I personally feel. I have to buy into a specific blockchain or platform, host in a specific way, hope that it gets picked up by other distributors, deal with hashing and immutability, etc. Maybe it's a difference in understanding of the word decentralized, or a different emphasis on certain parts of the definition.
> There is some level of centralization if I want to buy a domain name or do other certain things, but it's pretty minimal and I feel mostly in control, at least for the kind of things I want to do.
This actually related to a fundamental lack of control.
Lots of tech we buy now promises "smart" but because you need to get to it from your phone, some corporation "needs" to "help" connect your device and your phone.
for example - you want a "smart" thermostat. To connect to it from anywhere from your phone to change the temperature, the thermostat contacts the mannufacturer. Then your phone has to talk to them as an intermediary to contact your thermostat.
Conveniently, they can collect data from your thermostat, or your house, or your phone, maintain a "customer relationship" or even charge you for the "service".
if things were properly decentralized, your phone could just contact your "smart" stuff and nobody else would need to be involved.
Not advocating for either, but I believe the term is intended to contrast with Web 2.0, which was services like YouTube and Facebook, that centralized content hosting and delivery away from people running their own web sites and putting it into corporate services that offered individual user profiles instead. Web 3.0 was intended to provide a similar level of convenience and not require all people to self-host in order to be producers as well as consumers, but without corporate entities having ultimate veto power over what can and can't be hosted and delivered.
The short version of that is it's not as decentralized as Web 1.0, what you describe, but it's not as centralized as Web 2.0.
I run multiple websites since I have been roughly 14 and I never had any of those problems. I am not saying they could theoretically happen. What I am saying is that to most people running small websites they don't.
I don't use cloudflare because there was never a reason to. And if there was a DDOS I'd just wait. My websites are important to me, but I don't give a damn if they are down for some time. This isn't critical infrastructure.
I have been running internet services on shared hosts, VPS, dedicated servers, colocated own hardware, since 1996, including larger projects that have their own Wikipedia entry etc, and have never needed to put anything behind cloudflare or other ddos protection layers. The providers I have worked with mostly passed on DMCA takedown requests without issues. The small internet is still fine.
No, it hasn't. The small web is quite healthy to this day. I haven't heard anyone who actually runs services like this complain about any of the problems you suggest.
Those problems only happen to large entities. There's simply nothing to gain from DDoSing frankshobbytrainsandflowers.com other than sheer malice, and then it only lasts for as long as you care to spend resources. Frank won't care and likely won't notice if his website is unreachable for any length of time.
Conflating the small web with enterprise problems is an extremely ignorant take.
Besides, the single biggest threat today is not some motivated attacker trying to deprive you of your three visitors a week, the threat is gigacorps exploiting the public commons with reckless abandon. The only thing small web admins are actually concerned about is LLM scrapers.
You should try actually talking to small web admins before spouting off nonsense like this.
You should trying being a small web admin from around '95 to 2018 then. It was not rare for some random small site we were hosting, thats not infected with anything, to start getting flooded with 10mb+ of traffic for no apparent reason.
Whilst I disagree with what you have written, it is disappointing to see your genuine statements being downvoted.
Downvoting is meant for comments that detract from the discussion. Downvoting a comment simply because you disagree (and can't be arsed typing a rebuttal) leads to groupthink and unpopular ideas being hidden. This place is slowly turning into Reddit. And no, that isn't a cliché, it is a statement of fact.
I must be on a different internet for the last 30 years from other HN'ers. That or a lot of people say really boring things that won't matter if they are on the small web or consolidated web. It's the kind of things that get you kicked off the big providers that also make you targets to get blasted off the face of the internet.
Web3 and blockchain are not the only form of decentralization. Email is a decentralized protocol that has stood the test of time.
The bigger problem is having mega-entities like Google, Meta, and Amazon dominate the web. Instead of crypto, there should have been a focus on allowing mid-size players to have more power.
For anyone about to disagree with this; you've been lied to.
Many thousands of admins are currently operating independent email servers with zero problem, and have been for decades. Mine is going on 15 years. In practice, you have to actually try pretty hard to misbehave enough for Google and Microsoft to notice and block you. For the vast majority of independent servers, we have no problems at all routing to gmail or outlook.
Reports of the death of decentralized email have been greatly exaggerated. Independent mail servers are alive and well to this day, just as they always have been and (probably) always will.
> In practice, you have to actually try pretty hard to misbehave enough for Google and Microsoft to notice and block you.
As far as I can tell, the "misbehavior" that got me blocked by Microsoft was being hosted on Linode... where I'd been for around ten years at that time, all on the same IP address. Tiny server, had never emitted a single even slightly spammy message, all the demanded technical measures in place, including the stupid ones.
Because of the huge number of people stupid enough to receive their email through Microsoft, I had to spend a bunch of time "appealing". That's centralization.
On edit: Oh, and the random yahoos out there running freelance blocklists can do a lot of damage to you, too, by causing smaller operators to reject your mail.
They probably threw a fat CIDR block in their IP blacklist to fight off a spam campaign, and your IP was caught in the dragnet. This is how the big companies do it. They’ll evaluate for risk of false positives and as long as that stays below a threshold, they proceed.
> Because of the huge number of people stupid enough to receive their email through Microsoft...
Can you please elaborate on this? I use Outlook (@outlook.com) for my personal email management, but would definitely switch if there is a better alternative.
Microsoft will cut off your account if any of its abuse detection heuristics misfire. You will probably not get it back, ever, period. And, by the way, I don't remember if it was for email or storage or both, but Microsoft has admitted to going so far as trying to brute-force the passwords on encrypted ZIP files to run those heuristics.
Microsoft will (intentionally) lose your incoming mail if, say, it comes from the wrong AS. You have no control over most of that, and you will not be told about it.
Microsoft may or may not be data mining the actual content of your email. Possibly also after trying to break your encryption.
Since you not only use Microsoft to process mail, but have also inexplicably decided to use Microsoft's domain name for your address, you'll find it difficult to ever move away from Microsoft. If your account gets deleted without warning, you'll be permanently screwed. If Microsoft, say, decides to go out of the email business in 10 years, or start charging some ridiculous fee, you'll probably at least get some notice, so you'll only have to scramble to change your address before the hammer falls.
AND you're contributing to the centralization of the whole damned Internet.
The better alternative is to self-host. Even though you'll then constantly have to worry about your outgoing mail getting dropped, at least you'll be in a position to notice if that happens.
If for some reason you can't handle that, then at least register your own domain and host it with a smaller provider that gives you some guarantees. You will, of course, have to actually pay a small to moderate fee for the reliability.
Fantastic breakdown -- very much appreciated. I don't mind creating a new email address if needed, even though I've had this one for over a decade. What you mentioned are very real risks that would be catastrophic should they actually come to fruition.
I'm taking some networking courses online which will hopefully help in terms of being more tech literate and being able to do this.
Regardless, I'm definitely going to start the transition. I also use their cloud storage and have very important documents in there.
I had my mail hosted with my managed server for 10 years, then they started charging ridiculous amounts of money for SSL certs, so I moved all websites to a self-managed 10 times cheaper yet more performant webservers. Then they wanted to force Microsoft 365 on me and I migrated the mailserver in a similar fashion.
I use mailcow and did the migration over a weekend. It took a few hours after to get all the spam stuff right (SPF, DKIM etc), but now it works flawlessly since 3 years and I use maybe a few hours per year to run updates.
Who's your host? My primary VPS host and my home connection both block the only port I'm supposed to use for mail.
(A curiosity to me, as someone trained on HTTPS, which doesn't really care what port you're coming from. I wish there was just mail-over-HTTPS, since port 443 can't be blocked by a reasonable host.)
A lot of VPS hosts will unblock the port for you on request. I think that is better than open by default as that attracts spammers and gives IPs a bad reputation for future users.
Yeah... I think that one of the biggest "gotchas" with the web3 crowd is misunderstanding different levels of decentralization. Is it architectural decentralization? Institutional decentralization? Geographical decentralization? etc
Can you really say that email is a decentralised protocol that has stood the test of time? It’s more centralised than the web that you complain is dominated by Google, Meta, and Amazon. What proportion of mail isn’t sent or received by Google or Microsoft?
Even if there are 1 million independent email servers, if Google, Microsoft, Amazon, and Apple send, receive, or pass along 98-99% of emails, there's still a fair bit of centralization in the network which should be acknowledged, though this is admittedly not the same as the protocol being inherently centralized.
How do you messure this? Every university in my country has thousands of mail accounts that are constantly used and having worked at an university IT department I have never heard of any numbers collection coming by to ask us how many mails our users sent to each other.
I wouldn't be surprised if that percentage was made up by the big ones to give you the feeling that self hosting mailservers is useless.
You're probably right, I hadn't considered the volume of mail being sent within institutions that might host their own mailservers, so maybe we should say more than like 50-60%.
Is it still a good example of federation though if a vast majority (99%?) of mails between such institutions or outside of orgs large enough to have a campus, tend to go through Microsoft/Google/Apple/Amazon?
But those thousands and thousands of independent email servers mostly just send email to and receive email from Google and Microsoft. And even if they didn’t, compared with the total volume of email, they basically round to zero.
Quick search reveals Gmail and Outlook add up to about 62% of email accounts worldwide. That's a lot, but certainly not enough to make other providers a rounding error.
The problem with web3 is that it's a solution in search of a problem. People have decided they should build something around blockchain instead of starting with what they want to build and why
This is the main issue with it yes. Instead of having a compelling user- and/or howter-centric vision of how Web 3.0 could look and why, it was basically just like: like web but with blockchain and yet somehow more centralized and with changes, because we want to make a new thing.
That isn't how you elegantly solve problems and get people to use your solutions.
Ahh! It says a lot about the state of “Web 3.0” that I forgot that it was even a thing. I was hoping that IEEE Spectrum would cover the actual Web re: the efforts to un-silo it that I reckon will result in more silos with just less volume than before.
Isn't "more silos" the desired outcome, more or less? Has the web been anything but many silos? Of course it might have been different when everyone who wanted to have a presence on it was expected to build their own homepage. Now billions are online, courtesy of their phone, and don't even know what a computer is. So they naturally fall into silos where they and their friends are welcome.
There was a point where "the web" was literally a web, and it was literally "linked" together by...get this: hyperlinks!
I think that what killed that is that hyperlinks work great for browsing & discovery, but as the web matures, a lot of people want to use it for task-oriented things. And all of the Big Tech companies that came afterwards succeeded because they built a task-oriented interface that co-opted the links that were there before and turned them into ways to accomplish the task. Google took hyperlinks, used them to compute PageRank, and then used that to create a better way to solve the task of finding specific information. Facebook took user activity, aggregated it, presented it in a feed, and used it to improve the task of killing time. Amazon took product pages and direct links and used them as lead-gen to increase the ease with which you can buy things. Stripe took embedded Javascript and used it to make paying for things easier; Uber and Lyft took mobile phones and used it to make transportation easier; AirBnB took these large Internet markets and used it to make vacation rentals easier.
Where all these decentralization efforts fall down is that they tend to focus on content, the technical details of how they're going to spread bytes around, and nobody focuses on the task. It'd be interesting to recast the problem in terms of "Here's a common task of everyday life; how do you accomplish it in an adversarial environment where the government or major corporations are trying to shut off the Internet?"
I've become a bit cynic, and I think it will take time until it becomes better. Eventually though, there will be large protests that turn into law that change the dynamics of the now very centralized web.
London was known for having so thick fog from burning coal, and in 1952 it was such a bad event that 4k people died and 100k got badly affected from it. It is was started the Clean Air Act of 1956, that eventually led to clean (cleaner?) air in London.
Another one is Amsterdam. It was a car-centric city up until the 70s, where they started to rebuild it to be a walkable and cyclable city. This started because people protested the dangerous roads, which culminated with the "Stop de Kindermoord" protest.
Maybe if we live long enough, we get to experience a decentralized web again. Next time will come when it's supported by laws, that in turn stand atop a large understanding in the population of why those laws are good and needed.
> Where all these decentralization efforts fall down is that they tend to focus on content, the technical details of how they're going to spread bytes around, and nobody focuses on the task. It'd be interesting to recast the problem in terms of "Here's a common task of everyday life; how do you accomplish it in an adversarial environment where the government or major corporations are trying to shut off the Internet?”
Nice observation and can speak to why anyone intent on selling you content writes a “user story” to outline the tasks that lead to a purchase or engagement.
The cult of marketing ruined it. They're still fantasizing about personalisation, which in marketing speak means optimizing a mattress sale opportunity for Ted Waciznski of 932 West Moreland Ave in Walla Walla, Washington next Tuesday at 3:54pm as he reads CNN.com as he has every morning between 11:19am and 11:26am from a mobile device in the same spot every weekday for the past seven years, rather than, you know, "here's a cool thing you asked us to show you." The only links that matter are to an adtech/martech firm, and presumably the company that wants you to buy something.
Yes, I think so. I agree with you actually. But it doesn’t always seem like that’s the desired outcome for some people when they mention “decentralization”.
I have a particular interest group in mind here, who lament the current state of the Web to an extent unique to our time. It seems like they’re upset with their loss of agency/authority online. When a free, ad-less *.blogspot.com or LiveJournal presence could net you some clout you’d be precluded from otherwise in the real world.
Take for example any evidence of disapproval you may find from the proponents of ActivityPub toward the development of the AT Protocol in parallel. Also notice how moderation features on these federated platforms essentially centralize them and their respective enclaves. The desire for interoperability between social media platforms conflicts with the imminent re-siloing being examined. Or is it just the same ends with new wiry means attached?
“Decentralization” in this context seems more like separating from the whole with the hopes of replacing it. And I’m not confident that these many new “wholes” can resist jockeying one after the other for political influence.
"web3" is almost a blast from the past though, simpler times. Seeing it mentioned on HN now, I just had to take a look at https://www.aiisgoinggreat.com to see if it was available, it is not. Nothing there, but I'd love a similar site about AI debacles.
Wasn't Project Xanadu just HTML with backlinks? I was just thinking about it in the shower yesterday and realized I had no idea how it could have possibly worked for cross-site links, unless the web was either fully centralized, or fully distributed.
It was/is an pipe dream. The whole schtick was that links were visualized so you could, "trace" what you were researching visually without having to rely on jump links. In theory (more pipe dream) the whole topographical link structure of the internet would be capable of pretty easily being visualized by a person.
In my somewhat educated opinion, David Chaum is the OG inventor of crypto currency aka digital cash (eCash), back all the way in 1982.
He was last seen pitching family offices in the UAE pitching AstroCool, a concept involving spraying moon dust in orbit around the earth to block some of the sun light. You won't see this referenced anywhere. I was told about it by the person who introduced him to family offices over there.
Weirdly enough, in 2008, I was so pissed about the financial crisis, I went online to promote the idea of an absolute-valued energy-exchange-based currency called the Joule Token. It had no cryptographic component (was a centralized immutable store that is co-owned and co-administered by the community) The banner I promoted it under was "P2P Programmable Currency" and I kind of spammed Michel Bowen's P2P Research mailing list and wrote a giant post on the P2P Foundation Wiki titled the P2P Energy Economy. The idea was to decouple profit/interest from exchange, so you can value things based on the energy it took to create them and transport them, and instead of profit as an incentive it was about fulfilling the need for energy: you sold things because you needed to recoup the energy and you created the good that you sold because you could, no other reason. Obviously, it was a thought experiment, not an actual implementation or anything meant to becoone real. However, after I was kicked off the list for spamming (or rather for list etiquette -- Hi Bryan! if you still frequent HN lol), Satoshi@gmx.de replied to my emails 4 times between early 2009 and late 2010, so I literally ghosted him, but without knowing. I didn't know until Bryan Bishop, who was behind my banning, I think, told me here on HN 3 years later, in 2013, if I recall correctly.
Here is what ChatGPT knows about those times (I asked it to tell me what it knew without searching online)
I think a different type of decentralized web is growing. It's in Hetzner datacenters but it's a lot more "free". We're at an age where besides GPU inference an indie can afford to create cutting edge web services. There's still players that could really harm this new ecosystem (looking at you cloudflare) but the fact that the majors like Google and Facebook tried to kill it and failed to kill it proves something.
I'm still not a fan of how 'web3' has become a buzzword for crypto. The name seems like it should describe a much more interesting and general concept, not just "it's a service, except you can make money/sell NFTs/whatever".
But I'm not sure I'd write the decentralised web off yet. A lot of the issues it's faced are UX and design problems more than fundamental issues, and the way the world is going makes it seem like a comeback could be inevitable at some point.
After all, a good federated or decentralised system could be resistant to the age gating laws in places like the UK and Australia at the moment, and be a lot more tolerant of content that's critical of those in power than Twitter or Reddit would be (like with criticism of Israel and the goings on in Gaza, criticism of Trump and his administration, protests against authoritarian regimes worldwide, etc).
Large social media services are very willing to bend the knee to those with power (both political and corporate), while a more decentralised service may very well not be, especially if either the people using it are hard to identify or the infrastructure is spread across the world/in places with much more lax laws.
I would really love for new projects to find their own name instead of calling themselves the next version of something they don't like. "Web3" has never been a new version of the web, just a way to inject undeserved clout into a (IMHO fundamentally flawed) experiment.
There are so many great unused names! Please pick one.
a colleague of mine recently insisted that bittorrent be included in web3 techs since it uses hashing. if we toss torrents under the title of web3 then it's been a huge success.
atproto, however, could learn a lot from torrents and all of the other protos since then. for example, their recent push to centralize bookmarks.
> for example, their recent push to centralize bookmarks.
It's worth noting that the current bookmarks implementation is "temporary" and will be moved to the PDS as soon as private data repos are supported on the PDS.
The approach they are taking (and are recommending to other devs in the space) is "develop your private data as if it was stored on the PDS but stored on the appview. then migrate to the PDS once private data launches". If you maintain the same lexicon/schema under the hood, you can just copy the data to migrate it.
Private data is unfortunately really hard to do right so this "start workshopping it and get ready for us to roll support out" approach seems like the best temporary solution.
> a colleague of mine recently insisted that bittorrent be included in web3 techs since it uses hashing
Related pet-peeve: Folks who say "there's still great promise in private blockchains." This is equivalent to saying that self-balancing Segway-style devices can still become the dominant mode of transportation if we juuuuust make them bigger, enclosed, and add another pair of wheels so that they don't have to self-balance anymore.
Yea, the bookmarks thing is interesting, and wearing my conspiracy theory hat (it's not the red one), I wonder if it's not an experiment to see what the people will tolerate or how they will respond. It could equally be attributed to an early win for the new product manager, finally delivering a long requested feature in a POC form factor, while private data gets figured out.
I'm highly involved in the private data work and we'll have a better bookmarks in the long run
I like that atprotocol sits in the middle of web 2 & 3, ideas from both without being beholden to either
I'm pretty sure the private data model is going to be clearly defined by EOY, between the community group work and bluesky internal work. We also made sure that bookmarks will migrate cleanly to protocol private data when it lands, so no need for the tinfoil hat. Your other explanation is pretty close to the mark.
EDIT: also fwiw, I got my start in 2012 working on ssb and then in 2013 on dat, both of which were conceptualized as bittorrent variants. Atproto has a pretty clear lineage from that p2p work. You just can't do large scale social computing on user devices.
yes, and we are using computers in our pockets now that are way more powerful than the ones we had when we were working on ssb. local indexes are possible, you and dom showed that back then.
ssb had private bookmarks, let's dig into how those were enabled
Is there a way to use it without making an account? I really want a permission-less way for two browser clients to send a simple message on the public web.
And, then Hypercore, Hyperbee, Hyperdrive, Autobase, etc are all modules you need to get things done so you can make an app like their serverless Zoom clone https://keet.io/
I wish one of these networks could be the go-to clear default choice. Options are good, but it would be great to have a best option for a large discoverable community. Actually, their default choice Nostr sounds pretty awesome https://nostr.com/
Little known fact: Starlink should be a pure orbital internet backbone to communicate anywhere in the world, right? Because of the laser links between satellites?
Not so much. All the traffic from the US must transit through an earthbound surveillance hub, so the latency advantages are lost. Same with other countries.
(Please anyone correct me if I'm wrong about this, I was gravely disappointed to hear it and would love to hear that I'm wrong)
Gotta keep in mind that PRISM (or maybe something like it) is still a thing. If US traffic isn't sniffed by Starlink itself, then it's happening at the next hop in the nearest terrestrial network provider.
The term "web3" was coined in 2014 by Ethereum co-founder Gavin Wood, and the idea gained interest in 2021 from cryptocurrency enthusiasts, large technology companies, and venture capital firms.
i don't get it. is the fact that somebody misspelled a recently-coined nonsense-word product name/trademark an indication that the broader points are invalid?
The core tenets of web3; privacy, data sovereignty, encryption, open source, open protocols, and peer-to-peer networking are all good. The problem is the movement was never truly about this, and, even to the degree it was, it was taken over by crypto grifters. The ultimate indictment of web3 is that the old school nerds weren't interested, they are the perfect audience, and yet...
meh. "cryptographer" usually means someone who practices "cryptography," [or "cryptology" if you live near Baltimore] not someone who participates in "cryptocurrency" transactions.
As goes capital, so goes the web. Capital is becoming more concentrated, and so power over the internet becomes more concentrated. Because at the end of the day, the web serves capital, not individuals. And AI represents the ultimate centralization of information and power.
The failure of Mastodon to break into the mainstream especially in the wake of Twitter/X's debacles is a good example on how decentralized web is a good idea on paper but won't work in practice. It turns out that a) people will just join the largest instance/community possible for network effects b) internet discourse is bifuricated enough that preemtively banning instances just in case makes discourse even more bifuricated and c) maintaining and moderating said instances is expensive and time-consuming.
Even though Bluesky did break through to the mainstream with ATProto, it's unclear how ATProto is either a functional or competitive benefit, or if its users even know about it.
Mastodon might not have broken into the mainstream, but it's surprisingly viable. I am totally OK with its current state to be honest - feels like the cozy internet of 2000s.
This is, at least in part, a UX problem. Even as someone with a lot of technical experience I found Mastodon quite disorienting at first. Bluesky has solved this much better which is why they've won out over Mastodon.
It is easier, but it's still possible to solve the UX issues with a decentralized service, in my opinion. What I think is the main issue is that these decentralized services are made by programmers with little regard or intuition for UX, and there is also a lack of funding to work on UX problems.
I agree the UX challenge is much more challenging for decentralised services. I don't know enough about Bluesky to really comment on whether it is centralised or not.
Regardless, I think there's another thing that helped Bluesky: VC capital. In particular, to hire people to work on UX. It's a bit of a pet-peeve of mine, but I find it strange that designer don't contribute more to projects like Mastodon, which definitely need it. Even from the selfish angle of building a portfolio, helping solve Mastodon's UX challenges is much more impressive and realistic, than doing the millionth redesign of Gmail that will never get implemented.
People are running every element of the Bluesky / AtProtocol stack independently. Bluesky could disappear and it would continue to function as is (albeit with lots of Bluesky data lost).
PDS's to hold users data, relays/firehoses to aggregate & forward traffic, AppViews to create composite views of likes, replies, etc, resolvers to lookup DIDs, clients to access the network. Each of these has independent implementations. BlueSky is already decentralized & already has viable credible exit. It's not decentralized, and indeed the scalability & accessibility of having firehose consumers has the greatest scale out decentralization characteristics we've seen anywhere short of BitTorrent.
I don't know why you're being downvoted. I tried it the other day and this is indeed how it works. I could even see my home traffic rise and fall over the course of the day in time with activity on the network.
It seems quite reliable that Bluesky gets down voted. Whether it's mastodon/Fediverse folk or right wing pro twitter philes or both, I dunno, but it sucks to have aggressors out there, dark forest freaks sniping away.
I try very hard to find the positive & to upvote things I don't fully agree with, if well argued. I wish the social network of HN could do more against adversarial zero-sim thinking, didn't have people who insist on draining.
People keep thinking decentralization fails for technical reasons, when it’s economic. Without an economic model nobody can afford the massive effort required to make software polished and usable.
As a general rule I’d say polished user friendly software takes 10X the effort at a minimum vs software only nerds can use. That’s probably an underestimation. For consumer software it’s probably 100X. It’s because computers are incredibly confusing and hard to use and it takes immense effort to overcome that.
(If you don’t agree that computers are confusing and hard to use, you are part of a tiny highly educated minority.)
I’ve said this around here like a hundred times because it feels like I am the only one who gets it.
The tech to create an excellent decentralized network with an excellent experience and without the weaknesses of Mastodon exists. It doesn’t get used because nobody pays for it. Centralization gives you an economic model from either directly charging for access or selling data or ads, and none of that works in a decentralized world.
Right, but that's the entire point: Mastodon's UX problems are caused by its decentralization and mostly cannot be separated from it. Arguably all the problems of decentralization that make users disprefer it are UX problems-- that doesn't mean they are easy to solve.
I agree completely. This isn't the world of BBSes where everyone had a corner or two that they hung out on. As long as posts are gamified with likes and notifications, decentralization will never win. Even HN isn't immune and it has carved out a pretty specific niche of text based and no ads social network focused largely on tech. I don't think another decen-HN is likely to take off or break into other self organizing segments even though this is easy to attempt with some Mastodon servers.
> This isn't the world of BBSes where everyone had a corner or two that they hung out on.
That's exactly what Mastodon is. The whole reason mastodon exists is because a whole lot of people wanted exactly this experience.
Mastodon won't win because they aren't playing. Which again, is the whole point. It's extremely deliberate and intentional. People use mastodon because the "game" is abhorrent and objectively extremely bad for people individually and society as a whole.
So mastodon eschews the gamification and like farming and fake notifications. One of the most common jokes throughout the whole network is "wow, this post blew up and is doing Real Numbers" and the post has like ten boosts and twenty likes. It's extremely rare to see more than a couple hundred notes on any one post.
It's on purpose. Mastodon users like and actively want it to be this way.
>This isn't the world of BBSes where everyone had a corner or two that they hung out on.
Yes it is. It literally is. If we want to do that we can just do that. The internet is not a zero-sum game. We don't have to play to "win."
>I don't think another decen-HN is likely to take off or break into other self organizing segments even though this is easy to attempt with some Mastodon servers.
The only reason HN is popular is its connections to Silicon Valley and startup culture. It didn't really carve out a niche so much as succesfully create and market an image. Without that it isn't anything remarkable, and there are already tons of technically-focused communities on the fediverse.
Discoverability was an anti-goal for Mastodon initially. People from queer community felt it enables bullying. Mastodon has added search functionality since then, but it's still pretty basic.
IMO it's not necessarily bad though. It prevents the network from becoming centered around content and keeps it social. You discover new accounts as they get reposted by people you are already connected to
I feel like this is a clear case for "security through obscurity doesn't work" saying, and for the same reasons as the original meaning. "Normal" people won't find such community because it is pain in the ass to do, while bullies will get there just fine because they are motivated. It is like reverse moderation, filtering out less radical and less pro-active newcomers.
This attitude is weird to see on Hacker News of all places, a forum that considers itself a quarantine zone from the rest of the world and the modern web, and which is also irrelevant to almost everyone, including much of tech.
I'll post this every time mastodon comes up: to compare mastodon to Twitter, Facebook, bsky, etc. is fallacious in the extreme. Saying that mastodon is "losing" to bluesky is a complete non-sequitur.
Mastodon hasn't lost to these other services because they aren't competing. Nobody on mastodon actually wants to replace Twitter. Mastodon exists explicitly as a rejection of mainstream social media. Calling this "losing" is like saying that a bunch of kids playing basketball in the street are "losing" to pro hockey teams. Not the same game, not the same scale, the only similarity is that they're both playing a sport.
Mastodon doesn't want to compete. That's totally counter to the entire philosophy. Mastodon isn't even remotely playing the same game as twitter and Facebook. It just so happens that mastodon has a text box with limited characters. That's it, that's the only similarity.
Of course a chess club of 10 year olds in Japan isn't going to "win" and displace the entire American baseball league. You'd be laughed down for even suggesting that. And yet, people think that mastodon is "losing".
Mastodon is exactly where it wants to be. The small scale and diffusion are features, not bugs.
Besides that, it's functionally impossible for "mastodon" to go mainstream. To suggest that is a fundamental misunderstanding of what mastodon is, how it works, and how the people involved run it. At absolute most, you could get a few big servers to go 'mainstream'.
When a server gets too big or too popular, the network treats it as damage and routes around. A lot of admins do this purely on principle: huge servers are antithetical to the philosophy and actively bad for the network. The rest block them because the vast majority of attacks and spam originate from such servers.
If mastodon went mainstream, the network would fracture (again) and the majority of nodes would re-form the network in the obscurity they enjoy and value.
This is all extremely intentional and deliberate. Mastodon is an explicit rejection of all the things you think it's "losing" at. Mastodon doesn't want to "win", they largely want to be left alone to continue enjoying a small internet with limited connections.
Mastodon has won its own game. It's firmly established itself in a niche, and is not going to move from that niche to chase fads, engagement, metrics, profit. Mastodon wants to be exactly what it is because mainstream social media is a very bad thing
I just googled "science mastodon" and found plenty of results, including this reddit thread with some listed[0] and this list of academics on Mastodon[1]
Although, of course, you can follow any account you like from whatever instance you choose to use (assuming that instance doesn't block it.) A lot of people you may be interested in are on universeodon.com, although typically you'll discover things on Mastodon by searching hashtags, not people.
Discord servers are half of a good counterexample, since people try to avoid the extreme end of network effects and want their chat partners pre-screened.
But... it's all on Discord, not IRC or jabber or anything.
Now that you mention it, I vagely remember something about Bluesky being built on some distributed protocol. So yeah, I guess it's correct that doesn't mean much to it's popularity. Which is sad of course. But yeah, most people don't know about technology, and security and privacy, so yeah, why would they care..
I think there's an element of how decentralized services work that's not sufficiently refined yet. In a lot of ways it's a better way to have a good user experience with decentralized data that we're sort of looking for, and in some ways, it doesn't really fit how the web works.
yes, lots of stuff needs to be figured out. Like, you don't need decentralized auth servers.
bittorrent has webseeds. I dump large files and archives on my sever, give people the magnet or link it from an article, they download it fromthere. If for any reason I no longer want to host the file I just remove it. Usually I just want the space back, in theory there could be bandwidth issues or an excess of interested participants. You can put fairly interesting websites in torrents too! That no uninvited participant can find them is a feature not a bug.
Email does html too! PDF is also great for publishing. You can send people videos and even executables if you are close enough.
All of these decentralized social media concepts were/are too technically complicated for too long during the Twitter debacle. Even now, the top results for "how to use Mastodon" have jargon incompatible with mainstream usage.
If decentralized is great on paper and not in real life, then the paper is wrong and thus not great. Additionally, if decentralized doesn’t work, please explain the survival of email.
d) The kind of people who put a little thought into how Twitter could be done differently may come to the conclusion that they don't actually need a Twitter in their lives
Mastodon would be OK but the UI/UX is just terrible... interacting with someon on different instance is utterly cumbersome and I do/could miss a lot of posting in global view if someone didn't already followed someone from other instance...
Terrible compared to what? Mastodon is just fine for me. I'm an oblivious nerd here I guess. But I sometimes use Twitter too, can't say I find the UX refreshing by comparison.
For example bluesky. I tried masto a bunch of times, still have the account but each time I try to use it, it's so cumbersome. I tried alternative UIs but they are no that better...
The entire point and purpose of mastodon is to not be mainstream social media. Almost everyone there signed up because mainstream social media is so awful. The small and diffuse nature is one of the main benefits. Because it's distributed, there's an extremely niche community server for any subject you might be interested in.
Mastodon captures perfectly what twitter was at the beginning. Just people talking to each other and having fun. Not performative like fishing or whatever. No ads, algorithms, endless feeds, AI slop, or spying and tracking.
It turns out that a lot of people think mainstream social media is objectively abhorrent and want to connect with other humans in a more natural and user-driven way.
Frankly, "mainstream" is a dirty, disgusting concept here. A very large fraction of users would put a significant amout of effort to prevent that from happening. If one server did become "mainstream", as mastodon dot social did, the network treats it as a damage and routes around. Many servers just cut them off because the vast majority of attacks and spam originate from the big public servers.
Mastodon won't go mainstream because they don't want to and the system fundamentally cannot operate that way. A few individual servers may go mainstream, but we'd eventually consider it a hard fork. The network would fracture (as it has many times) and the network of small servers will go back to the obscurity we enjoy and cherish.
Case in point: Facebook tried to force their way into the fediverse network through threads or whatever it was called. There was a pretty hard split in the network as nodes that value privacy and safety cut themselves off from a literal hostile invader. Many servers went recursively through the network to cut off any servers who hadn't blocked Facebook. I haven't heard anything about Facebook trying activitypub again, so I guess the quarantine was effective.
> Many servers went recursively through the network to cut off any servers who hadn't blocked Facebook
A big problem is that this is happening with other servers as well. I've seen nodes that block all other nodes that run Pleroma because it's "for Nazis". And some of them also block nodes that don't share most of the blocklist.
The simplest way to answer this is to ask "whose dream?".
We have billions of data points conclusively showing that users don't care about nerd fantasies. No one cares if it's decentralized or not, if it's open source or not, if it's patent-encumbered or not. Maybe there are good philosophical reasons why people should care, but they don't.
So you can keep digging your hole, or you can build products that are simply good and happen to embrace your philosophy without that getting in the way of usability.
In this particular case, the dream the article is talking to was tainted even for the nerd audiences because of the sleaziness of a lot of what was happening under the banner of web 3.0.
From The Truth by Terry Pratchett, on a similar dilemma between pursuing the (journalistic) public-good versus satisfying the immediate desire of the average public:
> "Are you saying people aren't interested in the truth?"
> "Listen, what's true to a lot of people is that they need the money for the rent by the end of the week [...] This is a report of the annual meeting of the Ankh-Morpork Caged Birds Society [...] They've got no say in who runs the city but they can damn well see to it that cockatoos aren't lumped in with parrots. It's not their fault. It's just how things are."
> [...] "It's important! Someone has to care about the... the big truth. [...] if they don't care about anything much beyond things that go squawk in cages then one day there'll be someone in charge of this place who'll make them choke on their own budgies. You want that to happen?"
____
Our civilization is filled with uncountable things that were once "nerd fantasies" that the average person doesn't even remotely care about... Until it all goes tragically wrong.
If I'm being honest, your quote makes the first character sound like an out-of-touch loonie. It doesn't make the point you're trying to make. Maybe in its original context it did, but the way you've presented it, it's like going to an AA meeting with a large binder, intending to expose corruption in the senate. There is such a thing as time and place.
The context is that the Lawful-Neutral leader of the city-state has been framed and thrown in the dungeons by a shadowy conspiracy of powerful figures. The printing press was just recently invented the first character accidentally inventing Journalism wants to break the story while a coworker is arguing that it's not relevant to the citizens.
Expanded portion:
> "Someone has to care about the... the big truth. What Vetinari mostly does not do is a lot of harm. We’ve had rulers who were completely crazy and very, very nasty. And it wasn’t that long ago, either. Vetinari might not be ‘a very nice man,’ but I had breakfast today with someone who'd be a lot worse if he ran the city, and there are lots more like him. And what’s happening now is wrong. And as for your damn parrot fanciers [...]"
____
With respect to "we've had rulers", a bit from a previous book Men At Arms:
> "[...] He wielded the axe, you know. No-one else'd do it. It was a king's neck, after all. Kings are," he spat the word, "special. Even after they'd seen the... private rooms, and cleaned up the... bits. Even then. No-one'd clean up the world. But he took the axe and cursed them all and did it."
> "What king was it?" said Carrot.
> "Lorenzo the Kind," said Vimes, distantly.
> "I've seen his picture in the palace museum," said Carrot. "A fat old man. Surrounded by lots of children."
> "Oh yes," said Vimes, carefully. "He was very fond of children."
Yeah, I got that part. I took at face value that whatever the character's concerns were, they were legitimate. Even granting that much he still sounds silly, because he's trying to publish his findings at "the annual meeting of the Ankh-Morpork Caged Birds Society". Like I said, time and place. Of course it would not be welcome at such a venue; it's off-topic discussion. It would be like exposing a CP ring through a post in a programming forum. It'd just get your thread locked.
Trying to commandeer the attention of the attendees like that is simply disrespectful; it doesn't matter how important you think what you have to say is. Unless the building is on fire or there's an armed squad standing outside, whatever he had to say could have waited until the end, and whoever wanted to listen to it could have stuck around, and whoever didn't could leave.
> he's trying to publish his findings at "the annual meeting of the Ankh-Morpork Caged Birds Society". [...] Trying to commandeer the attention of the attendees like that is simply disrespectful;
Ah, I see the confusion, no, these are two people running an actual newspaper.
When the coworker says "this is a report" it's because they're holding it up in their hands to display it, as an example of something their average working-class reader might care more about than estoeric politics.
Alas, the 2-hour edit window is closed, but:
> She held up a piece of lined paper, crammed edge to edge with the careful looped handwriting of someone for whom holding a pen was not a familiar activity. "This is a report of the annual meeting of the Ankh-Morpork Caged Birds Society," she said. "They're just ordinary people who breed canaries and things as a hobby. Their chairman lives next door to me, which is why he gave me this. This stuff is important to him! My goodness, but it's dull. It's all about Best of Breed and some changes in the rules about parrots which they argued about for two hours. But the people who were arguing were people who mostly spend their day mincing meat or sawing wood and basically leading little lives that are controlled by other people, do you see? They've got no say in who runs the city but they can damn well see to it that cockatoos aren't lumped in with parrots. It's not their fault. It's just how things are. Why are you sitting there with your mouth open like that?"
As a side note, I find it really interesting and a bit disheartening to follow the history of “web 3”. The original “web 3” was a beautiful vision including the semantic web and its cousins. The term was hijacked by cryptobros through what could have been a genuine or malicious misunderstanding of the original web 3
There's a general phenomenon where terms get annexed, not by meaning that makes most sense, but by meaning that is most salient, most eye catching or emotionally riveting. This leads to Flanderization of many many important cultural concepts.
I fear this is the reason why intellectual discourse seems to have stagnated. We no longer pursue the spirit of important ideas but try to endlessly put them in uninteresting boxes.
Weird. The first time I came across "web3" was as an interface with Ethereum nodes. This is my first time hearing about the term ever having an earlier meaning.
> We have billions of data points conclusively showing that users don't care about nerd fantasies. No one cares if it's decentralized or not, if it's open source or not, if it's patent-encumbered or not. Maybe there are good philosophical reasons why people should care, but they don't.
This is focused on the technical aspects as if this is why people go to the big platforms. It is not the why for where they are or where they will go.
(1) The big platforms came first, most people don't know there are different ways for humans to organize social media.
(2) They have desires to leave for greener pastures, they know these companies don't have their best interest in mind.
We need a viable alternative and ATProto is in the best position right now. I have been able to get people excited and even create an account, knowing that the better way forward is still very young and a work in progress. I talk to them about features at the conceptual level, never about federation or any of that gobbledygook that they don't understand or want to
In the x86 days p2p filesharing applications unambiguously proved that people preferred a decentralized web, not only were they getting their music for free, they discovered new and hardly known niche artists in genre's they liked (some of the p2p software allowed checking out other music from the same seeder).
What really happens is the transition to ARM platforms (tablets, smartphones) with TrustZone technology, and the resulting lack of FOSS Operating Systems & Software being installed by the user, and a lack of development because average user can't install it either.
Perhaps one could argue that at world population scale people couldn't afford both an x86 system AND a smartphone, and preferred the latter, at the cost of user freedom.
>We have billions of data points conclusively showing that users don't care about nerd fantasies.
That's 100% BS. There is no such data or research. All the 'users don't care about x' comments all hold about as much water as if I were to claim `North Koreans don't care about basic freedoms`.
Everyone called us paranoid nerds and laughed at our concerns. We spent 20 years trying to tell the people we care about how to be safe and they didn't listen because the tech is ugly, with rough edges, and doesn't have rage bait and trillions of dollars of ad-tech money supporting it. I personally stopped advocating as loudly because I realized that it would take much, much more pain for regular people to get out of their comfort stupors and start giving a shit. Meanwhile, p2p tech slowly progresses in the shadows where the mainstream can't see it, can't support it, and doesn't care.
The best time for advocacy is when people get burned. It's not a matter of if, but when. People learn through pain.
New billionaire owner takes over your favorite microblogging platform? Pain. Your clubhouse has been destroyed. You suddenly realize you didn't own anything.
New billionaire owner decides they don't like the topics you talk about and bans you? Pain. Your ability to express yourself and communicate and make connections with like-minded people has been destroyed. You suddenly realize that you actually need to get permission to talk to the people you want to talk to.
Massive data leak that includes your intimate personal information, photos and videos? Pain. Your privacy and personal life have been irrevocably violated. You suddenly realize that you don't control any of your own private information, and it actually isn't safe out there in those clouds. Even your most intimate whispers are recorded forever and will eventually be listened to by strangers or published on the cover of the New York Times.
The list of reasons to get the fuck off these corporate, centralized apps is growing every day.
I'm even more demoralized than you, as I feel the problem with "the best time for advocacy is when people get burned" only manages to convert a handful of people and then a month later more new people have entered the system to replace them who know nothing about this and will adamantly claim that such a problem is impossible until it happens personally to them, which takes years and years, at which point the same thing repeats.
I understand the feeling of demoralization, but converting a handful of people is better than converting no one. Those people are now safer because of you, and that's something to be proud of.
We're hanging out on the decentralized web and/or private networks.
Sure it's not the same size as the "real web" but it's got no ads, less spam, less bots, and enough to people to blow out your dunbar ring and fill your cup.
You might never get to be a freelance marketer drop shipping some stupid crap - but that's a price some are willing to pay.
For us who remember the 20th Century? Yes, I agree and we can see the causes. For the next generation? It's more that they're wrapped up in centralized closed platforms for their hobbies. Most of them never considered anything else. They were born deep underwater and have no idea why they can't breathe.
We had BBSes, forums and IRC. Now it's Discord.
We had LAN parties. Now it's publisher-controlled match lobbies.
We had RSS feeds. Now it's curated social algos.
We had myriad self-hosted wikis and homespun fansites. Now it's all ad-choked Fandom.
For Discord and maybe some others, I wonder if it’s just a packaging and marketing problem that just hasn’t been solved.
Yes there’s Matrix, but it’s not packaged in a way that makes hosting a node accessible and the way it’s marketed, there’s no clear benefit to typical users.
Imagine if instead, the client and server were one in the same and starting a server is as simple as running the client and clicking “new server”, with the software figuring the rest out. Then, on the marketing side you can sell it as a way to get features that Discord puts behind a paywall for free.
There’s no way that wouldn’t have a dramatic impact on take-up.
This feels very strange: the clients on that page span a huge range of maturity and capability. If you put every email client or every web browser on a single page you’d get a crazy mix too - does that mean that email & the web have failed? Just the opposite, surely.
If you want to narrow it down to one or two really great ones which are well polished, supported and marketed just pick Element X, Beeper or maybe Fluffychat?
When I onboard users to my Matrix homeserver, I point them to a preconfigured Element-Web URL and the Element mobile apps. I also mention that there are other clients out there - a handful experiment with them, most don't.
One discovery problem is this client had three rebrandings, from Vector to Riot to Element. I've noticed users have had a hard time realizing the Element is a client for Matrix - even when they're actively chatting on Element. Usually they just refer to our chat by the homeserver's name.
> One discovery problem is this client had three rebrandings, from Vector to Riot to Element. I've noticed users have had a hard time realizing the Element is a client for Matrix - even when they're actively chatting on Element. Usually they just refer to our chat by the homeserver's name.
What this tells me is that having a canonical primary client with the same name as the protocol would do wonders. This doesn’t rule out third party clients, it just clarifies matters for users who don’t necessarily know what a client is and have trouble conceptualizing the protocol/service/client divide. Those who are technical enough will seek out their favorite client, but for everybody else chatting on Matrix means downloading the Matrix app.
So I've hosted a federated Matrix homeserver for almost eight years now, with a few dozen active users. Beyond the 800 pound gorillas of networking effects and the need to be your own infra engineer (mitigated by things like your integrated server/client idea), there are a few key features that I feel keep people on Discord:
- No distinct voice channels with push-to-talk. Yes there's Jitsi, but it's not the same as the TeamSpeak/Ventrilo/Mumble style functionality that gamers have used for 20+ years. Solve this, and I think Matrix will see much more adoption for casual voice chat in the same manner as Discord.
- No server-specific display names. I have multiple screen names across multiple games/communities, and Discord accommodates that. Matrix (or at least Element+Synapse) does not.
- No path for server-specific invites. If I attempt to invite someone to myexamplehomeserver.net in the Element UI, they are instead directed to make a matrix.org account and are federated in. It works I guess, but it's shit for homeservers with closed registration that want to invite users to that homeserver. Writing an invite bot has been on my "I'll get around to it" list for a very long time now, and I know other homeserver admins have considered the same thing.
- E2EE is clunky. Yes, it is much better now. Yes, there are concerted efforts to improve this. Yes, it is an extremely hard problem. However, many times I have had users lose access to past messages because they were signed in to only one device and don't remember their keys. I understand the security aspirations, the proles do not and never will.
There are a lot of other little things like custom reactions, but I think the four above would do a lot to foster more Matrix adoption.
Mattermost is probably an easier self-hosted Discord alternative.
For $2/month you can spin up an instance on Pikapods, based on their community edition. If you want advanced features (available in enterprise editions) you'll need to figure out how to re-enable them by un-feature-flag-gate-ing the source code and recompiling (and at that point move hosting), or work it into the plugin system (which gives you tons of power)
OK, I've investigated Matrix enough to suspect it's not going to work for the things I've been using Mattermost for, but I'm curious to hear what your stack is (client and server software)?
It sounds like, if Matrix worked seamlessly (which is not the experience of... anyone(?)... in this recent thread https://news.ycombinator.com/item?id=44617309 ) it would give us federation and better identity access management (compared to the Mattermost community edition anyway) which would be really nice
It also has video calls built in... also really nice, but people are saying since they migrated from jitsi to their MatrixRTC, setup & admin have been a nightmare.
Element X seems to be the more popular client, though it sounds like, while faster than Element, many features (such as threading) are not yet available, and people have been lamenting the extreme sluggishness of Element
It is absolutely possible to have a Matrix deployment that works seamlessly. It's true though that a lot of people are very unhappy at the transition from Element to Element X, hence the recent complaining threads. You can see a response from me as Matrix project lead (and Element CEO) here: https://gist.github.com/ara4n/190ad712965d0f06e17f508d1a45b5.... Meanwhile, we're working hard at landing both threads and spaces in Element X (they're available behind feature flags already).
The only reason people seem to complain about Element Call being hard to set up is that it's new and unfamiliar and doesn't have much docs yet, and VoIP is always a pain from a firewalling perspective. It's a hell of a lot simpler to setup and run than Jitsi, for instance. But if you run it via ESS Community (https://element.io/server-suite/community) or even https://github.com/element-hq/element-docker-demo it should work fine.
Element X shines with large rooms - think joining rooms with 100+ active users and a lot of events to sync. Using multiple clients is not a problem either, I think I have at least 10 active sessions active with my current account. Element largely "just works" for my users with few grumbles, but that is definitely not the case for a lot of others here.
Damn right. I don't hate tech, I fucking love it and I always have. I have a career because of tech. I have tons of friends I met online and continue talking to via tech. I hate tech companies because of what they did to tech.
We need more democratic institutions revolving around tech, too many rich individuals have an extreme amount of control in what technology gets pursued and pushed.
But this is where the disillusionment of this idol continues:
Do you think kids wouldn't develop the same mental health issues if they were on Mastodon instead of Facebook?
Do you think the largest Mastodon instances wouldn't start also harming their users once large enough?
There's nothing preventing the decentralized from becoming centralized over time, from the same network effects, and abusing the users all over again.
In which case, why fight for it? It's pointless; we started decentralized and became centralized; a re-decentralization moment just causes the same economic forces to pull everything together again.
>> In which case, why fight for it? It's pointless.
I fought for years and came to the same conclusion. I just focus on myself now and a few tight knit circles of friends. We still hang out together on Diaspora. We still muck and hack around with stuff like Ubuntu Touch and debate the latest Linux distros and see how long we can go without our smartphones.
A few of my friends have gone "analog" because of what's happened. They were there during the first dot com bust. They've seen what the internet and tech have become - they just want to opt out now. Kind of crazy the times we're living in. Decentralization was the dream - now its just a nightmare a lot of us no longer what to be apart of.
> Do you think kids wouldn't develop the same mental health issues if they were on Mastodon instead of Facebook?
I'm not sure, but Facebook has employed too many PhDs to name with the goal of fostering addiction to their product, as well as behavioral ad targeting, and to my knowledge, Mastadon has not.
> Do you think the largest Mastodon instances wouldn't start also harming their users once large enough?
I don't think user harm is correlated with size of the platform.
> There's nothing preventing the decentralized from becoming centralized over time, from the same network effects, and abusing the users all over again.
Which is why I support things like the AT Protocol, which enable decentralized social networks that can share data amongst each other, and permit users greater control over what they see and from where.
> In which case, why fight for it?
Because I've been fighting for a better internet for decades at this point and see nothing better to do with that time were I to stop.
Because according to UK research, just doing it once, in a supposedly safe way, causes biomarkers indicating brain damage in the blood. It can also apparently cause strokes decades afterwards.
“There’s no safe way to do it, no safe quantity of blood or oxygen you can cut off from her brain for fun,” says Jane Meyrick, a chartered health psychologist who leads work on sexual health at the University of the West of England. She describes being at a sexual health conference last year where data was presented on sexual strangulation – the prevalence and harms. “Usually, at those conferences, people will be talking about the extremes of what everyone is getting up to in a very sex-positive way,” she says. “When this was presented, you could feel the tension, the internal conflict, in the room, with professionals being unable to reconcile the gap between what they were hearing and their usual sex-positivity.”
Well shit, that's really unfortunate for a lot of people.
...but sexytime "choking" has an entire range of intensity that is more about intimacy and physical touch and doesn't restrict airflow or blood flow at all.
If a 14 year old watches porn and decides to copy what they see, that is a failure of sex education. Every child needs to be taught fairly graphic and frank realities of sex, sexuality, and sexual activity.
For some reason, this is basically only difficult in parts of the US.
Did you miss the part about how this affecting the UK despite educational warnings? This is hardly US-exclusive, and education is not solving the problem.
> If a 14 year old watches porn
This is not 14 year olds either, these are teenagers and fully grown consenting adults, notably not in the US.
"A survey by the Institute for Addressing Strangulation, established with Home Office funding in 2022, after strangulation became a standalone offence, found over a third of 16 to 34-year-olds had experienced this."
The reality is that people are influenced by pornography, despite education, despite warnings, despite common sense, despite age, despite region, and it is killing people. Even sex-positive educators are stuck in cognitive dissonance on this one.
Dang someone had that response locked and loaded, 2 minutes later... in any case I did not expect this entire sub-thread under a discussion on decentralized web!
If you’re going to insinuate on Hacker News, that pornography is inspiring behaviors that are literally killing people… you need to bring receipts. Being in denial of harm is the favorite pastime here, but that's just the first stage of grief.
Case in point, the sex-positive educators in the UK... forced to say that certain bedroom behaviors actually are too dangerous to participate in. Even that, yes, we will control what you do in the bedroom, with the force of law, and even restrict depicting that act regardless of consent, with the force of law. Extremely ironic that sex-positive educators are now forced to say what you can do in the bedroom, after they said "who are you to tell consenting adults what they can do" for decades.
> This is the first comment I've seen from you that leverages the UK's authority on anything related to online safety.
Edit for a (now deleted) reply: This has nothing to do with the UK; studies objectively say this activity is more dangerous than waterboarding. If that is true, action is objectively necessary, including against depictions.
We're still here. It's just that the mainsteam picked up on our technology, and now it's a bit harder to find us.
In the old days it was easy. Get into computers, and the internet when it was there, and you would for sure only meet nerds.
Now you mostly meet hipsters.
Just at that early time there was this new field that nobody was really much interested in, but some geeky people were, and that is what united us.
And then we discovered we often shared common causes too, like freedom of information and nerdism.
But now our meeting spaces have been invaded by hipsters making use of the great tools we once created to help each other.
Of course this is not fun, but there is one good thing that might come out of this.
The hipsters might have invaded our spaces, but they are still óúr spaces. We created the fundamentals, and the reason for it's success praises our vision.
Maybe that vision will rub off on the hipsters a little bit, and they get a bit of a feeling of what real freedom is actually all about.
It might take a while, and it will still get worse before it gets better, but we might have set the stage for changing the world.
The real challenge is network effects. You can't fight it. If the top dog has a considerable percentage of the population trapped in their platform, then it is already too late.
The only way to get people off the platform is to literally get hired and "trip over the cord" on purpose.. which at this point is likely going to involve very elevated permissions, and a really high chance of jail time.
I'm starting to imagine that diferent currencies, like money are converging with their platforms
each platform having your its own currency? well that's simply one nation. a federal government, registered at the UN. this translates the UN into Etherum, register your currency and be a part of what exactly?
I have to admit, I've always wondered what would happen if say, the old Twitter staff had just completely wiped the servers before they left. Would they even try to bring the site back, or just write off the whole company?
But seriously, a network effect isn't necessarily the end of the industry. Lots of sites and services make major screw ups that drive away a lot of users to an alternative in some way or another, or outcompete a huge incumbent. Like what happened with Digg and Reddit.
Digg was sold to someone that didn't understand the internet, and it was not social, so it was just like HN - not a place to put your "page" down and have your friends follow you there.
Reddit is a bit different - but as far as the data is concerned it has not once dropped in usage and is about double what it was a couple years ago.
What a load of nonsense. The dream isn't "dying", it was never alive: crypto and "Web3" were never about anything other than "number go up" speculation and gambling; anyone who thinks otherwise is either profoundly dumb or in on the con and trying to leave you with the bag.
Very disappointing to see IEEE cover this as though it's an actual thing, would they cover the "dying dream" of Amway knives?
To me "Web3" frequently feels more centralized than just the regular web.
I love that on the regular old internet I can stand up different websites, host them myself, and have anyone with a web connection be able to pull it up on their computer and see whatever it is I'm putting out there. There is some level of centralization if I want to buy a domain name or do other certain things, but it's pretty minimal and I feel mostly in control, at least for the kind of things I want to do.
That kind of standing up my own little server and running it myself while being accessible over the broader web feels decentralized. I can put whatever I want, you can put whatever you want, anyone can access it. Buying into some more distributed web3 type hosting feels more centralized at least in the way that I personally feel. I have to buy into a specific blockchain or platform, host in a specific way, hope that it gets picked up by other distributors, deal with hashing and immutability, etc. Maybe it's a difference in understanding of the word decentralized, or a different emphasis on certain parts of the definition.
> There is some level of centralization if I want to buy a domain name or do other certain things, but it's pretty minimal and I feel mostly in control, at least for the kind of things I want to do.
This actually related to a fundamental lack of control.
Lots of tech we buy now promises "smart" but because you need to get to it from your phone, some corporation "needs" to "help" connect your device and your phone.
for example - you want a "smart" thermostat. To connect to it from anywhere from your phone to change the temperature, the thermostat contacts the mannufacturer. Then your phone has to talk to them as an intermediary to contact your thermostat.
Conveniently, they can collect data from your thermostat, or your house, or your phone, maintain a "customer relationship" or even charge you for the "service".
if things were properly decentralized, your phone could just contact your "smart" stuff and nobody else would need to be involved.
As everyone knows, freedom is just a feeling.
Not advocating for either, but I believe the term is intended to contrast with Web 2.0, which was services like YouTube and Facebook, that centralized content hosting and delivery away from people running their own web sites and putting it into corporate services that offered individual user profiles instead. Web 3.0 was intended to provide a similar level of convenience and not require all people to self-host in order to be producers as well as consumers, but without corporate entities having ultimate veto power over what can and can't be hosted and delivered.
The short version of that is it's not as decentralized as Web 1.0, what you describe, but it's not as centralized as Web 2.0.
>That kind of standing up my own little server and running it myself while being accessible over the broader web feels decentralized.
I mean that is how the internet used to work.
>I can put whatever I want.
Then I make a series of false reports and copyright claims and your ISP boots you.
>anyone can access it.
Anyone can DDOS it, forcing you behind a centralized system like cloudflare.
The internet has turned into a dark forest. Standing out and saying something controversial will ensure something shows up to eat you.
I run multiple websites since I have been roughly 14 and I never had any of those problems. I am not saying they could theoretically happen. What I am saying is that to most people running small websites they don't.
I don't use cloudflare because there was never a reason to. And if there was a DDOS I'd just wait. My websites are important to me, but I don't give a damn if they are down for some time. This isn't critical infrastructure.
I have been running internet services on shared hosts, VPS, dedicated servers, colocated own hardware, since 1996, including larger projects that have their own Wikipedia entry etc, and have never needed to put anything behind cloudflare or other ddos protection layers. The providers I have worked with mostly passed on DMCA takedown requests without issues. The small internet is still fine.
No, it hasn't. The small web is quite healthy to this day. I haven't heard anyone who actually runs services like this complain about any of the problems you suggest.
Those problems only happen to large entities. There's simply nothing to gain from DDoSing frankshobbytrainsandflowers.com other than sheer malice, and then it only lasts for as long as you care to spend resources. Frank won't care and likely won't notice if his website is unreachable for any length of time.
Conflating the small web with enterprise problems is an extremely ignorant take.
Besides, the single biggest threat today is not some motivated attacker trying to deprive you of your three visitors a week, the threat is gigacorps exploiting the public commons with reckless abandon. The only thing small web admins are actually concerned about is LLM scrapers.
You should try actually talking to small web admins before spouting off nonsense like this.
You should trying being a small web admin from around '95 to 2018 then. It was not rare for some random small site we were hosting, thats not infected with anything, to start getting flooded with 10mb+ of traffic for no apparent reason.
Whilst I disagree with what you have written, it is disappointing to see your genuine statements being downvoted.
Downvoting is meant for comments that detract from the discussion. Downvoting a comment simply because you disagree (and can't be arsed typing a rebuttal) leads to groupthink and unpopular ideas being hidden. This place is slowly turning into Reddit. And no, that isn't a cliché, it is a statement of fact.
I agree somewhat with you, and it does happen, but in this case it is so alarmist it does not add anything to the discussion.
I must be on a different internet for the last 30 years from other HN'ers. That or a lot of people say really boring things that won't matter if they are on the small web or consolidated web. It's the kind of things that get you kicked off the big providers that also make you targets to get blasted off the face of the internet.
Web3 and blockchain are not the only form of decentralization. Email is a decentralized protocol that has stood the test of time.
The bigger problem is having mega-entities like Google, Meta, and Amazon dominate the web. Instead of crypto, there should have been a focus on allowing mid-size players to have more power.
For anyone about to disagree with this; you've been lied to.
Many thousands of admins are currently operating independent email servers with zero problem, and have been for decades. Mine is going on 15 years. In practice, you have to actually try pretty hard to misbehave enough for Google and Microsoft to notice and block you. For the vast majority of independent servers, we have no problems at all routing to gmail or outlook.
Reports of the death of decentralized email have been greatly exaggerated. Independent mail servers are alive and well to this day, just as they always have been and (probably) always will.
> Mine is going on 15 years.
I've been doing it for more like 30.
> In practice, you have to actually try pretty hard to misbehave enough for Google and Microsoft to notice and block you.
As far as I can tell, the "misbehavior" that got me blocked by Microsoft was being hosted on Linode... where I'd been for around ten years at that time, all on the same IP address. Tiny server, had never emitted a single even slightly spammy message, all the demanded technical measures in place, including the stupid ones.
Because of the huge number of people stupid enough to receive their email through Microsoft, I had to spend a bunch of time "appealing". That's centralization.
On edit: Oh, and the random yahoos out there running freelance blocklists can do a lot of damage to you, too, by causing smaller operators to reject your mail.
They probably threw a fat CIDR block in their IP blacklist to fight off a spam campaign, and your IP was caught in the dragnet. This is how the big companies do it. They’ll evaluate for risk of false positives and as long as that stays below a threshold, they proceed.
> Because of the huge number of people stupid enough to receive their email through Microsoft...
Can you please elaborate on this? I use Outlook (@outlook.com) for my personal email management, but would definitely switch if there is a better alternative.
Microsoft will cut off your account if any of its abuse detection heuristics misfire. You will probably not get it back, ever, period. And, by the way, I don't remember if it was for email or storage or both, but Microsoft has admitted to going so far as trying to brute-force the passwords on encrypted ZIP files to run those heuristics.
Microsoft will (intentionally) lose your incoming mail if, say, it comes from the wrong AS. You have no control over most of that, and you will not be told about it.
Microsoft may or may not be data mining the actual content of your email. Possibly also after trying to break your encryption.
Since you not only use Microsoft to process mail, but have also inexplicably decided to use Microsoft's domain name for your address, you'll find it difficult to ever move away from Microsoft. If your account gets deleted without warning, you'll be permanently screwed. If Microsoft, say, decides to go out of the email business in 10 years, or start charging some ridiculous fee, you'll probably at least get some notice, so you'll only have to scramble to change your address before the hammer falls.
AND you're contributing to the centralization of the whole damned Internet.
The better alternative is to self-host. Even though you'll then constantly have to worry about your outgoing mail getting dropped, at least you'll be in a position to notice if that happens.
If for some reason you can't handle that, then at least register your own domain and host it with a smaller provider that gives you some guarantees. You will, of course, have to actually pay a small to moderate fee for the reliability.
Fantastic breakdown -- very much appreciated. I don't mind creating a new email address if needed, even though I've had this one for over a decade. What you mentioned are very real risks that would be catastrophic should they actually come to fruition.
I'm taking some networking courses online which will hopefully help in terms of being more tech literate and being able to do this.
Regardless, I'm definitely going to start the transition. I also use their cloud storage and have very important documents in there.
Thank you for the education and the heads up.
I had my mail hosted with my managed server for 10 years, then they started charging ridiculous amounts of money for SSL certs, so I moved all websites to a self-managed 10 times cheaper yet more performant webservers. Then they wanted to force Microsoft 365 on me and I migrated the mailserver in a similar fashion.
I use mailcow and did the migration over a weekend. It took a few hours after to get all the spam stuff right (SPF, DKIM etc), but now it works flawlessly since 3 years and I use maybe a few hours per year to run updates.
I have sent out small volumes of email, and have never been able to deliver to hotmail addresses.
Oddly enough MS hosted organisational email seems to work fine.
Who's your host? My primary VPS host and my home connection both block the only port I'm supposed to use for mail.
(A curiosity to me, as someone trained on HTTPS, which doesn't really care what port you're coming from. I wish there was just mail-over-HTTPS, since port 443 can't be blocked by a reasonable host.)
A lot of VPS hosts will unblock the port for you on request. I think that is better than open by default as that attracts spammers and gives IPs a bad reputation for future users.
Yeah... I think that one of the biggest "gotchas" with the web3 crowd is misunderstanding different levels of decentralization. Is it architectural decentralization? Institutional decentralization? Geographical decentralization? etc
Can you really say that email is a decentralised protocol that has stood the test of time? It’s more centralised than the web that you complain is dominated by Google, Meta, and Amazon. What proportion of mail isn’t sent or received by Google or Microsoft?
Yes. There's thousands and thousands of independent email servers online. You just haven't heard of them. Mine is 15 years old soon.
Even if there are 1 million independent email servers, if Google, Microsoft, Amazon, and Apple send, receive, or pass along 98-99% of emails, there's still a fair bit of centralization in the network which should be acknowledged, though this is admittedly not the same as the protocol being inherently centralized.
How do you messure this? Every university in my country has thousands of mail accounts that are constantly used and having worked at an university IT department I have never heard of any numbers collection coming by to ask us how many mails our users sent to each other.
I wouldn't be surprised if that percentage was made up by the big ones to give you the feeling that self hosting mailservers is useless.
You're probably right, I hadn't considered the volume of mail being sent within institutions that might host their own mailservers, so maybe we should say more than like 50-60%.
Is it still a good example of federation though if a vast majority (99%?) of mails between such institutions or outside of orgs large enough to have a campus, tend to go through Microsoft/Google/Apple/Amazon?
But those thousands and thousands of independent email servers mostly just send email to and receive email from Google and Microsoft. And even if they didn’t, compared with the total volume of email, they basically round to zero.
Quick search reveals Gmail and Outlook add up to about 62% of email accounts worldwide. That's a lot, but certainly not enough to make other providers a rounding error.
The problem with web3 is that it's a solution in search of a problem. People have decided they should build something around blockchain instead of starting with what they want to build and why
This is the main issue with it yes. Instead of having a compelling user- and/or howter-centric vision of how Web 3.0 could look and why, it was basically just like: like web but with blockchain and yet somehow more centralized and with changes, because we want to make a new thing.
That isn't how you elegantly solve problems and get people to use your solutions.
Ahh! It says a lot about the state of “Web 3.0” that I forgot that it was even a thing. I was hoping that IEEE Spectrum would cover the actual Web re: the efforts to un-silo it that I reckon will result in more silos with just less volume than before.
Isn't "more silos" the desired outcome, more or less? Has the web been anything but many silos? Of course it might have been different when everyone who wanted to have a presence on it was expected to build their own homepage. Now billions are online, courtesy of their phone, and don't even know what a computer is. So they naturally fall into silos where they and their friends are welcome.
There was a point where "the web" was literally a web, and it was literally "linked" together by...get this: hyperlinks!
I think that what killed that is that hyperlinks work great for browsing & discovery, but as the web matures, a lot of people want to use it for task-oriented things. And all of the Big Tech companies that came afterwards succeeded because they built a task-oriented interface that co-opted the links that were there before and turned them into ways to accomplish the task. Google took hyperlinks, used them to compute PageRank, and then used that to create a better way to solve the task of finding specific information. Facebook took user activity, aggregated it, presented it in a feed, and used it to improve the task of killing time. Amazon took product pages and direct links and used them as lead-gen to increase the ease with which you can buy things. Stripe took embedded Javascript and used it to make paying for things easier; Uber and Lyft took mobile phones and used it to make transportation easier; AirBnB took these large Internet markets and used it to make vacation rentals easier.
Where all these decentralization efforts fall down is that they tend to focus on content, the technical details of how they're going to spread bytes around, and nobody focuses on the task. It'd be interesting to recast the problem in terms of "Here's a common task of everyday life; how do you accomplish it in an adversarial environment where the government or major corporations are trying to shut off the Internet?"
I've become a bit cynic, and I think it will take time until it becomes better. Eventually though, there will be large protests that turn into law that change the dynamics of the now very centralized web.
London was known for having so thick fog from burning coal, and in 1952 it was such a bad event that 4k people died and 100k got badly affected from it. It is was started the Clean Air Act of 1956, that eventually led to clean (cleaner?) air in London.
Another one is Amsterdam. It was a car-centric city up until the 70s, where they started to rebuild it to be a walkable and cyclable city. This started because people protested the dangerous roads, which culminated with the "Stop de Kindermoord" protest.
Maybe if we live long enough, we get to experience a decentralized web again. Next time will come when it's supported by laws, that in turn stand atop a large understanding in the population of why those laws are good and needed.
> Where all these decentralization efforts fall down is that they tend to focus on content, the technical details of how they're going to spread bytes around, and nobody focuses on the task. It'd be interesting to recast the problem in terms of "Here's a common task of everyday life; how do you accomplish it in an adversarial environment where the government or major corporations are trying to shut off the Internet?”
Nice observation and can speak to why anyone intent on selling you content writes a “user story” to outline the tasks that lead to a purchase or engagement.
The cult of marketing ruined it. They're still fantasizing about personalisation, which in marketing speak means optimizing a mattress sale opportunity for Ted Waciznski of 932 West Moreland Ave in Walla Walla, Washington next Tuesday at 3:54pm as he reads CNN.com as he has every morning between 11:19am and 11:26am from a mobile device in the same spot every weekday for the past seven years, rather than, you know, "here's a cool thing you asked us to show you." The only links that matter are to an adtech/martech firm, and presumably the company that wants you to buy something.
Yes, I think so. I agree with you actually. But it doesn’t always seem like that’s the desired outcome for some people when they mention “decentralization”.
I have a particular interest group in mind here, who lament the current state of the Web to an extent unique to our time. It seems like they’re upset with their loss of agency/authority online. When a free, ad-less *.blogspot.com or LiveJournal presence could net you some clout you’d be precluded from otherwise in the real world.
Take for example any evidence of disapproval you may find from the proponents of ActivityPub toward the development of the AT Protocol in parallel. Also notice how moderation features on these federated platforms essentially centralize them and their respective enclaves. The desire for interoperability between social media platforms conflicts with the imminent re-siloing being examined. Or is it just the same ends with new wiry means attached?
“Decentralization” in this context seems more like separating from the whole with the hopes of replacing it. And I’m not confident that these many new “wholes” can resist jockeying one after the other for political influence.
A big chunk of my favorite websites to this day are hosted by motivated individuals in their spare time. And that hasn't changed.
Maybe I am an exception as I find social media nearly worthless by this point, but hey: Just because you don't see it doesn't mean it is not there.
Is the title a nod to Molly White's skepticism?
https://www.web3isgoinggreat.com/about
Came here looking for this. Love that site.
"web3" is almost a blast from the past though, simpler times. Seeing it mentioned on HN now, I just had to take a look at https://www.aiisgoinggreat.com to see if it was available, it is not. Nothing there, but I'd love a similar site about AI debacles.
Pivot to AI probably fits the bill pretty well? https://pivot-to-ai.com/
The real Web 3.0 is the Ted Nelson internet we never got.
Wasn't Project Xanadu just HTML with backlinks? I was just thinking about it in the shower yesterday and realized I had no idea how it could have possibly worked for cross-site links, unless the web was either fully centralized, or fully distributed.
It was/is an pipe dream. The whole schtick was that links were visualized so you could, "trace" what you were researching visually without having to rely on jump links. In theory (more pipe dream) the whole topographical link structure of the internet would be capable of pretty easily being visualized by a person.
In my somewhat educated opinion, David Chaum is the OG inventor of crypto currency aka digital cash (eCash), back all the way in 1982.
He was last seen pitching family offices in the UAE pitching AstroCool, a concept involving spraying moon dust in orbit around the earth to block some of the sun light. You won't see this referenced anywhere. I was told about it by the person who introduced him to family offices over there.
Weirdly enough, in 2008, I was so pissed about the financial crisis, I went online to promote the idea of an absolute-valued energy-exchange-based currency called the Joule Token. It had no cryptographic component (was a centralized immutable store that is co-owned and co-administered by the community) The banner I promoted it under was "P2P Programmable Currency" and I kind of spammed Michel Bowen's P2P Research mailing list and wrote a giant post on the P2P Foundation Wiki titled the P2P Energy Economy. The idea was to decouple profit/interest from exchange, so you can value things based on the energy it took to create them and transport them, and instead of profit as an incentive it was about fulfilling the need for energy: you sold things because you needed to recoup the energy and you created the good that you sold because you could, no other reason. Obviously, it was a thought experiment, not an actual implementation or anything meant to becoone real. However, after I was kicked off the list for spamming (or rather for list etiquette -- Hi Bryan! if you still frequent HN lol), Satoshi@gmx.de replied to my emails 4 times between early 2009 and late 2010, so I literally ghosted him, but without knowing. I didn't know until Bryan Bishop, who was behind my banning, I think, told me here on HN 3 years later, in 2013, if I recall correctly.
Here is what ChatGPT knows about those times (I asked it to tell me what it knew without searching online)
https://chatgpt.com/share/6893129d-1a20-8002-a08b-acd66bf91b...
If you remember me from those days, I moved from Evolving Trends on Wordpress to Substack https://understoryai.substack.com/
Good times. They're back again. (sarcasm)
I think a different type of decentralized web is growing. It's in Hetzner datacenters but it's a lot more "free". We're at an age where besides GPU inference an indie can afford to create cutting edge web services. There's still players that could really harm this new ecosystem (looking at you cloudflare) but the fact that the majors like Google and Facebook tried to kill it and failed to kill it proves something.
Please check the new Braid HTTP, it's shaping up the as the killer feature of the Web 3 to enable killer web based local-first application [1],[2]:
[1] A Synchronous Web of State:
https://braid.org/meeting-107
[2] Local-first software You own your data, in spite of the cloud:
https://www.inkandswitch.com/essay/local-first/
I'm still not a fan of how 'web3' has become a buzzword for crypto. The name seems like it should describe a much more interesting and general concept, not just "it's a service, except you can make money/sell NFTs/whatever".
But I'm not sure I'd write the decentralised web off yet. A lot of the issues it's faced are UX and design problems more than fundamental issues, and the way the world is going makes it seem like a comeback could be inevitable at some point.
After all, a good federated or decentralised system could be resistant to the age gating laws in places like the UK and Australia at the moment, and be a lot more tolerant of content that's critical of those in power than Twitter or Reddit would be (like with criticism of Israel and the goings on in Gaza, criticism of Trump and his administration, protests against authoritarian regimes worldwide, etc).
Large social media services are very willing to bend the knee to those with power (both political and corporate), while a more decentralised service may very well not be, especially if either the people using it are hard to identify or the infrastructure is spread across the world/in places with much more lax laws.
Oh, this is about crypto. I thought it was about the decentralized web for a moment.
Is Web3 just regular internet with ponzi schemes bolted onto it?
I would really love for new projects to find their own name instead of calling themselves the next version of something they don't like. "Web3" has never been a new version of the web, just a way to inject undeserved clout into a (IMHO fundamentally flawed) experiment. There are so many great unused names! Please pick one.
Missed opportunities over decades, in EU especially.
Which areas support community‑owned or open‑access networks that enable multi‑ISP competition and affordable symmetric service (aka "true internet")?
We're reimagining what web 3 means in atprotocol, without the blockchain and probably avoiding the term "web 3"
a colleague of mine recently insisted that bittorrent be included in web3 techs since it uses hashing. if we toss torrents under the title of web3 then it's been a huge success.
atproto, however, could learn a lot from torrents and all of the other protos since then. for example, their recent push to centralize bookmarks.
> for example, their recent push to centralize bookmarks.
It's worth noting that the current bookmarks implementation is "temporary" and will be moved to the PDS as soon as private data repos are supported on the PDS.
The approach they are taking (and are recommending to other devs in the space) is "develop your private data as if it was stored on the PDS but stored on the appview. then migrate to the PDS once private data launches". If you maintain the same lexicon/schema under the hood, you can just copy the data to migrate it.
Private data is unfortunately really hard to do right so this "start workshopping it and get ready for us to roll support out" approach seems like the best temporary solution.
> a colleague of mine recently insisted that bittorrent be included in web3 techs since it uses hashing
Related pet-peeve: Folks who say "there's still great promise in private blockchains." This is equivalent to saying that self-balancing Segway-style devices can still become the dominant mode of transportation if we juuuuust make them bigger, enclosed, and add another pair of wheels so that they don't have to self-balance anymore.
Yea, the bookmarks thing is interesting, and wearing my conspiracy theory hat (it's not the red one), I wonder if it's not an experiment to see what the people will tolerate or how they will respond. It could equally be attributed to an early win for the new product manager, finally delivering a long requested feature in a POC form factor, while private data gets figured out.
I'm highly involved in the private data work and we'll have a better bookmarks in the long run
I like that atprotocol sits in the middle of web 2 & 3, ideas from both without being beholden to either
I'm pretty sure the private data model is going to be clearly defined by EOY, between the community group work and bluesky internal work. We also made sure that bookmarks will migrate cleanly to protocol private data when it lands, so no need for the tinfoil hat. Your other explanation is pretty close to the mark.
EDIT: also fwiw, I got my start in 2012 working on ssb and then in 2013 on dat, both of which were conceptualized as bittorrent variants. Atproto has a pretty clear lineage from that p2p work. You just can't do large scale social computing on user devices.
yes, and we are using computers in our pockets now that are way more powerful than the ones we had when we were working on ssb. local indexes are possible, you and dom showed that back then.
ssb had private bookmarks, let's dig into how those were enabled
agreed! we should continue this discussion over on atproto
one thing I would like in atproto is some form of smart contracts
(transactional semantics over accounts and xrpc calls)
Is there a way to use it without making an account? I really want a permission-less way for two browser clients to send a simple message on the public web.
>Web3
Why the humility? If I'm ever going to create a payment system, I will call it "Universe v2".
Or slightly less grandiose on a mere planetary scale, "World[coin]" :p
If only we had an totally open universal way to signal between two WebRTC clients. It wouldn't be perfect but it would go a very long way.
I never see mention of https://holepunch.to/ I think it would be great if we had more serverless apps.
I'm trying to understand its github if holepunch itself open-source or not. Hypercore is but is holepunch different?
AFAIK, "Holepunch" is the overall project which is all open source. It is composed of https://docs.pears.com/ , https://github.com/holepunchto/libpear which runs on https://github.com/holepunchto/bare/ which runs on a V8 interface to libuv.
And, then Hypercore, Hyperbee, Hyperdrive, Autobase, etc are all modules you need to get things done so you can make an app like their serverless Zoom clone https://keet.io/
https://github.com/dmotz/trystero is the closest we currently have.
I wish one of these networks could be the go-to clear default choice. Options are good, but it would be great to have a best option for a large discoverable community. Actually, their default choice Nostr sounds pretty awesome https://nostr.com/
868mhz/902mhz ISM bands
That would be pretty cool if everyday computers could do that. I'm wishing for something that would work in a browser like with HTTP.
I think, the Permaweb is on a good way.
They have decentralised storage (Arweave), decentralized HTTP gateways (ar.io), and decentralised name service (ArNS).
What's some content on Arweave that's good to follow?
How much of the world's traffic goes through Cloudflare? Won't be surprised if in 75 years we all learn it was a NSA front lol.
I was just browsing Neocities yesterday and having fun online for the first time in a long long long long long time.
Little known fact: Starlink should be a pure orbital internet backbone to communicate anywhere in the world, right? Because of the laser links between satellites?
Not so much. All the traffic from the US must transit through an earthbound surveillance hub, so the latency advantages are lost. Same with other countries.
(Please anyone correct me if I'm wrong about this, I was gravely disappointed to hear it and would love to hear that I'm wrong)
Gotta keep in mind that PRISM (or maybe something like it) is still a thing. If US traffic isn't sniffed by Starlink itself, then it's happening at the next hop in the nearest terrestrial network provider.
>The term Web3 was originally coined by Etherium ...
Sure, mistakes happen, but it's hard to take an article seriously when such fundamentals are messed up 8 words into the article...
I also responded this way until I realized this original comment was talking about the misspelling of Ethereum as Etherium.
i don't get it. is the fact that somebody misspelled a recently-coined nonsense-word product name/trademark an indication that the broader points are invalid?
Tangentially relevant: https://www.web3isgoinggreat.com/
The core tenets of web3; privacy, data sovereignty, encryption, open source, open protocols, and peer-to-peer networking are all good. The problem is the movement was never truly about this, and, even to the degree it was, it was taken over by crypto grifters. The ultimate indictment of web3 is that the old school nerds weren't interested, they are the perfect audience, and yet...
meh. "cryptographer" usually means someone who practices "cryptography," [or "cryptology" if you live near Baltimore] not someone who participates in "cryptocurrency" transactions.
"cryptogrifter"
Article title was either changed or wasn't correct to begin with!
I'm 100% OK with "Web3" dying a slow death. Garbage in, garbage out.
> “Hey, this is a technology that has something to do with blockchains.”
This.
As goes capital, so goes the web. Capital is becoming more concentrated, and so power over the internet becomes more concentrated. Because at the end of the day, the web serves capital, not individuals. And AI represents the ultimate centralization of information and power.
The failure of Mastodon to break into the mainstream especially in the wake of Twitter/X's debacles is a good example on how decentralized web is a good idea on paper but won't work in practice. It turns out that a) people will just join the largest instance/community possible for network effects b) internet discourse is bifuricated enough that preemtively banning instances just in case makes discourse even more bifuricated and c) maintaining and moderating said instances is expensive and time-consuming.
Even though Bluesky did break through to the mainstream with ATProto, it's unclear how ATProto is either a functional or competitive benefit, or if its users even know about it.
Mastodon might not have broken into the mainstream, but it's surprisingly viable. I am totally OK with its current state to be honest - feels like the cozy internet of 2000s.
My Mastodon feed has a very "early to mid 2000s LiveJournal friends page" feel to it, and I honestly adore this.
This is, at least in part, a UX problem. Even as someone with a lot of technical experience I found Mastodon quite disorienting at first. Bluesky has solved this much better which is why they've won out over Mastodon.
It’s arguably much easier to solve the UX issues when you are designing a centralized service, which is what Bluesky is.
It is easier, but it's still possible to solve the UX issues with a decentralized service, in my opinion. What I think is the main issue is that these decentralized services are made by programmers with little regard or intuition for UX, and there is also a lack of funding to work on UX problems.
I agree the UX challenge is much more challenging for decentralised services. I don't know enough about Bluesky to really comment on whether it is centralised or not.
Regardless, I think there's another thing that helped Bluesky: VC capital. In particular, to hire people to work on UX. It's a bit of a pet-peeve of mine, but I find it strange that designer don't contribute more to projects like Mastodon, which definitely need it. Even from the selfish angle of building a portfolio, helping solve Mastodon's UX challenges is much more impressive and realistic, than doing the millionth redesign of Gmail that will never get implemented.
It's a "distributed protocol" but there's really only a single server using it.
People are running every element of the Bluesky / AtProtocol stack independently. Bluesky could disappear and it would continue to function as is (albeit with lots of Bluesky data lost).
PDS's to hold users data, relays/firehoses to aggregate & forward traffic, AppViews to create composite views of likes, replies, etc, resolvers to lookup DIDs, clients to access the network. Each of these has independent implementations. BlueSky is already decentralized & already has viable credible exit. It's not decentralized, and indeed the scalability & accessibility of having firehose consumers has the greatest scale out decentralization characteristics we've seen anywhere short of BitTorrent.
I don't know why you're being downvoted. I tried it the other day and this is indeed how it works. I could even see my home traffic rise and fall over the course of the day in time with activity on the network.
It seems quite reliable that Bluesky gets down voted. Whether it's mastodon/Fediverse folk or right wing pro twitter philes or both, I dunno, but it sucks to have aggressors out there, dark forest freaks sniping away.
I try very hard to find the positive & to upvote things I don't fully agree with, if well argued. I wish the social network of HN could do more against adversarial zero-sim thinking, didn't have people who insist on draining.
Arguable indeed: plenty of email clients have great UI. Both RSS readers I used are better than Facebook.
People keep thinking decentralization fails for technical reasons, when it’s economic. Without an economic model nobody can afford the massive effort required to make software polished and usable.
As a general rule I’d say polished user friendly software takes 10X the effort at a minimum vs software only nerds can use. That’s probably an underestimation. For consumer software it’s probably 100X. It’s because computers are incredibly confusing and hard to use and it takes immense effort to overcome that.
(If you don’t agree that computers are confusing and hard to use, you are part of a tiny highly educated minority.)
I’ve said this around here like a hundred times because it feels like I am the only one who gets it.
The tech to create an excellent decentralized network with an excellent experience and without the weaknesses of Mastodon exists. It doesn’t get used because nobody pays for it. Centralization gives you an economic model from either directly charging for access or selling data or ads, and none of that works in a decentralized world.
> This is, at least in part, a UX problem.
Right, but that's the entire point: Mastodon's UX problems are caused by its decentralization and mostly cannot be separated from it. Arguably all the problems of decentralization that make users disprefer it are UX problems-- that doesn't mean they are easy to solve.
I agree completely. This isn't the world of BBSes where everyone had a corner or two that they hung out on. As long as posts are gamified with likes and notifications, decentralization will never win. Even HN isn't immune and it has carved out a pretty specific niche of text based and no ads social network focused largely on tech. I don't think another decen-HN is likely to take off or break into other self organizing segments even though this is easy to attempt with some Mastodon servers.
> This isn't the world of BBSes where everyone had a corner or two that they hung out on.
That's exactly what Mastodon is. The whole reason mastodon exists is because a whole lot of people wanted exactly this experience.
Mastodon won't win because they aren't playing. Which again, is the whole point. It's extremely deliberate and intentional. People use mastodon because the "game" is abhorrent and objectively extremely bad for people individually and society as a whole.
So mastodon eschews the gamification and like farming and fake notifications. One of the most common jokes throughout the whole network is "wow, this post blew up and is doing Real Numbers" and the post has like ten boosts and twenty likes. It's extremely rare to see more than a couple hundred notes on any one post.
It's on purpose. Mastodon users like and actively want it to be this way.
>This isn't the world of BBSes where everyone had a corner or two that they hung out on.
Yes it is. It literally is. If we want to do that we can just do that. The internet is not a zero-sum game. We don't have to play to "win."
>I don't think another decen-HN is likely to take off or break into other self organizing segments even though this is easy to attempt with some Mastodon servers.
The only reason HN is popular is its connections to Silicon Valley and startup culture. It didn't really carve out a niche so much as succesfully create and market an image. Without that it isn't anything remarkable, and there are already tons of technically-focused communities on the fediverse.
Those sound like reasons why federation doesn't work. Some of us don't think federation is decentralized enough anyway.
I asked people who use platforms like X and Substack (Notes) why they don't use Mastodon and they said they can't discover new people and new content.
Discoverability was an anti-goal for Mastodon initially. People from queer community felt it enables bullying. Mastodon has added search functionality since then, but it's still pretty basic.
IMO it's not necessarily bad though. It prevents the network from becoming centered around content and keeps it social. You discover new accounts as they get reposted by people you are already connected to
I feel like this is a clear case for "security through obscurity doesn't work" saying, and for the same reasons as the original meaning. "Normal" people won't find such community because it is pain in the ass to do, while bullies will get there just fine because they are motivated. It is like reverse moderation, filtering out less radical and less pro-active newcomers.
Interesting. Thanks for the history on this.
Mastodon is completely detached from the world, it's irrelevant.
This attitude is weird to see on Hacker News of all places, a forum that considers itself a quarantine zone from the rest of the world and the modern web, and which is also irrelevant to almost everyone, including much of tech.
Sure, but hacker news isn't an alternative to Twitter, nor does it market itself as such.
I'll post this every time mastodon comes up: to compare mastodon to Twitter, Facebook, bsky, etc. is fallacious in the extreme. Saying that mastodon is "losing" to bluesky is a complete non-sequitur.
Mastodon hasn't lost to these other services because they aren't competing. Nobody on mastodon actually wants to replace Twitter. Mastodon exists explicitly as a rejection of mainstream social media. Calling this "losing" is like saying that a bunch of kids playing basketball in the street are "losing" to pro hockey teams. Not the same game, not the same scale, the only similarity is that they're both playing a sport.
Mastodon doesn't want to compete. That's totally counter to the entire philosophy. Mastodon isn't even remotely playing the same game as twitter and Facebook. It just so happens that mastodon has a text box with limited characters. That's it, that's the only similarity.
Of course a chess club of 10 year olds in Japan isn't going to "win" and displace the entire American baseball league. You'd be laughed down for even suggesting that. And yet, people think that mastodon is "losing".
Mastodon is exactly where it wants to be. The small scale and diffusion are features, not bugs.
Besides that, it's functionally impossible for "mastodon" to go mainstream. To suggest that is a fundamental misunderstanding of what mastodon is, how it works, and how the people involved run it. At absolute most, you could get a few big servers to go 'mainstream'.
When a server gets too big or too popular, the network treats it as damage and routes around. A lot of admins do this purely on principle: huge servers are antithetical to the philosophy and actively bad for the network. The rest block them because the vast majority of attacks and spam originate from such servers.
If mastodon went mainstream, the network would fracture (again) and the majority of nodes would re-form the network in the obscurity they enjoy and value.
This is all extremely intentional and deliberate. Mastodon is an explicit rejection of all the things you think it's "losing" at. Mastodon doesn't want to "win", they largely want to be left alone to continue enjoying a small internet with limited connections.
Mastodon has won its own game. It's firmly established itself in a niche, and is not going to move from that niche to chase fads, engagement, metrics, profit. Mastodon wants to be exactly what it is because mainstream social media is a very bad thing
You sold me on mastodon. Is there a node for scientists? How do I find it?
I just googled "science mastodon" and found plenty of results, including this reddit thread with some listed[0] and this list of academics on Mastodon[1]
[0] https://old.reddit.com/r/Mastodon/comments/yy71wh/is_there_a...
[1] https://github.com/nathanlesage/academics-on-mastodon
Although, of course, you can follow any account you like from whatever instance you choose to use (assuming that instance doesn't block it.) A lot of people you may be interested in are on universeodon.com, although typically you'll discover things on Mastodon by searching hashtags, not people.
Discord servers are half of a good counterexample, since people try to avoid the extreme end of network effects and want their chat partners pre-screened.
But... it's all on Discord, not IRC or jabber or anything.
Now that you mention it, I vagely remember something about Bluesky being built on some distributed protocol. So yeah, I guess it's correct that doesn't mean much to it's popularity. Which is sad of course. But yeah, most people don't know about technology, and security and privacy, so yeah, why would they care..
I think there's an element of how decentralized services work that's not sufficiently refined yet. In a lot of ways it's a better way to have a good user experience with decentralized data that we're sort of looking for, and in some ways, it doesn't really fit how the web works.
yes, lots of stuff needs to be figured out. Like, you don't need decentralized auth servers.
bittorrent has webseeds. I dump large files and archives on my sever, give people the magnet or link it from an article, they download it fromthere. If for any reason I no longer want to host the file I just remove it. Usually I just want the space back, in theory there could be bandwidth issues or an excess of interested participants. You can put fairly interesting websites in torrents too! That no uninvited participant can find them is a feature not a bug.
Email does html too! PDF is also great for publishing. You can send people videos and even executables if you are close enough.
All of these decentralized social media concepts were/are too technically complicated for too long during the Twitter debacle. Even now, the top results for "how to use Mastodon" have jargon incompatible with mainstream usage.
If decentralized is great on paper and not in real life, then the paper is wrong and thus not great. Additionally, if decentralized doesn’t work, please explain the survival of email.
d) The kind of people who put a little thought into how Twitter could be done differently may come to the conclusion that they don't actually need a Twitter in their lives
Mastodon would be OK but the UI/UX is just terrible... interacting with someon on different instance is utterly cumbersome and I do/could miss a lot of posting in global view if someone didn't already followed someone from other instance...
Terrible compared to what? Mastodon is just fine for me. I'm an oblivious nerd here I guess. But I sometimes use Twitter too, can't say I find the UX refreshing by comparison.
For example bluesky. I tried masto a bunch of times, still have the account but each time I try to use it, it's so cumbersome. I tried alternative UIs but they are no that better...
> Terrible compared to what?
Email?
The entire point and purpose of mastodon is to not be mainstream social media. Almost everyone there signed up because mainstream social media is so awful. The small and diffuse nature is one of the main benefits. Because it's distributed, there's an extremely niche community server for any subject you might be interested in.
Mastodon captures perfectly what twitter was at the beginning. Just people talking to each other and having fun. Not performative like fishing or whatever. No ads, algorithms, endless feeds, AI slop, or spying and tracking.
It turns out that a lot of people think mainstream social media is objectively abhorrent and want to connect with other humans in a more natural and user-driven way.
Frankly, "mainstream" is a dirty, disgusting concept here. A very large fraction of users would put a significant amout of effort to prevent that from happening. If one server did become "mainstream", as mastodon dot social did, the network treats it as a damage and routes around. Many servers just cut them off because the vast majority of attacks and spam originate from the big public servers.
Mastodon won't go mainstream because they don't want to and the system fundamentally cannot operate that way. A few individual servers may go mainstream, but we'd eventually consider it a hard fork. The network would fracture (as it has many times) and the network of small servers will go back to the obscurity we enjoy and cherish.
Case in point: Facebook tried to force their way into the fediverse network through threads or whatever it was called. There was a pretty hard split in the network as nodes that value privacy and safety cut themselves off from a literal hostile invader. Many servers went recursively through the network to cut off any servers who hadn't blocked Facebook. I haven't heard anything about Facebook trying activitypub again, so I guess the quarantine was effective.
> Many servers went recursively through the network to cut off any servers who hadn't blocked Facebook
A big problem is that this is happening with other servers as well. I've seen nodes that block all other nodes that run Pleroma because it's "for Nazis". And some of them also block nodes that don't share most of the blocklist.
> decentralized web is a good idea on paper but won't work in practic
This claim is so vague it's meaningless. I could just as well argue that torrents are the purest example of decentralisation, and work amazing.
That’s a key feature
The simplest way to answer this is to ask "whose dream?".
We have billions of data points conclusively showing that users don't care about nerd fantasies. No one cares if it's decentralized or not, if it's open source or not, if it's patent-encumbered or not. Maybe there are good philosophical reasons why people should care, but they don't.
So you can keep digging your hole, or you can build products that are simply good and happen to embrace your philosophy without that getting in the way of usability.
In this particular case, the dream the article is talking to was tainted even for the nerd audiences because of the sleaziness of a lot of what was happening under the banner of web 3.0.
> users don't care
From The Truth by Terry Pratchett, on a similar dilemma between pursuing the (journalistic) public-good versus satisfying the immediate desire of the average public:
> "Are you saying people aren't interested in the truth?"
> "Listen, what's true to a lot of people is that they need the money for the rent by the end of the week [...] This is a report of the annual meeting of the Ankh-Morpork Caged Birds Society [...] They've got no say in who runs the city but they can damn well see to it that cockatoos aren't lumped in with parrots. It's not their fault. It's just how things are."
> [...] "It's important! Someone has to care about the... the big truth. [...] if they don't care about anything much beyond things that go squawk in cages then one day there'll be someone in charge of this place who'll make them choke on their own budgies. You want that to happen?"
____
Our civilization is filled with uncountable things that were once "nerd fantasies" that the average person doesn't even remotely care about... Until it all goes tragically wrong.
If I'm being honest, your quote makes the first character sound like an out-of-touch loonie. It doesn't make the point you're trying to make. Maybe in its original context it did, but the way you've presented it, it's like going to an AA meeting with a large binder, intending to expose corruption in the senate. There is such a thing as time and place.
The context is that the Lawful-Neutral leader of the city-state has been framed and thrown in the dungeons by a shadowy conspiracy of powerful figures. The printing press was just recently invented the first character accidentally inventing Journalism wants to break the story while a coworker is arguing that it's not relevant to the citizens.
Expanded portion:
> "Someone has to care about the... the big truth. What Vetinari mostly does not do is a lot of harm. We’ve had rulers who were completely crazy and very, very nasty. And it wasn’t that long ago, either. Vetinari might not be ‘a very nice man,’ but I had breakfast today with someone who'd be a lot worse if he ran the city, and there are lots more like him. And what’s happening now is wrong. And as for your damn parrot fanciers [...]"
____
With respect to "we've had rulers", a bit from a previous book Men At Arms:
> "[...] He wielded the axe, you know. No-one else'd do it. It was a king's neck, after all. Kings are," he spat the word, "special. Even after they'd seen the... private rooms, and cleaned up the... bits. Even then. No-one'd clean up the world. But he took the axe and cursed them all and did it."
> "What king was it?" said Carrot.
> "Lorenzo the Kind," said Vimes, distantly.
> "I've seen his picture in the palace museum," said Carrot. "A fat old man. Surrounded by lots of children."
> "Oh yes," said Vimes, carefully. "He was very fond of children."
Yeah, I got that part. I took at face value that whatever the character's concerns were, they were legitimate. Even granting that much he still sounds silly, because he's trying to publish his findings at "the annual meeting of the Ankh-Morpork Caged Birds Society". Like I said, time and place. Of course it would not be welcome at such a venue; it's off-topic discussion. It would be like exposing a CP ring through a post in a programming forum. It'd just get your thread locked.
Trying to commandeer the attention of the attendees like that is simply disrespectful; it doesn't matter how important you think what you have to say is. Unless the building is on fire or there's an armed squad standing outside, whatever he had to say could have waited until the end, and whoever wanted to listen to it could have stuck around, and whoever didn't could leave.
> he's trying to publish his findings at "the annual meeting of the Ankh-Morpork Caged Birds Society". [...] Trying to commandeer the attention of the attendees like that is simply disrespectful;
Ah, I see the confusion, no, these are two people running an actual newspaper.
When the coworker says "this is a report" it's because they're holding it up in their hands to display it, as an example of something their average working-class reader might care more about than estoeric politics.
Alas, the 2-hour edit window is closed, but:
> She held up a piece of lined paper, crammed edge to edge with the careful looped handwriting of someone for whom holding a pen was not a familiar activity. "This is a report of the annual meeting of the Ankh-Morpork Caged Birds Society," she said. "They're just ordinary people who breed canaries and things as a hobby. Their chairman lives next door to me, which is why he gave me this. This stuff is important to him! My goodness, but it's dull. It's all about Best of Breed and some changes in the rules about parrots which they argued about for two hours. But the people who were arguing were people who mostly spend their day mincing meat or sawing wood and basically leading little lives that are controlled by other people, do you see? They've got no say in who runs the city but they can damn well see to it that cockatoos aren't lumped in with parrots. It's not their fault. It's just how things are. Why are you sitting there with your mouth open like that?"
Fair enough, then.
As a side note, I find it really interesting and a bit disheartening to follow the history of “web 3”. The original “web 3” was a beautiful vision including the semantic web and its cousins. The term was hijacked by cryptobros through what could have been a genuine or malicious misunderstanding of the original web 3
There's a general phenomenon where terms get annexed, not by meaning that makes most sense, but by meaning that is most salient, most eye catching or emotionally riveting. This leads to Flanderization of many many important cultural concepts.
I fear this is the reason why intellectual discourse seems to have stagnated. We no longer pursue the spirit of important ideas but try to endlessly put them in uninteresting boxes.
Weird. The first time I came across "web3" was as an interface with Ethereum nodes. This is my first time hearing about the term ever having an earlier meaning.
> We have billions of data points conclusively showing that users don't care about nerd fantasies. No one cares if it's decentralized or not, if it's open source or not, if it's patent-encumbered or not. Maybe there are good philosophical reasons why people should care, but they don't.
This is focused on the technical aspects as if this is why people go to the big platforms. It is not the why for where they are or where they will go.
(1) The big platforms came first, most people don't know there are different ways for humans to organize social media.
(2) They have desires to leave for greener pastures, they know these companies don't have their best interest in mind.
We need a viable alternative and ATProto is in the best position right now. I have been able to get people excited and even create an account, knowing that the better way forward is still very young and a work in progress. I talk to them about features at the conceptual level, never about federation or any of that gobbledygook that they don't understand or want to
That is a very selective truth.
In the x86 days p2p filesharing applications unambiguously proved that people preferred a decentralized web, not only were they getting their music for free, they discovered new and hardly known niche artists in genre's they liked (some of the p2p software allowed checking out other music from the same seeder).
What really happens is the transition to ARM platforms (tablets, smartphones) with TrustZone technology, and the resulting lack of FOSS Operating Systems & Software being installed by the user, and a lack of development because average user can't install it either.
Perhaps one could argue that at world population scale people couldn't afford both an x86 system AND a smartphone, and preferred the latter, at the cost of user freedom.
>We have billions of data points conclusively showing that users don't care about nerd fantasies.
That's 100% BS. There is no such data or research. All the 'users don't care about x' comments all hold about as much water as if I were to claim `North Koreans don't care about basic freedoms`.
It's not just the dream that it's dying - I mean, it's not like it's just external circumstances that are killing the dream.
Where are the dreamers?
Back in the day tech nerds, coders, and hackers were far more vocal in favor of open standards and decentralized web.
Everyone called us paranoid nerds and laughed at our concerns. We spent 20 years trying to tell the people we care about how to be safe and they didn't listen because the tech is ugly, with rough edges, and doesn't have rage bait and trillions of dollars of ad-tech money supporting it. I personally stopped advocating as loudly because I realized that it would take much, much more pain for regular people to get out of their comfort stupors and start giving a shit. Meanwhile, p2p tech slowly progresses in the shadows where the mainstream can't see it, can't support it, and doesn't care.
The best time for advocacy is when people get burned. It's not a matter of if, but when. People learn through pain.
New billionaire owner takes over your favorite microblogging platform? Pain. Your clubhouse has been destroyed. You suddenly realize you didn't own anything.
New billionaire owner decides they don't like the topics you talk about and bans you? Pain. Your ability to express yourself and communicate and make connections with like-minded people has been destroyed. You suddenly realize that you actually need to get permission to talk to the people you want to talk to.
Massive data leak that includes your intimate personal information, photos and videos? Pain. Your privacy and personal life have been irrevocably violated. You suddenly realize that you don't control any of your own private information, and it actually isn't safe out there in those clouds. Even your most intimate whispers are recorded forever and will eventually be listened to by strangers or published on the cover of the New York Times.
The list of reasons to get the fuck off these corporate, centralized apps is growing every day.
I'm even more demoralized than you, as I feel the problem with "the best time for advocacy is when people get burned" only manages to convert a handful of people and then a month later more new people have entered the system to replace them who know nothing about this and will adamantly claim that such a problem is impossible until it happens personally to them, which takes years and years, at which point the same thing repeats.
I understand the feeling of demoralization, but converting a handful of people is better than converting no one. Those people are now safer because of you, and that's something to be proud of.
We're still vocal, but everyone has to get paid, and only the supervillains are paying these days
> Where are the dreamers?
Getting paid $$$$ at bigtech companies to build the centralized web.
The writers and the artists, all are paid to tell us lies
To keep us locked inside, they keep us locked inside
(Janelle Monae)
We're hanging out on the decentralized web and/or private networks.
Sure it's not the same size as the "real web" but it's got no ads, less spam, less bots, and enough to people to blow out your dunbar ring and fill your cup.
You might never get to be a freelance marketer drop shipping some stupid crap - but that's a price some are willing to pay.
They are disillusioned.
They watched big tech clobber the internet.
They watched network effects cause irreversible centralization.
They watched Cambridge Analytica and "openness" get abused.
They watched toddlers being given iPads to watch YouTube by their parents.
They watched kids have mental health issues from online participation.
They watched teens start strangling each other after watching porn.
They watched robots become 40% of internet traffic, and that was before AI.
They watched crypto go from decentralized payments into a get-rich-quick scheme.
And they questioned whether it's worth standing behind this.
For us who remember the 20th Century? Yes, I agree and we can see the causes. For the next generation? It's more that they're wrapped up in centralized closed platforms for their hobbies. Most of them never considered anything else. They were born deep underwater and have no idea why they can't breathe.
We had BBSes, forums and IRC. Now it's Discord.
We had LAN parties. Now it's publisher-controlled match lobbies.
We had RSS feeds. Now it's curated social algos.
We had myriad self-hosted wikis and homespun fansites. Now it's all ad-choked Fandom.
So on and so forth.
For Discord and maybe some others, I wonder if it’s just a packaging and marketing problem that just hasn’t been solved.
Yes there’s Matrix, but it’s not packaged in a way that makes hosting a node accessible and the way it’s marketed, there’s no clear benefit to typical users.
Imagine if instead, the client and server were one in the same and starting a server is as simple as running the client and clicking “new server”, with the software figuring the rest out. Then, on the marketing side you can sell it as a way to get features that Discord puts behind a paywall for free.
There’s no way that wouldn’t have a dramatic impact on take-up.
Definitely agree. Take a look at the number of Matrix clients:
https://matrix.org/ecosystem/clients/
The world doesn't need that many clients, it needs 1 or 2 really good ones that are well polished, supported and marketed.
This is a problem in other open-source ecosystems. Eg. We don't need more web browsers, we need Firefox to focus on being a great web browser.
This feels very strange: the clients on that page span a huge range of maturity and capability. If you put every email client or every web browser on a single page you’d get a crazy mix too - does that mean that email & the web have failed? Just the opposite, surely.
If you want to narrow it down to one or two really great ones which are well polished, supported and marketed just pick Element X, Beeper or maybe Fluffychat?
> If you put every email client or every web browser on a single page you’d get a crazy mix too -
And yet how many web browsers are actually well known and used by the average web user?
Safari, chrome and Firefox (if you're lucky).
My point is that if there are too many choices for users then the network effect is lost.
When I onboard users to my Matrix homeserver, I point them to a preconfigured Element-Web URL and the Element mobile apps. I also mention that there are other clients out there - a handful experiment with them, most don't.
One discovery problem is this client had three rebrandings, from Vector to Riot to Element. I've noticed users have had a hard time realizing the Element is a client for Matrix - even when they're actively chatting on Element. Usually they just refer to our chat by the homeserver's name.
> One discovery problem is this client had three rebrandings, from Vector to Riot to Element. I've noticed users have had a hard time realizing the Element is a client for Matrix - even when they're actively chatting on Element. Usually they just refer to our chat by the homeserver's name.
What this tells me is that having a canonical primary client with the same name as the protocol would do wonders. This doesn’t rule out third party clients, it just clarifies matters for users who don’t necessarily know what a client is and have trouble conceptualizing the protocol/service/client divide. Those who are technical enough will seek out their favorite client, but for everybody else chatting on Matrix means downloading the Matrix app.
So I've hosted a federated Matrix homeserver for almost eight years now, with a few dozen active users. Beyond the 800 pound gorillas of networking effects and the need to be your own infra engineer (mitigated by things like your integrated server/client idea), there are a few key features that I feel keep people on Discord:
- No distinct voice channels with push-to-talk. Yes there's Jitsi, but it's not the same as the TeamSpeak/Ventrilo/Mumble style functionality that gamers have used for 20+ years. Solve this, and I think Matrix will see much more adoption for casual voice chat in the same manner as Discord.
- No server-specific display names. I have multiple screen names across multiple games/communities, and Discord accommodates that. Matrix (or at least Element+Synapse) does not.
- No path for server-specific invites. If I attempt to invite someone to myexamplehomeserver.net in the Element UI, they are instead directed to make a matrix.org account and are federated in. It works I guess, but it's shit for homeservers with closed registration that want to invite users to that homeserver. Writing an invite bot has been on my "I'll get around to it" list for a very long time now, and I know other homeserver admins have considered the same thing.
- E2EE is clunky. Yes, it is much better now. Yes, there are concerted efforts to improve this. Yes, it is an extremely hard problem. However, many times I have had users lose access to past messages because they were signed in to only one device and don't remember their keys. I understand the security aspirations, the proles do not and never will.
There are a lot of other little things like custom reactions, but I think the four above would do a lot to foster more Matrix adoption.
Mattermost is probably an easier self-hosted Discord alternative.
For $2/month you can spin up an instance on Pikapods, based on their community edition. If you want advanced features (available in enterprise editions) you'll need to figure out how to re-enable them by un-feature-flag-gate-ing the source code and recompiling (and at that point move hosting), or work it into the plugin system (which gives you tons of power)
It's not federated though
Mattermost always struck me as more of a Slack alternative. Though the line between the two was always a big blur.
> It's not federated though
This is a big one for my server - our community has a handful of active "visitors" from other homeservers and the experience is pretty seamless.
OK, I've investigated Matrix enough to suspect it's not going to work for the things I've been using Mattermost for, but I'm curious to hear what your stack is (client and server software)?
It sounds like, if Matrix worked seamlessly (which is not the experience of... anyone(?)... in this recent thread https://news.ycombinator.com/item?id=44617309 ) it would give us federation and better identity access management (compared to the Mattermost community edition anyway) which would be really nice
It also has video calls built in... also really nice, but people are saying since they migrated from jitsi to their MatrixRTC, setup & admin have been a nightmare.
Element X seems to be the more popular client, though it sounds like, while faster than Element, many features (such as threading) are not yet available, and people have been lamenting the extreme sluggishness of Element
It is absolutely possible to have a Matrix deployment that works seamlessly. It's true though that a lot of people are very unhappy at the transition from Element to Element X, hence the recent complaining threads. You can see a response from me as Matrix project lead (and Element CEO) here: https://gist.github.com/ara4n/190ad712965d0f06e17f508d1a45b5.... Meanwhile, we're working hard at landing both threads and spaces in Element X (they're available behind feature flags already).
The only reason people seem to complain about Element Call being hard to set up is that it's new and unfamiliar and doesn't have much docs yet, and VoIP is always a pain from a firewalling perspective. It's a hell of a lot simpler to setup and run than Jitsi, for instance. But if you run it via ESS Community (https://element.io/server-suite/community) or even https://github.com/element-hq/element-docker-demo it should work fine.
Element X shines with large rooms - think joining rooms with 100+ active users and a lot of events to sync. Using multiple clients is not a problem either, I think I have at least 10 active sessions active with my current account. Element largely "just works" for my users with few grumbles, but that is definitely not the case for a lot of others here.
As for my stack, it hasn't changed much since this post: https://news.ycombinator.com/item?id=34784816
I think this ignores an important point: many of them participated in this process. A lot of those idealists turned out to prefer money.
Damn right. I don't hate tech, I fucking love it and I always have. I have a career because of tech. I have tons of friends I met online and continue talking to via tech. I hate tech companies because of what they did to tech.
We need more democratic institutions revolving around tech, too many rich individuals have an extreme amount of control in what technology gets pursued and pushed.
But this is where the disillusionment of this idol continues:
Do you think kids wouldn't develop the same mental health issues if they were on Mastodon instead of Facebook?
Do you think the largest Mastodon instances wouldn't start also harming their users once large enough?
There's nothing preventing the decentralized from becoming centralized over time, from the same network effects, and abusing the users all over again.
In which case, why fight for it? It's pointless; we started decentralized and became centralized; a re-decentralization moment just causes the same economic forces to pull everything together again.
>> In which case, why fight for it? It's pointless.
I fought for years and came to the same conclusion. I just focus on myself now and a few tight knit circles of friends. We still hang out together on Diaspora. We still muck and hack around with stuff like Ubuntu Touch and debate the latest Linux distros and see how long we can go without our smartphones.
A few of my friends have gone "analog" because of what's happened. They were there during the first dot com bust. They've seen what the internet and tech have become - they just want to opt out now. Kind of crazy the times we're living in. Decentralization was the dream - now its just a nightmare a lot of us no longer what to be apart of.
> In which case, why fight for it?
Eternal vigilance is the price of liberty. Among other things left unsaid.
> Do you think kids wouldn't develop the same mental health issues if they were on Mastodon instead of Facebook?
I'm not sure, but Facebook has employed too many PhDs to name with the goal of fostering addiction to their product, as well as behavioral ad targeting, and to my knowledge, Mastadon has not.
> Do you think the largest Mastodon instances wouldn't start also harming their users once large enough?
I don't think user harm is correlated with size of the platform.
> There's nothing preventing the decentralized from becoming centralized over time, from the same network effects, and abusing the users all over again.
Which is why I support things like the AT Protocol, which enable decentralized social networks that can share data amongst each other, and permit users greater control over what they see and from where.
> In which case, why fight for it?
Because I've been fighting for a better internet for decades at this point and see nothing better to do with that time were I to stop.
Hey wait a minute what’s so bad about a little strangling? One of these things is not like the other
Because according to UK research, just doing it once, in a supposedly safe way, causes biomarkers indicating brain damage in the blood. It can also apparently cause strokes decades afterwards.
https://pubmed.ncbi.nlm.nih.gov/40062485/
https://www.theguardian.com/lifeandstyle/2025/jul/07/no-safe...
“There’s no safe way to do it, no safe quantity of blood or oxygen you can cut off from her brain for fun,” says Jane Meyrick, a chartered health psychologist who leads work on sexual health at the University of the West of England. She describes being at a sexual health conference last year where data was presented on sexual strangulation – the prevalence and harms. “Usually, at those conferences, people will be talking about the extremes of what everyone is getting up to in a very sex-positive way,” she says. “When this was presented, you could feel the tension, the internal conflict, in the room, with professionals being unable to reconcile the gap between what they were hearing and their usual sex-positivity.”
Reminds me of old net, Maddox. How to kill yourself like a man.
https://maddox.xmission.com/c.cgi?u=manly_suicide"%3B>...
It's from 2004 and dark humor.
Well shit, that's really unfortunate for a lot of people.
...but sexytime "choking" has an entire range of intensity that is more about intimacy and physical touch and doesn't restrict airflow or blood flow at all.
If a 14 year old watches porn and decides to copy what they see, that is a failure of sex education. Every child needs to be taught fairly graphic and frank realities of sex, sexuality, and sexual activity.
For some reason, this is basically only difficult in parts of the US.
Did you miss the part about how this affecting the UK despite educational warnings? This is hardly US-exclusive, and education is not solving the problem.
> If a 14 year old watches porn
This is not 14 year olds either, these are teenagers and fully grown consenting adults, notably not in the US.
"A survey by the Institute for Addressing Strangulation, established with Home Office funding in 2022, after strangulation became a standalone offence, found over a third of 16 to 34-year-olds had experienced this."
The reality is that people are influenced by pornography, despite education, despite warnings, despite common sense, despite age, despite region, and it is killing people. Even sex-positive educators are stuck in cognitive dissonance on this one.
>If a 14 year old watches porn and decides to copy what they see, that is a failure of sex education
Sex education is just another school class that in goes one ear, out goes another.
The kids will follow what they see (including in porn), what the culture has normalized, not what some teacher tells them in class...
So, "porn normalizes X, but it's the duty of sex-education to prevent kids copying it" is a lousy and losing approach.
Dang someone had that response locked and loaded, 2 minutes later... in any case I did not expect this entire sub-thread under a discussion on decentralized web!
If you’re going to insinuate on Hacker News, that pornography is inspiring behaviors that are literally killing people… you need to bring receipts. Being in denial of harm is the favorite pastime here, but that's just the first stage of grief.
Case in point, the sex-positive educators in the UK... forced to say that certain bedroom behaviors actually are too dangerous to participate in. Even that, yes, we will control what you do in the bedroom, with the force of law, and even restrict depicting that act regardless of consent, with the force of law. Extremely ironic that sex-positive educators are now forced to say what you can do in the bedroom, after they said "who are you to tell consenting adults what they can do" for decades.
> This is the first comment I've seen from you that leverages the UK's authority on anything related to online safety.
Edit for a (now deleted) reply: This has nothing to do with the UK; studies objectively say this activity is more dangerous than waterboarding. If that is true, action is objectively necessary, including against depictions.
I appreciate your receipts; I was just surprised at the rapid response.
The bad part is it's strangling.
Any perversion relegated to a handful of people, now is part of porn "sex-ed" to millions of kids
We're still here. It's just that the mainsteam picked up on our technology, and now it's a bit harder to find us.
In the old days it was easy. Get into computers, and the internet when it was there, and you would for sure only meet nerds.
Now you mostly meet hipsters.
Just at that early time there was this new field that nobody was really much interested in, but some geeky people were, and that is what united us.
And then we discovered we often shared common causes too, like freedom of information and nerdism.
But now our meeting spaces have been invaded by hipsters making use of the great tools we once created to help each other.
Of course this is not fun, but there is one good thing that might come out of this.
The hipsters might have invaded our spaces, but they are still óúr spaces. We created the fundamentals, and the reason for it's success praises our vision.
Maybe that vision will rub off on the hipsters a little bit, and they get a bit of a feeling of what real freedom is actually all about.
It might take a while, and it will still get worse before it gets better, but we might have set the stage for changing the world.
> tech nerds, coders, and hackers
This is the people that build it.
"Oh, somebody made me do it"
Great excuse!
The real challenge is network effects. You can't fight it. If the top dog has a considerable percentage of the population trapped in their platform, then it is already too late.
The only way to get people off the platform is to literally get hired and "trip over the cord" on purpose.. which at this point is likely going to involve very elevated permissions, and a really high chance of jail time.
I'm starting to imagine that diferent currencies, like money are converging with their platforms
each platform having your its own currency? well that's simply one nation. a federal government, registered at the UN. this translates the UN into Etherum, register your currency and be a part of what exactly?
I have to admit, I've always wondered what would happen if say, the old Twitter staff had just completely wiped the servers before they left. Would they even try to bring the site back, or just write off the whole company?
But seriously, a network effect isn't necessarily the end of the industry. Lots of sites and services make major screw ups that drive away a lot of users to an alternative in some way or another, or outcompete a huge incumbent. Like what happened with Digg and Reddit.
Digg was sold to someone that didn't understand the internet, and it was not social, so it was just like HN - not a place to put your "page" down and have your friends follow you there.
Reddit is a bit different - but as far as the data is concerned it has not once dropped in usage and is about double what it was a couple years ago.
Is it possible for any tech to become mainstream without monetization being the fuel? Especially social media
Possible. Very, very unlikely.
What a load of nonsense. The dream isn't "dying", it was never alive: crypto and "Web3" were never about anything other than "number go up" speculation and gambling; anyone who thinks otherwise is either profoundly dumb or in on the con and trying to leave you with the bag.
Very disappointing to see IEEE cover this as though it's an actual thing, would they cover the "dying dream" of Amway knives?
>crypto and "Web3" were never about anything other than "number go up" speculation and gambling
You forgot organized crime