| ▲ | gjsman-1000 3 days ago |
| 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. |
|
| ▲ | snerbles 3 days ago | parent | next [-] |
| 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. |
| |
| ▲ | cosmic_cheese 3 days ago | parent [-] | | 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. | | |
| ▲ | Flere-Imsaho 3 days ago | parent | next [-] | | 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. | | |
| ▲ | Arathorn 3 days ago | parent [-] | | 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? | | |
| ▲ | Flere-Imsaho 3 days ago | parent [-] | | > 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. | | |
| ▲ | snerbles 3 days ago | parent [-] | | 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. | | |
| ▲ | cosmic_cheese 3 days ago | parent [-] | | > 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. |
|
|
|
| |
| ▲ | snerbles 3 days ago | parent | prev [-] | | 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. | | |
| ▲ | pcthrowaway 3 days ago | parent [-] | | 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 | | |
| ▲ | snerbles 3 days ago | parent [-] | | 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. | | |
| ▲ | pcthrowaway 3 days ago | parent [-] | | 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 | | |
| ▲ | Arathorn 2 days ago | parent | next [-] | | 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. | |
| ▲ | snerbles 2 days ago | parent | prev [-] | | 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 |
|
|
|
|
|
|
|
| ▲ | fidotron 3 days ago | parent | prev | next [-] |
| I think this ignores an important point: many of them participated in this process. A lot of those idealists turned out to prefer money. |
|
| ▲ | ToucanLoucan 3 days ago | parent | prev | next [-] |
| 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. |
| |
| ▲ | azemetre 3 days ago | parent | next [-] | | 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. | |
| ▲ | gjsman-1000 3 days ago | parent | prev [-] | | 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. | | |
| ▲ | at-fates-hands 3 days ago | parent | next [-] | | >> 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. | |
| ▲ | snerbles 3 days ago | parent | prev | next [-] | | > In which case, why fight for it? Eternal vigilance is the price of liberty. Among other things left unsaid. | |
| ▲ | ToucanLoucan 3 days ago | parent | prev [-] | | > 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. |
|
|
|
| ▲ | kouru225 3 days ago | parent | prev [-] |
| Hey wait a minute what’s so bad about a little strangling? One of these things is not like the other |
| |
| ▲ | gjsman-1000 3 days ago | parent | next [-] | | 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.” | | |
| ▲ | doublerabbit 3 days ago | parent | next [-] | | 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. | |
| ▲ | mrguyorama 3 days ago | parent | prev | next [-] | | 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. | | |
| ▲ | gjsman-1000 3 days ago | parent | next [-] | | 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. | |
| ▲ | coldtea 2 days ago | parent | prev [-] | | >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. |
| |
| ▲ | password4321 3 days ago | parent | prev [-] | | 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! | | |
| ▲ | gjsman-1000 3 days ago | parent [-] | | 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. | | |
|
| |
| ▲ | coldtea 3 days ago | parent | prev [-] | | 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 |
|
