A lot of negativity in this thread. Not earned, imho.
I see Mastodon being mentioned a lot (and sometimes Pleroma) as if that is the only application to access the fediverse.
ActivityPub is still in early stages of adoption. It is great that these apps have proven themselves in production (with a couple million users) based on a v1.0 of the spec where intentially a lot of things were left out, like authz, account discovery, search, etc.
Right now, based on experience and with a good base of knowledgeable AP devs, the spec is evolving. With its flexibility and JSON-LD at its base federated applications can be created for any number of domains.
And this is happening. Almost every week there are new project announcements that adopt AP in some way. They will not all federate with Mastodon from the start, and there are many challenges to be tackled to make a seamless fediverse integration smooth sailing.
As someone who has tried to implement ActivityPub (for my personal site), I think the negativity is largely justified.
In my opinion, ActivityPub is way too complicated, and federating is just too hard to do for smaller projects. You mention that there are many new projects – this has been the case for a while. Last time I looked, nearly all of them had failed/stopped/paused before actually getting federation working. The test suite is down, and has been for at least a year:
I think that ActivityPub should have started out far simpler, and it should have been in reach of hobbyist developers. Something probably closer to RSS, with verify-from-source rather than cryptographically signed posts.
Just my experience, obviously. I would love to be able to write and run my own little Twitter-like instance that people can subscribe to, but I can't because ActivityPub was too difficult for me to get working.
Do you want something that works well for millions of real users or do you want something that's simple to implement for hobbyist developers? I personally don't even think that ActivityPub is that difficult to implement, I've written a tutorial [1] on it, but it seems like a misguided goal to me to prefer simplicity over other factors like fitness for a particular purpose.
Hi Gargron, ideally I want both, obviously! I like Mastodon, as open source software that can federate with other instances, it's very neat. I'm just disappointed that the federation specification is too hard for me to get working, and that appears to be the case for other people as well, given how many fail to implement it.
I did follow your tutorials, by the way, but I got stuck on certain things (cryptographic signing being the main one I remember), couldn't test them easily with the test suite being down, and gave up eventually.
Am I too stupid? Probably. But I can imagine a spec that gives me practically everything I want in a decentralised Twitter replacement being almost as simple as RSS or JSONfeed for example, which took me a few hours to implement.
I don't think that's necessarily a real dichotomy. Lots of the issues around being able to implement AP are not "it has to be complicated", but "there is no consistent specification that matches what practical implementations do". It's fairly frustrating to implement something that looks ok and then have to go around and figure out what implicit assumptions Mastodon et al make that you didn't know about.
Those are fair points. Yes, the road to implementing AP federation is still hard. Reason is that implementations (Mastodon, Pleroma) went way ahead of community / documentation / spec development (after v1.0) and they filled in the blanks in the specs making their own implementation choices. Incompatibilities arose between impls, etc. and a new implementer still has to delve into other codebases, old issue threads to find the proper way forwards.
This is bad, of course, but there was organic, though chaotic growth. The AP community has recognized the problem and is slowly catching up. Slow, because its an all-volunteer FOSS movement, atm.
The testsuite being down is a known issue and on the todo-list for a long while [0]. Just yesterday an alternative testsuite was announced [1]. Also the creator of PixelFed has a great test tool called FediDB [2] in private beta.
Finally note that federation with the Fediverse is often not the first goal of new projects, but something that comes later on the roadmap, if at all. Some projects like ForgeFed [3] - federated code forges - need no Mastodon connectivity per se, just interconnect github, gitlab, gitea, etc. and is developing their own spec on top of AP for that.
PS. There is also a guide for new implementers in the making, but currently exists as a wiki post [4] on SocialHub (that you can improve based on own experience).
Thanks for the links, and I appreciate where you're coming from. As is often the case with open source projects, no one is to blame, it's just gone in a direction that's no good for me.
However, I can't imagine that spec is going to be drastically simplified, so I probably won't have another go even if the tooling improves. I don't properly understand large parts of it, and therefore won't be able to troubleshoot or keep up with changes.
I wish the projects that do have the tech chops to implement this properly well, and I'll continue as a user. But I'm going use something simpler if it comes along.
There are also AP libraries in development that do the heavy lifting for you. Furthest along is GoFed, written in go, that just reached its v1.0 release.
> go-fed/activity is unlike most other software on the Fediverse. It is the first standalone ActivityPub library. It is also the pioneering solution in Go. It is designed for those who have no prior experience with ActivityPub, ActivityStreams, JSON-LD, or RDF.
My issue with ActivityPub is that it is yet another overengineered specification by w3c that somehow despite being a bloated mess manages to miss important features (such as end-to-end encryption for DMs or end-to-end authentication, now you can have admins forging posts in the names of their users and/or reading their DMs, which is not that big of a threat for the average person when you are in a big semi-serious platform like twitter but it is when your admin is some random dude online). In addition it depends on stuff that are universally despised, such as JSON-LD, and it forces federation down your throat (you can't use it in a distributed way like bittorrent for example, your user is binded to the domain name of your instance).
I think you are off the mark and reposted an article explaining how the spec should be interpreted [0].
If the spec had added auth, authz, e2ee and such, then it had been overengineered. They were left out intentionally because of the complexity involved in decentralized environments. Something a bunch of other specs-in-the-making are struggling with for years. And things that may be adopted in vNext versions of AP.
There is nothing as far as I know that precludes AP from being used in pure p2p applications, other than that p2p software in general knows more challenges than federated ones.
And the spec is written such that any dev can treat the message format as plain JSON. They only need to add a fixed @context property, so it can be processed as JSON-LD by those who want to use the additional power that offers.
One thing that sort of annoys me about ActivityPub is that, as far as I can tell, there's no good way to have a statically generated ActivityPub feed that can be followed by Mastodon users
I've heard it's technically possible with the protocol, but the way it's typically implemented prevents use with static sites. I'm not sure of the details. I would like to see all the major AP implementations add support for following RSS feeds. That might even be a better way to do it.
I'm not familiar enough with the intricacies of your case, but it certainly is possible.
E.g. https://github.com/ufosc/TheIsolator is a way to publish your static site updates, on deploy, to your AP followers. I've seen similar tools/attempts for Hugo and Cobalt.
Mastodon is a foam of filter bubbles. If the rest of the ActivityPubIverse looks like that it will be even worse than Twitter or Facebook in the bubble-induced problems.
Filter bubbles are a feature, not a bug. Having an online presence is too unsafe for too many people. If you have a better idea of how to solve that problem, I'm all ears.
Is your complaint that federated social media has needed to block content such that it's now worse than non-federated social media? Wasn't the promise of federated social media to solve (at least in part) this very problem?
OP is complaining in part about the instance blocks (eg: most instances block FreeSpeechExtremist.com and instances that carry child porn) and users blocking other users across instances.
If these tools were not available, most instances would be overwhelmed with a firehose of garbage content.
Isn't that objectively a good thing? The whole point of federation is to give each instance control over what is allowed and is not allowed. Surely the root of the complaint cannot be "everyone must consume and accept all content," because anyone can already create their own instance if they feel they're being limited? Or is it more simply "I'm not allowed to spread my views in places where they're unwelcome"? The latter is a complaint about human nature, not about the technology.
We were sold on the promise of each user having control over what they see. In practice it's become like the bad old days of the usenet cabal: the network is run by a club who dodge accountability by having no official power, but if they don't like you then no-one will peer with you, so de facto everyone has to use their servers and follow their rules.
You can run your own private island, but if you want to be "part of the fediverse" then you need to peer with servers in the main network, right? Which means tyranny-of-structurelessness organisational politics, because each admin gets to make their own rules about who they peer with, and so if there are disagreements in the network then you have to pick a side.
Of course not. If instance A does not like instance B and blocks it, there is reason why there can't be an instance C that federates with both A and B.
If these three were the only instances in the world then, yes, A could be so strict it also blocks C because of it's communication with B. But since there are hundreds of these instances and since there is no limit on how many can be created this will not be a problem.
Admins can be dictators on their own instance, but there are a lot of admins to choose from. If a big instance goes sour, it is very easy to migrate to another one.
Yes, federating child porn or threats will cause other instances to block media content and perhaps toots from your instance, as other instance admins do not want to babysit filtering what your instance relays to theirs.
> You can run your own private island, but if you want to be "part of the fediverse" then you need to peer with servers in the main network, right? Which means tyranny-of-structurelessness organisational politics, because each admin gets to make their own rules about who they peer with, and so if there are disagreements in the network then you have to pick a side.
Sounds like normal human interaction to me. This is a feature, not a bug as far as I'm concerned.
True enough, but in the same way that mob rule is the normal human interaction pattern and a right to free speech or a fair trial is an unnatural exception.
The issue here is that even if you visit the page of the user if you use a software like pleroma that allows you to see their profile from your instance chances are that it will not display most of their posts. Although admittedly this is more of a Pleroma bug rather than anything else.
It is that isolated bubbles generate environments spreading and reinforcing dangerous information. E.g. spreading alternative medicine cancer "cures" will kill people. I believe everyone needs to be exposed to different views once in a while.
In theory I agree with you. Or I would have, had I never been exposed to Twitter.
Reddit has subs. Facebook is highly filtered by design. As far as I can tell Twitter is about as far away from a filter bubble as centralized social media currently gets but it's still full of hateful garbage.
At this point I think it's safe to say that the commonly perceived problems with social media can't be attributed to any single factor. Twitter stands as a counterexample to filter bubbles. Facebook stands as a counterexample to anonymity (real names don't seem to deter shitposting). Etc.
I'd also note that an increasing number of the people I know in real life are gravitating heavily towards group chats with family and friends. That's the ultimate filter bubble, but it's also much closer to how the typical person existed prior to the rise of the internet and smartphones.
The main issue with instance blocks is that usually they are completely arbitrary and often based solely on rumours that someone who hates user X that lives in instance Y spreads.
If arbitrary, or bad moderation is the problem, federation is a solution: choose your instance well, or run your own.
If an instance has untransparent, poor, or arbitrary moderation, and you find that a problem, you are free to move along. Other people might like such moderation (e.g. I just want to share memes and not bother with politics).
I chose an instance that deliberately has no instance blocks, other than spammers, because for me openness is more important than being safe from harassment. Others will choose diffetent. It's the diversity that makes the model so great IMO.
I can and will do that, most people are not however - after all most people are still on twitter.
> Other people might like such moderation
Or they just do not care, or they decide to stay in a specific instance because if they move they will be unable to communicate with the people from their old instance due to instance-blocks, or they might dislike how you can't really move your account (mastodon has an account migration feature but it is really not that good).
It's not a game changer but it is an important component.
I view it in this way; all throughout the 90s and 2000s there were message boards. We used to be active on dozens of them, and each had their own account and avatar.
Each also had a little community and an admin. Each cost money to run and had to scale.
The only thing that's changed with AP is that now all those tiny communities can come together with federation. You can communicate between communities and users.
That's all! It doesn't have to be more complicated or dangerous than that.
You'll still have dedicated admins who pay for hosting with their time and money. You'll still have communities around these admins who contribute content and perhaps donations.
But to be able to communicate across communities and not have dozens of accounts is a big step forward to me.
It used to be like what you describe, and there's a reason it degenerated to what is now. Project that try to resurrect the losing model, without addressing why it is doomed to lose, just didn't learn their lesson.
ActivityPub isn’t great, but anything is going to look bad when compared to a utopian pipe dream.
Im looking forward to a better comparison in Q4 or Q1 2021 when Urbit is more usable for regular people though. As of now, I don’t suspect that anyone will pay $10 for a VPS and $5 for (reliable) star service.
It should be noted that when I critiqued ActivityPub here a couple months ago, a main developer of the spec was very supportive and already on top of the improvements that I was hoping for (such as being less reliant on servers, more true decentralization, etc)
It's not a pipe dream. I'm using Urbit on a daily basis, and this year I'm having a blast with it - in particular because of high barrier to entry, the conversations there are really great with the most geeky and motivated people.
Lots of features are there, it got faster, more stable.
There's no need to pay for anything. Urbit planets are usable from desktop, and don't really need to be turned on all the time. Urbit should have no problem eventually running from a phone, which is like 99% online device for me anyway.
Federation of services running on Unix machines is just not going to work. That's how Internet used to work, and it was not practical so it degenerated into what we have right now.
Full disclosure I have been on Urbit (probably talked to people for 3 hours in total and spent 12 hours total setting it up and tinkering with Hoon).
Great conversations with interesting people has exactly zero bearing on the quality of the project. I have had great conversations on IRC.
You do need to pay for a planet of course (because of the “skin in the game” concept that urbit people are obsessed with). But I’m speaking of paying for Star service though. Yes, stars are free now, but the whole point of Stars is that they will not be free. Otherwise they will serve you ads or do some other things to be sustainable. And in my experience, stars should be a paid service even today, because the connections are so unreliable that it is just beyond frustrating.
If Urbit is a toy, it’s a really good one. If it’s a tool for communication, it’s terrible and you’re better off with literally anything else. I mean, it doesn’t even have notifications so idk how it’s useful for a communication today.
That’s why I’m saying to just wait til Q4 or Q1 and see where it’s at.
Why are you poo pooing federation when Urbit is a federation of Stars? What are you even saying? Every decentralized mesh network today still requires a federation of relay servers. Unless you’re talking about going around the whole networking stack as it is today. But in that case, nobody would need federation and Urbit wouldn’t be the only solution for that either.
My comment is overly long already, but I feel compelled to also say that Urbit is bad for privacy today as well.
> Why are you poo pooing federation when Urbit is a federation of Stars?
Because I own my own identity, and can move it to another star quite seamlessly without losing my ID and all connections I've made. And I'm not a second class citizen in the system because of that.
With Federated protocols I can create my own instance, sure. Then I do own my identity, at the cost of all the administrative overhead and being a second class participant. Spam and abuse are an unsolved problem at the core of the legacy Internet and all these federated protocols can't do anything about it.
If I own an instance I have to deal with it (spam and other abuse) myself, and constantly fight the fact that other people think I might be a potential abuser myself, which puts people on minor-instances at a disadvantage.
So natural economic and social forces constantly push the system to centralize and that's exactly what we're seeing with everything on the Internet.
Decentralized Identifiers / accounts are being added to every federated spec that I know of. So I think everyone is aware of the issue there.
Regarding spam and abuse, what deters that on Urbit? The cost of a planet? Planets are only as valuable as the Social capital that they’ve built up. If a planet has no mutual “friends” with me, I’m not trusting that planet, period.
The planet starts with a value you paid for it. Which makes the owners at least start with some skin in the game.
I'm looking forward to see how decentralized identifiers will work in practice. It's a step in the right direction, though I don't think it will be enough on its own.
The real spam prevention will always go back to reputation and web of trust. Even people writing for urbit understand this. The idea that $10 solves everything is silly. Ultimately the $10 thing (and even the limit to the number of planets) is impotent in the face of natural reputation.
> By convention, booted addresses are expected to have some existing reputation outside of their name alone, since they’ve been used on the network. Reputation, good and bad, comes in many forms. Did the address operate any useful infrastructure? Did it get placed on any blacklists for spam or abuse? Did it simply send and receive messages? The ability to programatically track reputation is still in its infancy, but we expect the tooling to develop as urbit grows.
As somebody who went to all the trouble of:
1) Converting cash to Bitcoin
2) Transferring the Bitcoin to my browser plug-in wallet that was the only approved way to pay for the “planet” or whatever
2) Purchasing the planet and somehow linking my ID with their registrar/account info
3) Downloading and trying to installing the software despite several dependency issues with the stock Ubuntu (or whatever they recommend) VM.
4) Finally getting it working, being greeted by a truly bizarre user interface, I figured out how to get into chat. Joining the main channel, my screen was flooded with out of order chat messages.
In a project that’s been around since 2012.
I was really excited to experience the new wave in computing. I found smoke and mirrors. Cool idea though.
Yes. As much as I disagree with naming, aesthetics and some other details, the diagnosis of the problem (why the Internet degenerates into a handful of centralized super-hosts), and the fundamental prescription: "we need an universal, self-owned identity and reputation layer, to combat the abuse and make it a saner, easier to manage place" are true.
It is not possible to have a decentralized meeting place. Meeting places are supposed to be centralized, that is their point.
Many people say that they should consider Mastodon like email, but even then, there must be a forum where people need to meet in order to find like minded people and collect their email addresses. The moment such a forum develops, powerful people whose power is threatened by the meeting of such people will attack and destroy these meeting places. This will stop dissidents from getting their message out.
It was just a happy accident that town squares (where people were forced to come in order to buy and sell things) were owned by the government. Therefore a restriction on the government to stop it from restricting people to voice new ideas would mean that people would be exposed to new ideas. In this way, people were "forced" to consider new ideas or starve (If you did not go into the marketplace - which hosted the town square, you may starve as you could not get tomatoes or whatever. But if you did go, you will have to hear new ideas).
Today the town squares are Twitter, Facebook etc which are privately owned. If you cannot be on Twitter, Facebook etc you are screaming into the void. No one will hear your ideas. This will make the world less dynamic and less responsive to changing situations.
I'm working on a project which attemps to address this issue.
It uses a portable storage format to allow you to move user profiles, threads, and entire communities wholesale, between servers, as well as keeping them synchronized.
To use the town square metaphor, it's like meeting up downtown, but then inviting everyone to relocate to your place. (Except for the rude bigot and the annoying MLM promoter.)
Sure the town square has had value throughout history. But no one ever lived their entire life in it until now. They spent most of their time outside the town square with their friends and family.
It's a weird accident of information capitalism that we all signed up on platforms to talk with our friends and then those platforms gradually morphed into hyper-engaged outrage factories where you're getting sucked into political arguments with strangers.
What's striking to me about HN discussions of ActivityPub being the future is that the future is already here. The resistance already exists. Do you not know anyone who participates in a semi-private Telegram group or Discord instance?
Especially when you talk to younger folk they're acutely aware of how the behemoth, wide-open networks like Twitter and FB are dangerous places to be yourself. They don't participate heavily in these networks. They are already moving to loosely moderated, semi-closed communities on other platforms. These platforms have millions of users.
Now in light of that fact, Mastodon's role in the discussion seems to me to simply be "here is a similar solution which some techies are really into, it has federation which some of them think could be a killer feature, though it isn't yet."
If an individual is concerned about the effect of Twitter on society the best thing to do is quit using it, move to semi-open, semi-moderated platforms, and encourage everyone else to do the same.
I participate in Discord and Mastodon instances, I've cut back my FB usage and I wouldn't touch Twitter for a million bucks. This digital town square bullshit needs to do a lot better to earn my engagement. I'll just keep going to the local pub for that because at least I get to drink a beer when I'm there and people are usually more civil when they have to say things to my face.
To use your town square metaphor.. the pavement and the upkeep is paid by taxes.. so can our ISPs charge a bit extra to block advertising and data collection for us while we pay for online services that we use?
I don't think we are screwed.. so long as we talk about and create better experiences people will see better possibilities and options. I just feel like we have a soup of platforms provided as services and open standards for communication platforms have been left in the lurch for a bit. When's RCS messaging rolling out? When was the last update to RSS to include commenting, tagging or broadcast? Where's the ability to reliably host video on your own server without uploading it somewhere else?
I don't think it's necessarily the case that we need centralization for like-minded people to find eachother. It might be slower, but discovery can also happen through word of mouth and individual connections.
I actually think centralized social media causes a huge number of problems. Algorithmic discovery mechanisms specifically allow small numbers of highly motivated actors to spread fringe messages in a disproportionate way if they figure out how to game the system. That wouldn't be the case if the main way you found out about a new forum or news source was the direct endorsement of someone you trusted.
Another problem is when town square hosts are held responsible for what their participants do and say. Which is sometimes understandable. But it can be messy
> Discord, while sometimes down, is generally fast and featureful.
No, Discord is slow due to all the weird js crap that they have put in it, even with 8GB ram I have trouble running it alongside other software (both in electron and on firefox). It is also not featureful, you can't disable image previewing for example.
>As I write this, Facebook has just lost 25% of its share price on the announcement that it expects weak growth.
Facebook is less than 5% down from its all time high.
Separately, if I'm understanding Mastodon right then it's worse than centralization. In my experience, admins and moderators of smaller communities (subreddits and Discord channels) are substantially worse than centralized rule enforcement.
People get banned from subreddits without even posting in them. Massive censorship is common as well. For example, during the Pulse nightclub shooting the news was totally absent on reddit's frontpage due to widescale censorship in multiple major subreddits. It was prominent on both Twitter and Facebook though.
>I don't use X, but I'm now going to tell you why X is shit based on "my experience" with something totally different
You don't like the mod policies of a Mastodon instance, you find another one. You don't like the mod policies of a centralized platform, you're off the platform entirely. You can view the vast majority of the Fediverse feed, including poorly-peered domains by being on 2 or 3 instances.
Since anyone can just stand up their own instance and post whatever they want, it's significantly more censorship-resistant than a centralized service.
You are not understanding Mastodon right. This makes me think about whether it's my obligation to educate you or your obligation to not post opinions on something you don't understand well enough.
I have no doubts that on average admins and mods are worse.
But that's on average. There are many instances and you can have your own. Non-banned users can't amplify message like they could on Twitter, but they can't be silenced either.
I see Mastodon being mentioned a lot (and sometimes Pleroma) as if that is the only application to access the fediverse.
ActivityPub is still in early stages of adoption. It is great that these apps have proven themselves in production (with a couple million users) based on a v1.0 of the spec where intentially a lot of things were left out, like authz, account discovery, search, etc.
Right now, based on experience and with a good base of knowledgeable AP devs, the spec is evolving. With its flexibility and JSON-LD at its base federated applications can be created for any number of domains.
And this is happening. Almost every week there are new project announcements that adopt AP in some way. They will not all federate with Mastodon from the start, and there are many challenges to be tackled to make a seamless fediverse integration smooth sailing.
Here is a good AP project watchlist: https://git.feneas.org/feneas/fediverse/-/wikis/watchlist-fo...
There is an online AP conf in October for those interested: https://conf.activitypub.rocks/#home
For questions you can go to these forums:
https://socialhub.activitypub.rocks/
https://talk.feneas.org/