- cross-posted to:
- [email protected]
- cross-posted to:
- [email protected]
Yeah, this is actually why the Fediverse will shine, there is no central node, and in the time it takes for the government to shut down a node, many more can be set up.
If governments start shutting down instances, phew… I suspect that will be a pretty bleak time to be alive. Or just bleaker, haha. 😑
Sneakernets, where people trade data on USB sticks and MicroSD cards (I remember back in the day reading about people smuggling data in and out of countries in fake coins with a MicroSD inside). Air-gapped LANs (not internet connected) that cover neighborhoods with people managing local web servers and services for the community. I have probably over 20 services running on a handful of physical servers, most of them old, most of them underpowered, but they could serve a small community on a local network easily.
I honestly think that’s where people will end up to escape the corporate internet.
Hmm. What’s your scenario for the “corporate internet” killing off the Fediverse? Deliberate botnet attacks and/or overwhelming the instances with spam / junk bot accts?
I don’t think it’s necessarily escaping places like the Fediverse, but the internet as it exists… I don’t know if there’s a way to “heal” it or stop it from getting worse, honestly. So I very much feel like there’s opportunity for people to “disconnect” while also being somewhat connected. Local websites tailored to community needs as opposed to corporate ones.
I mean, if “Big Tech Won’t Broadcast Your Dissent” and the costs of maintaining alternative communities rise, due to the cost of renting servers, it may just make more sense to make a local intranet with your buddies who all have their own Jellyfin servers too, and that becomes a community resource so nobody in the complex needs a streaming service. I mean I have RomM, which allows me to play tons and tons of old video game ROMs in a web browser, so that could become a community resource, too.
There could be community forums and community weblogs, allowing people to update each other on new developments in the community, and updates to community servers, scheduled maintenance times etc. Sneakernets could share local community updates among different local LANs in the area, and share files via traditional file transfer from one disk to another so other communities can get new entertainment, etc.
Especially with Microsoft demanding everyone have a TPM 2.0 chip for Windows 11 there is currently a lot of shunted hardware that could easily be turned into local server stacks and then the only cost is replacement parts if things go down and the regular cost of electricity and maybe a handful of uninterruptible power supplies, and that’s only really if you’re on a system that power loss could cause problematic data corruption.
TL;DR: Seizing the means of communication may be just as important as seizing the means of production. In this case we don’t “seize it” we make our own parallel systems.
Oh I forgot to include this:
https://loriemerson.net/2024/08/31/a-brief-history-of-barbed-wire-fence-telephone-networks/
This is piece on the history of Barbed Wire Phone networks which were an early way to roll out local phone access in remote areas without rolling out much infrastructure. It essentially would be an “always on” community line. No ringer, you just pick up your local receiver, and see if anyone’s on any of the other receivers to talk. So if you had 3+ receivers connected to the barbed wire network, all of those would be able to be part of a big community phone call.
This was actually was initially got me thinking about these ideas and in particular how especially dissenting protest groups must stop trying to communicate using the tools of their oppressors. Secured messaging isn’t enough, we need to have messaging entirely removed from the wider, corporate and government controlled internet. Build parallel communications systems and begin to use those instead.
Oh, I also forgot to include LoRa texting for remote areas as well:
The point being that all these rely on building out our own networks instead of relying on all the privately owned fiber, cable, dsl, and ethernet that exists between us and our destination on the internet. We can’t avoid passing through someone else’s property on the wider internet, which makes it more at risk of interception and possible decryption. Removing it from the wider internet and closing the loop tighter resolves this.
Every “hop” to a new server online between you and your destination is a place where your data has been replicated to be able to pass through. Parallel networks avoids this entirely.
I have thought about airgapped networks for a long time, there is nothing to stop peoplw from buying networking equipment and setting up a secondary network that is completely physically separated from the normal internet.
Used enterprise networking and server gear exists and is fairly cheap.
It doesn’t even need to be enterprise, honestly.
I run a rather significant local network with off the shelf consumer stuff and OpenWrt.
How do air-gapped LANs that span neighborhoods work?
A lot of managed switches, a lot of ethernet cable, maybe some point-to-point WiFi connections for longer distances, and local DNS and DHCP servers that have failovers for redundancy. You technically don’t need a router because routers are for connecting different networks, and if it’s not connected to the internet, as long as you have local DHCP and DNS, you’re kind of golden for local services. You can even have internal TLS for local security. I do that with Caddy, it’s dead simple, and the certificates are locally generated for a 10 year period. I’m connected to the internet but I don’t expose any of these services to the internet, just my local network.
With a little more work, you could set up a VoIP server too so people could connect their phones that don’t have cellular service and use a VoIP app to be able to call anyone in the neighborhood by their VoIP number.
We are headed there, sadly.
I don’t know… as long as you have corporate social media, I’m thinking most people will remain there out of convenience, precedent, and buy-in. Which would leave places like the Fediverse small fry that probably wouldn’t be going after statistically. Also, the decentralised aspect means that you’d have to get much closer to a 1984 situation for this to be a serious concern, with instances currently spread out all over the world.
But yeah, I agree that “1984” is certainly on the near-horizon of strong possibileries.
I think you’re far too optimistic. There very much are central nodes: DNS registrars, ISPs and data-centers. The prerequisites for participating have increased. Domestic Australian providers now require copies of government-issued IDs (copies which then make their way into America’s private-intelligence spinoffs like Palantir, because it’s these spinoffs that are selling verification services). The West is expanding its capability to surveil domestic and foreign populations. To participate on the open-internet is to identify yourself, provide documentation and have government- or private-databases verify your identity and to pay perpetual gratuity rents.
A thousand virtual-servers hosted on Microsoft’s or Amazon’s computers. That’s the reality. It already feels as if Cloudfare is MITMing everyone. The internet has become more centralized than it was and that’s a trend that will continue.
It seems to me absurd that the majority of a developed-nation’s citizens carry around pocket-computers that are connected to the internet 24/7/365 yet none of them are allowed to host a blog on their computers (pocket or otherwise). (The reality is your mobile-computer is forbidden by economic-rent-seekers from talking to another mobile-computer even though they may both be in the same room; they force your data to transit over their infrastructure and gate-keep it so that the other mobile-computer is forced also into paying rent so that it can retrieve your data from the economic-rent-seeking entity [that may be located in another country; so that message you sent to a friend may travel half-way across the world even though their computer is three meters apart from yours! What a farce!].) Consumers are behind CGNATs and are barred from hosting by default. I was optimistic about IPv6–everyone could have a unique public-IPv6-address. The market for static IPv4s should have collapsed. It’s still here. It’s still demanding perpetual rents.
I’m an advocate of OpenNIC–a domain should not require a perpetual rent be paid to a registrar. The foundation’s of the internet are centralised. To seek and agitate for de-centralisation is correct, but it’s counter-productive to redeploy the same centralised software that got us to this point we want to move away from. That software was designed for a bygone era and it’s been taken over by corporate interests. Capitalism is doing what it does best: consolidating! You’re not going to de-consolidate by taking their software. We need a new internet. We need GNUnet.
GNUnet’s still juvenile, but I do encourage users to deploy and play with it. We can get to a better internet that reflects all of us. Computers have become locked dumb-terminals for ad-networks and tech-oligarch mouth-pieces. They constrict the user absolutely whilst using your hardware, electricity and data connection to surveil the mobile-computer’s environment and your interactions with it. If you dislike that future you need to participate.
I2P is a mature project and is ready to deploy now. There’s two implementations: one in Java and one in C++.
You need to pay perpetual rents to host on the old-web: you need to pay rent to a DNS-registrar and you need to pay rent for a publicly-routable IP address; you’re also coerced into paying rents for compute-time in a data-center (it’s often cheaper than paying for a publicly-routable IP address which are priced for businesses). These are economic-burdens that exist purely for rent extraction. You can host your own Lemmy instance over I2P (and/or GNUnet) on your own machines on the data-connection you already pay for right now. The additional cost is $0.00 (excluding the increased compute-burden on your machines). If you data-connection already grants you a public routable address you can use OpenNIC for a free or self-administered domain that is accessible over the old-web. The software exists. Users need to deploy it, use it and agitate for it.
I will be using solar-panels, batteries and arm CPUs running GNU/Linux to host content over I2P and GNUnet. The compute becomes free. The only cost that remains is the nominal amount required for a consumer-connection. Using software like I2P and GNUnet on your own hardware on your data-connection–that you’re using to read this–is the closest thing we have to “no-barrier” self-hosting or decentralising. The reality is that’s a significant information barrier and I am surprised that no-one has made a mobile-computer ‘app’ for self-hosting blogs/forums that runs a I2P-router in the background for non-technical users. I type that… but I know why it probably hasn’t been done (if it hasn’t): Google’s Android will kill any long-running background process… and Apple would not let you publish it.
edit: this was a duplicate
Google’s Android will kill any long-running background process… and Apple would not let you publish it.
tbh portable mobile phones are not the right place for running public services or an i2p router. an i2p router produces a constant traffic, and mobile data plans have a monthly traffic limit. you can’t afford having 4 TB traffic a month on mobile data. also its a waste to use wireless networks for it, because of the increased power consumption and because it is taking airtime from other devices, with wifi too but even more so on the cellular network.
I’m always available to be bashing google, but it makes sense to kill most long running things. smartphones only have a day long or more battery life because they kind of suspend themselves frequently. if you are running a blog on your phone, it will either get drained more quickly, or your blog won’t be accessible most of the time. but also it won’t kill all apps, you can make exceptions. likely depends on the brand, but on my phone I have several long running apps that are allowed to run freely. one of them the ntfy app, which is capable of receiving my notifications quickly, and I don’t have google services installed.
I prefer the term mobile-computer because it’s a more accurate description of what it is. “Smartphone” is a marketing-term that immediately limits users’ perception around mobile-computers. Your post is a reflection of that.
One of the qualities of libre or libre-inspired -software is that it gives users complete control. That control is usually restricted behind technical knowledge though. Thankfully the developers of the Android port of i2pd (hosted on f-droid and github) have created a GUI where users can set limits arbitrarily. The user can even disable transit for their router, but this is parasitic and I encourage all operators to contribute what they can. The notion that you shouldn’t run i2p, i2pd or other such software on a mobile-computer is… nonsensical when you take it out of the consumer-smartphone context.
Mobile-computers trend on being more energy-efficient with larger battery capacities over time. If we’re to take your concerns about data and energy scarcity seriously we need to address the CPU-cycles and data that are forcefully wasted against the users’ wishes. We should tell Google not to transmit any data from the mobile-computer that’s related to ad-networks. We don’t want that and it’s a complete waste of CPU-cycles and scare data that users have to pay for. Likewise we will prohibit Google from transmitting ad related data to the mobile-computer. Imagine wasting wireless networks by transmitting ads in high-resolutions and wasting CPU-cycles to decode it!
The reality is this: mobile-computers are incredibly sophisticated and capable. When corporations abuse your mobile-resources it becomes the norm and is not even questioned. What I suggest is users take control over their property and use libre or libre-adjacent -software to redploy those otherwise stolen resources towards things you actually want. You may want a lot or a little. If you dream big pack a few batteries.
Google does not honour that flag that ‘grants’ permission to specific processes to be long lived. It will kill it.
The notion that you shouldn’t run i2p, i2pd or other such software on a mobile-computer is… nonsensical when you take it out of the consumer-smartphone context.
it really is not. it sounds like you skippped reading my comment after 2 sentences. sure run i2p with transit on a normal, corded computer, with wired network if at all possible, that’s a good thing. but not on a “mobile computer” with a very constrained battery and mostly wireless only connection. unless you repurposed it as a server, and not using it as a smartphone anymore, then at least it doesn’t have to power the screen.
Mobile-computers trend on being more energy-efficient with larger battery capacities over time.
my phone’s battery has lower mAh capacity than my former one I bought roughly 10 years ago.
If we’re to take your concerns about data and energy scarcity seriously we need to address the CPU-cycles and data that are forcefully wasted against the users’ wishes
obviously, if a user was scrolling facebook all day instead of running an i2p router in transit mode, that would drain the battery quickly too. that is a choice to make, whether to scroll that much or not. compared to that, an i2p router would not need the screen to be on, but it would have constant CPU and wireless network activity, not letting any of these go to low power mode.
We should tell Google not to transmit any data from the mobile-computer that’s related to ad-networks.
which is not a constant activity, but a bursty one. for every x seconds of not transmitting anything, it’ll only transmit that data for less than a second. it does not keep the cpu and wireless adapters awake constantly. neither does facebook in the background.
Imagine wasting wireless networks by transmitting ads in high-resolutions and wasting CPU-cycles to decode it!
I agree in part, but you don’t have a sense of scales. how much resources consume this activity and that one.
The reality is this: mobile-computers are incredibly sophisticated and capable.
capable, sure. except the battery. i2p does not need much resources, but the hardware still needs a lot of power compared to what the battery can store.
When corporations abuse your mobile-resources it becomes the norm and is not even questioned.
but even they can only abuse it when the screen is on! when its off, their capabilities are more limited.
What I suggest is users take control over their property and use libre or libre-adjacent
I am running a fairphone with a custom rom. I don’t play games and the only social media app I have is for lemmy. my battery lasts for a day, but if I had a transiting i2p router in it, properly set to be allowed running all the time, it would not last 4 hours.

You’ve gone from stating i2pd on Android would use 4TB of data to a mobile-computer would last only 4 hours if an i2p-router was running. Just state clearly: you’re making up shit. You haven’t used the software and you have no interest in it. The above is a screenshot of i2pd running on a Pixel 6 Pro under GrapheneOS. In 6 hours i2pd has used 94.31Mb. In a month with 31 days that would extrapolate to 11.69444Gb. How much data you’re willing to ‘donate’ is entirely up to you as it’s configurable. My internal-battery is full at 80% (charging optimization enabled which halts charging at 80%) whilst the external battery is at 66%. Bluetooth has been on the whole time sending audio to a headset. This mobile-computer also runs Briar non-stop.
I stick an external-battery to the back using hook and loop tape. It will last more than 24 hours. I honestly don’t know what you’re trying to say. People doomscroll on Tiktok all day and I’m pretty sure that’s just video content, probably hi-res too. No one’s pretending data is precious or limited except for you. (Australian retailers sell unlimited mobile-data plans.) You don’t want to use i2p on a mobile-computer? Ok. Other people do and I’m one of them. As I already stated in the previous post: if you actually want to do things with your mobile-computer pack batteries. Your imagination’s the limit.
I administer a cluster of irc-nodes that talk to each other over i2p. Some nodes live on mobile-computers whilst others are on laptops or desktops. Decentralised chat hosted on your own hardware using your own connections. No need to pay perpetual rents for whitelisted IP addresses or a domain. You get encrypted communications with anonymization built-in that supports both TCP and UDP. It’s great.
Australian retailers sell unlimited mobile-data plans
in eastern Europe unlimited mobile data plans are pricy, it’s not too common to have them.
This is something that doesn’t make sense in the Australian market. Consumers are lazy here. Felix sells an unlimited plan for $40.00 AUD a month. Other retailers are selling 30Gb a month for $44.00 AUD. I think the cheapest seller here is Aldi who resells Tesltra-capacity for $28.00 AUD a month with 12Gb of data. The price isn’t exceptional, often being cheaper than the dominant sellers.
You’ve gone from stating i2pd on Android would use 4TB of data
surely not in 6 hours you dummy! over a month, which network providers use as a billing cycle! my i2p router on my server is up almost a month, and in that time it has had 4 TB traffic, in both directions. so my example 4 TB was not even correct.

I have been running this i2p router for the past 3 or 4 years, yes I do know it! it always has this much traffic. would likely have even more if I had a better upload bandwidth and a higher limit set.
but I have to admit I forgot about the bandwidth limit setting.how many mAh is the capacity of the external battery?
The ones I use for mobile-computers are 10,000mAh batteries. At these sizes they fit nicely on the back and the mobile-computer can still easily fit into pockets.
restricted behind technical knowledge
Knowledge is power.
I think there are barriers to this.
- Getting people to know about the fediverse at all
- The UX of the sign-up is (or at least was if it’s changed in the last few months) confusing and bad – what is an instance? How do I pick one? What are the consequences? etc.
- If an instance gets taken down suddenly, depending upon how that looks, the user has to find another and sign up again, and then start over not necessarily under the same name (and a bad actor could purposefully start taking, for example, an successful organizer’s name on other instances to sow confusion or at least block them from being identified as quickly/easily).
We figured shit out in the 90s. People will figure it out now. People are figuring it out now. I come here more than anywhere else, and there is always new activity.
90s people are different from 2020s people entirely. Not a good way of looking at it.
Every older generation says that about every younger generations, this is nothing new.
I don’t know that the demographic that is primarily, if not only, reachable via meta/xitter is at all incentivized to do that, though. I know even less about “digital natives” whom I presume have different experience around that. Certainly, I hope they do, but that doesn’t change the fact that making it more visible and with a better UX would improve and accelerate adoption. I conversely find people opposing that because something about “normies” (a word I loathe, to be honest).
I conversely find people opposing that because something about “normies” (a word I loathe, to be honest).
Same, I don’t disagree with you there. I’m not trying to gatekeep anything, and I think that even a slightly more polished UX will make a vast improvement.
If an instance gets taken down suddenly, depending upon how that looks, the user has to find another and sign up again, and then start over not necessarily under the same name (and a bad actor could purposefully start taking, for example, an successful organizer’s name on other instances to sow confusion or at least block them from being identified as quickly/easily).
that also applies to email. and centralized social media too.
If an instance gets taken down suddenly, depending upon how that looks, the user has to find another and sign up again, and then start over not necessarily under the same name (and a bad actor could purposefully start taking, for example, an successful organizer’s name on other instances to sow confusion or at least block them from being identified as quickly/easily).
that also applies to email. and centralized social media too.
the crimes of ice community alone makes the fediverse worth it for me.
I wish you were right.
The central node is world and ml.
Extremis or radicalised behaviour is increasingly normalised, sooner or later someone will cause harm and visibility and this utopia will end.
We are shrinking for months on end now too.
Cut off one head and two more shall rise to take its place
The revolution will not be televised.
But the apocalypse will be! On intergalactic TV!
It’s always funny to see this comment and then read the replies that don’t get the reference
deleted by creator
The revolution will not be hosted on AWS.
The first thing I want to say is…mandate, my ass.
Most big tech companies are the new oil and railroad robber barons and need to be dealt with yesterday. This is fucking out of control and they partially rule the country not the people. Break up big tech! Bring back the trust busters!
The new plutocratic class are exponentially richer than the oil and railroad robber barons, and are exponentially worse, though some like JP Morgan were monsters. (I really wonder how a bank unironically names itself after him.)
The aristocrats of the 21st century don’t even hide their world-domination / autocratic rule / global genocide intentions and fantasies.
Also their ill-gotten gains are not merely from ruthless anticompetitive business methods, but from financialization, which is rent-seeking jacked on crack cocaine. They didn’t actually produce anything but sinkholes in the global economy.
And they have way better technology to control and contain the masses when they dissent.
This may be our last chance to end the owner class. With water running out, it may be our last chance anyway.
The revolution will not be televised
Rest in power, Gil!
Fediverse needs to work on onboarding if they want the job.
There should be “default” instances with no local posting. People can browse all and experience instances first before making their decision.
Plus the ability to transfer your account to a new instance without making a new one.
Or instances flag themselves as accepting new signups and then new users are round robined in.
This isn’t a new problem. We’ve seen centralized cellular service get shut down during protests and populist political actions before, such as during the Arab Spring.
A similar issue is centralized media, owned or captured by plutocratic interests. It was a problem during the Pulitzer/Hearst era through the cold war when General Electric used its advertising footprint on the three major television networks to leverage against the broadcast of news stories it didn’t like.
Decentralization would be great if we could, say, get every thousandth household to host a Fediverse instance.
It does raise a question of whether it would be able to handle all the moderation problems that emerge at scale, especially when hostile actors backed by state or business interests can bankroll contributors with antagonistic input. This is going to get worse with agentic LLMs.
It also raises the question if the Fediverse will segregate itself into ideology silos. We already don’t like tankies and tend to defederate from instances that harbor them. Can we do that if we’re trying to create a unified social media network for the global public? We might have to error to the side of hosting content we don’t like (and focus moderation, instead, on content that causes harm).
If we can make it work, it will serve as a great example of anarchist/ libertarian social models working. If we can make it work.
Wait, you mean they are cancelling the left? Are they…shadowbanning?
Where are all the “ERMAGAWD Section 230!!!” rightwing assholes and FREEZED PEACH absolutists now, I wonder?
It’s our only way out.
Your dissent will be live!
(Presuming the headline is a reference)
Imagine being a fly between these people. Musk looks high as shit and wants a gummie from grumpeh, and grumpeh doesn’t know what’s going on but will try to end anybody that tells him children are off limits.
Edit: they’re all musk to me.

I dunno, Mother Jones.
Good joke, but no. Your Google News may very well be wildly different than other people’s Google News.
That sounds ungood
Big Tech will get it’s comeuppance. Look back in history at all of the greedy robber barons exploiting the latest new industrial age technology, like cars and trains. It all settled down eventually and most of those Snidely Whiplash motherfuckers were never heard from again. People of the Globe…UNITE!
People are complaining about “ew bad onboarding” while it’s the best barrier holding up the flood of very, very irritating people.
People figured harder things before, we do not have to make everything easiest thing in the world - it makes fediverse a quality place actually. A lot of people who join, really want to be there and it’s great.
The problem is we ideally want the fediverse to grow to be the new normal in social media and not just a little club for cool kids.











