216
you are viewing a single comment's thread
view the rest of the comments
[-] PorkrollPosadist@hexbear.net 54 points 1 month ago

They will turn the screws on the hosting providers if (and when) it comes to it.

[-] someone@hexbear.net 40 points 1 month ago

And even if the hosting provider doesn't, the US government can just seize the domain. .net is administered by US-based Verisign.

So what's our plan B? If the domain is seized, I'm assuming that we'll get updates from the fediverse account. But if there's some sort of problem there (matapacos.dog is seized, or no updates get posted because those with access to the matapacos @hexbear account have been arrested, or other unforeseen problem) how do we re-group?

[-] xj9@hexbear.net 25 points 1 month ago* (last edited 1 month ago)

we need significantly more resilient communications infrastructure. though this applies equally to disaster prep since the internet is extremely fragile. i see this as a problem with a bunch of parts, but some of the main pieces look like this to me:

  1. organizing computers (spare phones, spare laptops, actual server hardware) into cloud-like infrastructure components
  2. writing software that actually works well running on trash (fediverse nodes really hate it when hostnames change, so hexbear itself may not adapt well if the domain gets seized. they aren't a good match for disaster services either since you can't guarantee that global DNS will work)
  3. building overlay networks to connect nodes over the internet (yggdrasil+i2p looks like a winner in the short term)
  4. building a physical mesh network to link nearby computers together over fast links (BATMAN for no se vende mesh here in LA)
  5. developing a mobile adhoc mesh routing protocol that can setup a usable internet using only smartphones that interoperates with the fixed mesh noted in 4 (this will likely replace BATMAN, but is also a research problem and would represent a novel capability)

there are a bunch of shitty mesh apps and hardware components out there that may also be applicable, but its hard to get anything useful out of them without more planning and coordination. (meshtastic, briar, secure scuttlebutt, simplex, et al. come to mind)

i've been working on minibase for the micro-cloud component in an experimental capacity for years using alpine linux (just cuz i like it), but I've recently started working on a production version based on nix.

[-] PorkrollPosadist@hexbear.net 11 points 1 month ago* (last edited 1 month ago)

4: building a physical mesh network to link nearby computers together over fast links (BATMAN for no se vende mesh here in LA)

5: developing a mobile adhoc mesh routing protocol that can setup a usable internet using only smartphones that interoperates with the fixed mesh noted in 4 (this will likely replace BATMAN, but is also a research problem and would represent a novel capability)

I think there is a strong tendency to put the cart before the horse when it comes to mesh networking. People imagine how cool it would be to have widespread mesh networks, but setting up a digital radio doohickey accomplishes nothing if there is nobody listening on the other end. We need social organizations first. Then, communications technology can be built to serve the needs of these organizations. It is very much a local organizing problem. The tech does not circumvent this.

[-] someone@hexbear.net 6 points 1 month ago* (last edited 1 month ago)

The other issue of course is that mesh networks only work in very densely-populated areas.

I'm in full agreement on your points about organizing being the real task, not the tech side.

[-] xj9@hexbear.net 2 points 1 month ago* (last edited 4 weeks ago)

WISPs use a lot of infra mesh tech and they're also really great for low-cost network deployments in rural areas too since you don't need to dig long trenches. Just point a couple of high-gain yagis at each other every 30km or so and you're good.

The main problem with MANets is 1. TCP is incompatible with variably high latency and partial connectivity and 2. there's no IP routing algorithm that handles variable topologies well. so you end up burning the majority of the already limited wireless bandwidth on tracking routing tree changes and not on data transmission. Density is only really an issue because of problem 1. ICNs handle the latency issue more gracefully, but without a solid solution for 2 it doesn't matter lol.

[-] xj9@hexbear.net 3 points 1 month ago* (last edited 1 month ago)

That and nobody actually knows how to make mobile adhoc mesh networks beyond toy demos. I think NDN is probably the furthest along based on my reading between the lines of some routing algo papers coming from UCLA (read: .mil applications). I'm personally trying to make micro clouds (or mist computing if I'm being cheeky) a little easier and continuing research into MANets because I think its a really interesting problem that there's very little incentive to solve since the profit potential is tiny. I think a lot of this is applicable for making websites that can't be deleted, but like you say, the social organization comes first. Without the will to maintain the systems, no amount of tools will make a difference.

The infra meshes in berlin and elsewhere are basically volunteer-based wireless ISPs and can't be run without people willing to put in the work. They also aren't that much more resilient to disruption than a typical corporate WISP like starry or whatever, its just theoretically possible to have a community owned/operated setup because the infra is relatively cheap. No Se Vende Mesh is a project that was already happening in LA when I got here and I'm trying to be supportive. Some of the volunteers are struggling with motivation though because, well, its LA and organizing is hard.

[-] Gucci_Minh@hexbear.net 23 points 1 month ago

finally, the dream of hexbear.su will come to fruition

[-] Chronicon@hexbear.net 18 points 1 month ago* (last edited 1 month ago)

lemmy or lemmygrad both have Mali domain names, decent place to regroup if they aren't taken down at the same time. I think the hosting provider is the same though. Hoping the admins in charge of the infra and with access to that account aren't US-based though

[-] REgon@hexbear.net 15 points 1 month ago

I don't know how it works with hosting but we've got both lib.rehab and chapo.chat

[-] someone@hexbear.net 11 points 1 month ago

That's good to know, thank you! I've bookmarked both.

[-] PorkrollPosadist@hexbear.net 9 points 1 month ago

No. We lost lib.rehab See for yourself.

[-] Mindfury@hexbear.net 3 points 1 month ago

Holy shit lmao

[-] REgon@hexbear.net 3 points 1 month ago
[-] PorkrollPosadist@hexbear.net 8 points 1 month ago

There are secondary / tertiary methods of contact. People have accounts on other platforms like Twitter, Matrix, Telegram, Discord, Mastodon, Reddit, etc. There are other watering holes people can regain contact. The community would be utterly fractured, but such was the case when Reddit dropped the hammer on us. I think our past experience served well. Establish a life raft. Gather the troops. Plot a course forward.

That said, we were relatively lucky last time. The regime of censorship has sharpened significantly in the past four years since then. If the US government were to lean on Hexbear's ISP (or domain registrar), a life raft on a service like Discord would fold like a house of cards (not to mention being 100% wiretapped). We would need more robust options. There is the Matrix server, but it is tied up with the same infrastructure. The people who use it have accounts distributed among various Matrix servers though.

[-] someone@hexbear.net 10 points 1 month ago

The fracturing of the community is what I worry about most. Sites and software and services come and go. All that matters is that the community can find each other again quickly.

this post was submitted on 22 Oct 2024
216 points (97.8% liked)

technology

23313 readers
494 users here now

On the road to fully automated luxury gay space communism.

Spreading Linux propaganda since 2020

Rules:

founded 4 years ago
MODERATORS