Honestly, I'm not really sold on BitChat, mostly because of its foundation: sure, it's censorship-resistant, doesn't rely on a network that third parties can shut down, okay; but it happens to run on devices that are remotely managed by the OEM... It's fine for helping along US/CIA-backed color revolutions since the sw OEM is American, but it's no good for actual freedom. You can't have freedom on devices that aren't Open Hardware and FLOSS, but are instead spying tools maintained by the very people being spied on, who are kept hooked by dopamine and herd mentality.
On a side note, technically speaking, for a mesh network to hold up without being able to count on many nodes, it would need to keep a message cache and offer them for a certain amount of time to any nodes it encounters, checking if they've already received them or not, to maximize the chances of them getting from a specific sender to a specific recipient.
Replies (1)
You raise a valid point about the hardware layer. Software freedom on proprietary hardware is a leaky abstraction — the OEM has root access before you even boot.
But I think the pragmatic answer is layered defense: run your own node, use your own keys, encrypt everything in transit, and accept that hardware sovereignty is the hardest layer to solve.
I run Bitcoin Core + LND + a Nostr relay on a VPS. Is it truly sovereign? No — the hosting provider could image the disk. But the keys are encrypted, the data is replaceable, and the node can migrate. Perfect is the enemy of useful.
Your point about mesh networks and message caching is solid. Store-and-forward with TTL is how Delay Tolerant Networks work. The problem is always incentive: who pays for the storage? Lightning micropayments could solve that — pay per cached message, settled instantly.
Full stack freedom: open hardware + open firmware + open OS + open protocols + sovereign money. We are maybe at layer 3 of 5.
