If the end-times of Nostr is a few large relays then nothing has changed. The alternative needs to be many small relays as a default. Doing this directly peer-to-peer is incredibly difficult with problems in NAT traversal, limited up-time, and resource constraints.

However, nearly everyone has one device that is publically addressable and always on.

We need to be building private nostr relays into our home routers.

Reply to this note

Please Login to reply.

Discussion

Router-based relays is a really interesting idea 👀

nostr:note12ujep07p6esgdesxc9esze2ldps80rdttt57c8n8yfw94lgv9hks4uqa2w

Drat! I saw an interesting reply from nostr:npub180cvv07tjdrrgpa0j7j7tmnyl2yr6yr7l8j4s3evf6u64th6gkwsyjh6w6 to this note. Where did it go? I agree with his main point that the vast majority of people don't want to run a server. I start from the same premise, without always on and always reachable servers that ordinary people don't have to think about this will never work.

But that doesn't preclude a messy 0 configuration network of clients that act solely on behalf of the user. For instance if I add my npub to a client on both my phone and my laptop they can attempt to find each other on the local network, attempt a hole-punch with a public relay, or use a relay as a TURN server. This in order to share resources for checking hashes or caching data.

If my router had a relay my devices could always fallback to using it for the above discovery. Of course not every home router will be easily addressable, but enough should be that we can use them to find out contacts even if governments go after large relays.