I've only been around for 5 or 6 weeks but this is something that I've always foreseen as happening. Clients are the super tool here. Relays are just dumb backends that the average user doesn't care about. They are extremely important, but the average person on the street does not care about that. Because of this, all major clients will run relays and they'll bootstrap their users into use them. We'll go down the centralization path a little bit, but users will have the choice to use other big and popular relays. I do hope that we don't go full down this path. It's still too early to tell. So, I guess the question remains: how do we make relays more important to users? How do we make relay selection important? Should it be part of every client's onboarding process? Do we advocate for that? If a priv/pub key signs in and doesn't have any relays attached to their profile, then display the relay onboarding process? This still leads to clients choosing a default relay list and most users will never go back and update those again because it's not something we've had to do in the last 20 years of social networking. It will take a long time to change that mentality.

Reply to this note

Please Login to reply.

Discussion

If I run my own relay and users engage with my account, “centralized relays” shouldn’t matter? Isn’t this similar to Bitcoin where every person doesn’t need to run their own node, but they MUST have the ability/option TO run their own node? This is a serious question because I’m not sure I’ve completely wrapped my head around how relays interact with each other on Nostr…

That's a good way of looking at it.

Bitcoin gives you a view on a shared global ledger, and you can run your own node to verify that ledger.

Nostr gives you ability to converse freely. Or to help others converse freely. Anyone can run a relay, but you dont have to run a relay.

The difference is, bitcoin is one single ledger state. Nostr is sharded into many thousands or millions of conversations, that can be independent of each other.

Ultimately things that work don't matter until they stop working in some way. Relays censoring notes for example?

It doesn’t have to be a choice between all users managing their relays vs. full blown relay centralization.

As others have said in different ways, we can also have a system where for the default user, “relay autopilot” intelligently figures out which relays to connect to based on the most common ones that their follows use, while also evenly distributing the load between the relays.

There can then be an “advanced user mode” for those who want to manage their relays manually.

On the relay side, multiplexing/mirroring can make it so that relays can sync with other relays.

With this system, you get a scenario where you average user doesn’t need to worry about relays, but it also prevents the need/trend of relays centralizing and keeps the load evenly distributed.

As for funding of relays, I imagine long term, they’ll be a mix of free volunteer run relay, paid relay, and free ad-supported relays.

#[2]

The relays will be chosen according to the user's interests. You'll want to be on the relay of your friends or anyone looking for information.

There will also be relays of specific themes. I quickly see a relay just about football emerging, news will probably be another topic that will soon emerge. and of course we can't forget the sex and dating relay (Tinder).

If one were to make clear some kind of cause and effect to the user relating to what an active selection of relays is responsible for then its kind of easy. I've been messing around on nostr for only a week or so and its not at all clear. Otherwise its pretty awesome :-)