I think the model of the client that tries to follow someone automatically is not incompatible at all with the model of the client that allows you to switch relays view and puts relays at the forefront of the experience. They are very compatible and the synergy between the two models can create interesting experiences. I wouldn't be unhealthy for some clients to specialize in one model of browsing while others specialize on the other, but a mix would be desirable too.

Also the algorithmic follow-by-heuristics method may work on Twitter-like experiences, but for most of other use cases that can be done with Nostr the browse-relays experience seems to fit much better.

Anyway, the only thing I think is terribly unhealthy is the current ubiquitous model of a "settings" page with a list of boring and meaningless relay URLs and people selecting these once (if at all) and never thinking about them again. This is horrible, centralizating, unscalable, a terrible anti-pattern. I take the blame for writing the first Nostr client with that experience.

Reply to this note

Please Login to reply.

Discussion

(In my defense I did have in mind that these relays would become more personalized and more an integral part of the experience as time went on, I just didn't know how to do it at the time and it wouldn't make sense since there were no relays or users. Again in my defense I also think that pattern was kinda "obvious" and if I hadn't done it others would do it anyway since it "fits" the basic patterns that most non-nostr apps use today.)