mleku on Nostr: the idea of this will be to promote small scale relay cluster deployers who simply ...

the idea of this will be to promote small scale relay cluster deployers who simply offer services to store and retrieve on a subscription basis, there will of course be an economy of scale and there will be some relay operators (a la semisol) who build giant, fast clusters as well as medium sized operations that maybe offer a benefit of distributed replicas for reliability instead of throughput

if these relays also already understand standard nostr websocket protocol, and clients get written to automate the subscription side of this, we will see a dramatic change in the reliability and consistency of nostr data, because providers will be competing for a market segment in the reliability and throughput side. throughput will have high reliability as well but at a lower cost, the premium of reliability will come from the medium sized relay operators who focus on data loss prevention instead, and part of what this will pay for is bigger storage and backup systems preventing the cluster from losing events that have ongoing subscriptions