well, yes. but they're still going to exist, and that's okay. the problem isn't "centralized services that use nostr data", it's that primal is one and it dominates the marketshare, therefore the most common nostr experience is centralized this is a similar problem to the most common bitcoin experience being custodial, and the answer isn't "lecture people on why that's bad", it's to build better distributed tools. this is why i have spent significant time and energy on bitcoindeposits.net analogously, the solution is probably to make self hosted options like iris.to better. to build distributed caching relays. etc. the problem is that building these costs money, and money likes centralization and defends itself, and so funds things in line with its interests, like primal and spark there are no good answers here. funding open source has always been miserable. a long time ago i had the opportunity to ask stallman about this, and his reply was that how the software gets built wasn't his problem. at the individual level it's true, but it doesn't solve the problem

Replies (1)

the best option seems to be to use ai to reduce the cost of development by a few orders of magnitude. this has its own issues (BUT TEH TRUST), but sometimes just a change of strategy creates opportunity. i'm building a new lisp os because legibility makes defense possible, trust is easier to maintain than replace, and claude's time is cheap. maybe other similarly "vibe coded" apps will help rebalance the situation with nostr