I love that Nostr can be as simple as using 2-3 relays. It makes getting started easy, and honestly it has made my life easier. I have been a large relay girl from the get go.

But the more I think about building on Nostr, I also see the challenges of big relays. Without careful design, both users and conversations tend to centralize around the largest and most convenient relays. That feels like the opposite of what Nostr was meant to be. How much will client defaults and user inertia end up influencing the network?

The downside is what keeps me thinking. If I build on this, what are my users going to see? Porn? Political views they strongly disagree with? Is that the image I want associated with my brand? Do I then start censoring things? If I do, where do I draw the line? Am I suddenly deciding what people should and shouldn't see, and based on what, my own judgement? That doesn't feel right either.

I believe users should be able to choose for themselves. But right now the only real option is to mute or block after they've already seen it, not actually design their own ecosystem around their interests. Are users going to end up muting 500 content providers they don't like, or just leave Nostr altogether after the first 10? That pushes the onus back onto the client.

But without some form of filtering on the client side, do we end up back at an all-or-nothing centralized platform? I myself have been annoyed from time to time by the infighting and gender degradation. Do I want other people seeing that all the time?

And somehow I end up going around in circles on this.

But then there is also an incredibly attractive upside. Large relays make it viable to build serverless apps without needing to run your own infrastructure, which is a huge deal. These guys have been carrying a lot of the load for a long time now, for which I am grateful for. So there is a real trade off.

Then there is the UX hurdle. The cryptographic key based identity system is incredibly attractive (secp256k1 keys underneath, same curve as Bitcoin), but login is still a major barrier for non tech users. It is a problem we all know needs solving.

On the technical side, I am really excited about the possibilities of FROST for key security and projects like Pomegranate. It is great to see fiatjaf and other devs exploring tools in this space, even if they are still early. My only hope is that momentum holds and these early projects don’t lose steam before they mature.

Discovery is still a big open question, but that is also what makes this space exciting. The potential for building and solving these problems is what keeps a whole bunch of problem solvers coming back.

Honestly I was not thinking about building on Nostr until recently. But the more conversations I have been having with @daniele over the past few weeks, the more tempted I am to migrate what I have been building ("the other stuff" related). Some people see problems and complain. I have always believed that identifying problems is the first step toward finding solutions. To me, the bigger problem is not knowing what the problem is. And I may not know enough.

But the more I think about it, the more I see vast opportunities here.