Replying to Avatar Melvin Carvalho

So, just as a quick reminder—Nostr has actually had quite a lot of funding. Even with around $1M each month, it hasn’t really seen sustained growth, so we have to consider what might happen without that level of support. If the funding were to dry up, it’s quite likely we’d see a real drop-off in users.

It feels like the relay network could have used more attention and funding early on. With that neglected, we’re seeing operators start to move on. And right now, there’s not much of a sustainable business model to rely on—practically no revenue streams. This makes it hard to envision a future without ongoing appeals for funding, especially to Jack, whose generosity has been key.

I think part of the challenge is that scaling just hasn’t been factored in. Simplicity can be great, but it doesn’t inherently scale. Now there’s talk of adding complexity through negentropy, which may end up fragmenting the network rather than improving things. It’s also easy to assume things scale by default, but most systems don’t—scaling takes deliberate work and careful planning. Nostr has struggled in this regard, and the NIPs have grown more centralized, which has also limited the flexibility needed for use cases and architecture. Only #ditto has really made an effort there.

The social graph portability in Nostr is actually very promising. It’s an elegant solution—using hashes to identify users is a technique we know well and can apply in many contexts. Other projects like #pubky are experimenting with similar approaches, and it’s encouraging to see the potential there.

However, Nostr’s main challenge remains quite fundamental. For it to become financially viable, it would need to grow by a factor of 100, yet it’s currently at its limit in terms of capacity. And since investment in relays has been limited, the network is unfortunately in decline.

That said, I can see Nostr’s relevance to the Bitcoin ecosystem, and supporting it can make sense from that perspective, even if it’s a philanthropic investment. But acknowledging the current challenges is essential if we’re ever going to fix them. Right now, the developer experience could be much better, and the relay network needs more focus. Some clients are hindering others, and the project’s leadership has room for improvement, especially compared to other open-source initiatives. And there are substantial opportunities we’re missing, particularly around taproot and financial markets, again due to centralization and group-think.

In short, Nostr offers a rich space for innovation that’s still largely untapped. Proof of concepts are valuable, but they don’t necessarily scale. Generous donations have provided stability, yet a more targeted strategy could make a significant difference. At this point, though, it’s still a cost center with a limited runway, reliant on the goodwill of its supporters. #ditto, on the other hand, has potential because it’s built on a different architecture. And by focusing on Taproot-based solutions and market-driven models, we’d allow the free market to decide what works—letting successful ideas flourish while phasing out those that don’t.

#Reticulum 's low bandwidth protocol LXMF is easily scaled. The networking stack is quite complex though

https://github.com/markqvist/Reticulum/discussions/59

Reply to this note

Please Login to reply.

Discussion

A notable quote from the thread:

"Exactly, and that makes the cost of running one completely neglible. LXMF is so efficient that it can handle millions of users on a Raspberry Pi and a 256 GB SD card. When something is that cheap, how are you realistically going to monetise it? There's no point. Leaves you to focus on the things that are hard to provide for monetisation.

...a very general theme in everything I have implemented around Reticulum is to base reasoning on physical resource limits and prioritising how they are utilised.

Within this conceptual framework, it's possible to create functional (truly) decentralised application where cost is so distributed that it is neglible for every user. Not using memory resources that are available is actually more costly than giving them away for free, on a societal level.

When you have Reticulum fully deployed, there is no need for costly "servers" that centrally handle compute and memory demands. The concept is unneccesary when everyone can reach everyone. At our current level of technology, we have all the compute and memory we can possibly need, it's only the transport layer that is lacking immensely behind.

There's no point in trying to monetise the water you swim in."