This image is slightly outdated, but it’s still a good representation of how our payment verification works, including built-in splits and referrals.

Think of it like signing a contract with our tiny backend before paying for an event. It attests what you’re purchasing and sends back a signed contract. After payment, you provide proof that the token was spent. We cross-check that against the contract, and if everything matches, you get the decryption key.

We’re also experimenting with open web-of-trust systems, particularly those being built by the Nostr community. We’re open to letting authors decide which mints they trust, and our system can adapt around that.

There’s also a key recovery mechanism if a key gets leaked. It’s still being developed, but it already works. Key derivation itself is deterministic based on the event, its author, and a specialized tag, while our backend only holds half of the secret.

Eventually, we want to expose all of this through an API so paid content can be integrated across applications. And what if that API could be self-hosted on a $2.50 ESP32?

We know we’ll make mistakes and there will be trade-offs along the way, but we’ll do our best. Feel free to ask me anything about it.

This is where the Fanfares architecture gets really interesting.

We use Nostr, relays, and Cashu for most of the heavy lifting. Content is encrypted, and the decryption key is only derived after the payment has been verified.

With Cashu, that verification can be extremely lightweight. Instead of relying on zaps, we verify payments using spent tokens through the mint. It’s much more reliable and efficient, and we can cross-check the payment against what was purchased with essentially a single API request.

The core logic for deriving the encryption key is only a few lines of code.

That makes the ESP32 a surprisingly good candidate for running it.

It costs around $2.50, has Wi-Fi built in, and is powerful enough to run this tiny server. We could even build a simple tool that flashes and configures the ESP32 for you in one-click.

Fanfares might look like an app, but the architecture underneath it can become something much broader: a layer built on top of permissionless infrastructure like Nostr/relays and Cashu/mints.

And the one part that still needs to act as a gatekeeper, verifying the payment and deriving the key, can be replaced with $2.50 ESP32 sitting on your desk.

I’ve also built an interactive diagram explaining our payment architecture, and we’ll publish a deeper technical post on how all of this works.

I’ve been part of the core Fanfares team for more than a year now.

There’s so much cool stuff we’ve built, and so much I’ve wanted to talk about. I went through a lot personally this year, so I wasn’t able to share nearly as much as I wanted to. But behind the scenes, we kept building together.

Fanfares has come a long way.

We now have on-device media processing and encryption, split payments powered by Cashu, Blossom support, and an architecture built on top of Nostr.

And one of my favorite parts: the backend is lightweight enough to be hosted on a $2.50 ESP32 board.

A lot went into getting here. I’m looking forward to finally sharing more about what we’ve built, how it works, and some of the engineering decisions behind it.

When I was 16, I got a Royal Enfield Desert Storm 500cc. I absolutely loved it and rode it everywhere. Then I got into an accident. I was covered in scratches, but my helmet and riding gloves probably saved me from much worse injuries. After that, I never got back on a bike. Now I only drive cars.

Just when I thought it couldn’t get any worse. How is that even real? It sounds like something out of a dystopian novel.

I hope they make an example of those responsible and that this becomes a turning point for the entire nation.

RIP Henry Nowak.

Yeah, I think hatred toward obese people is definitely playing a part there. It started as a pushback against the glamorization of obesity through magazine covers and all that, which was completely justified. But hating obese people as a whole is equally wrong, in my opinion.

I don’t know much about the effects in that particular area. Are there any peer-reviewed research papers I could read about it?

I completely agree. As far as I know, the effects are often temporary if someone stops taking the medication and returns to their previous lifestyle. In many cases, people regain a significant amount of the weight they lost.

That’s also my understanding of it. I’ve heard some people say it reduces cravings, long-term studies are still needed, but I really hope it works without some major side-effects

That’s actually a fair point. As far as I know, there’s no reliable way to predict long-term side effects without studying it for a long time.