Gonna do - https://bountsr.org/relay-display/ - https://bountsr.org/follow-hints/ - https://bountsr.org/nevent-nprofile/ for iris client ( #[0] ) They all seem very linked together, ez cash. Possibly also https://bountsr.org/gossip-model/ -- it seems similar to the prev 3. https://bountsr.org/relay-publish/ will be easy. Might even do that one first.

Reactions (likes) should get tossed. Instead, a frontend+backend should be made where users get access 1 sat per day to spend on zaps, stacking up to 10. A frontend with a custodial lightning node married to it is just waiting to happen, no? Or are lightning tx fees too hefty for that? How much does a 1sat zap go for?

I am quite pessimistic about the future of nostr.

Apparently langchain+llama.cpp can make a locally trained chatgpt interface for...things. The langchain example shows how they train it on the declaration of independence and then ask it questions about it, and it works. That example is on openai, but llama.cpp produces similar results. There definitely is a need for that over NIPs. Among other things, of course.

LangChain overview - Docs by LangChain

LangChain is an open source framework with a prebuilt agent architecture and integrations for any model or tool—so you can build agents that adap...

GitHub

GitHub - ggml-org/llama.cpp: LLM inference in C/C++

LLM inference in C/C++. Contribute to ggml-org/llama.cpp development by creating an account on GitHub.

It would be good manners for client developers to include a link to the source code somewhere prevalent? So far none I've seen have done so? why

I think the `like` button on nostr is 100% doomed. It'll just get spam botted to oblivion, and it's way more effort to combat that. Zaps make more sense.

Now liking the iris.to client more than coracle. It's faster?

Here we go:

GitHub

Nested replies by npub1zenn0 · Pull Request #16 · fiatjaf/nocomment

Diff is quite huge because of moving code around. The second commit has actual changes. (the clipped border was like that from the start, prob loc...

I made this:

GitHub

GitHub - npub1zenn0/go-git-nostr: Send and receive git patches over nostr

Send and receive git patches over nostr. Contribute to npub1zenn0/go-git-nostr development by creating an account on GitHub.

I think it ticks all the boxes.

After some time developing on nostr, "simple protocol" makes development a real pain. Nostr isn't picking up steam anytime soon. Less devs => worse UX => slow death. There is a need for good frameworks/programming libraries, but from the experience of Ethereum web3, I can confidently state that no breakthrough libraries are on the horizon. I'm thinking that NIPs should be gatekept by introducing libraries to deal with said NIP first.

Will it be enough to create a web app that scans through relays on the client-side? It will not work for TOR relays then.

testing

Holy moly the UX of the first 2 nostr clients I try is atrocious. Make a post and nothing happens. I cannot tell if my post went through or not.

Whoever is the developer of primal.net, I have a feature request: please add `draggable=false` to replies, so I can copy paste text. Also I would flip post/cancel positions but that's just me.