Having an audio conversation with gpt-4o is surreal. Incredibly smooth.

The NIP-05 page doesn't say I should do anything else: https://github.com/nostr-protocol/nips/blob/master/05.md

Anyone know what I need to do to get a purple checkmark on Nostrudel instead of a yellow caution icon?

Apparently adding my relays to my nip-05 doesn't work. Is it a domain whitelist thing?

Whoa, you can browse all the notes from a single relay in Nostrudel. Super cool!

Adding that '?' at the end of a Kagi search is magical

So happy overleaf now has drag and drop file uploads, small feature but so convenient

Time for stack overflow on Nostr?

(Please don’t call it stackoverflowstr)

nostr:note1zcfp22esrtcvvxq5dfsl8mlamyw6dj4v3q44tapx3gn5cvl2wjpq5rnkax

I haven’t found anything great yet. Depending on how big your project is, context window limits are still a problem (I.e. you can’t just give the whole code base along with your prompt for every query).

Cursor is the best thing I’ve tried, it can be helpful for answering questions about big codebases, although it doesn’t always work for things like what you asked for. I assume it’s using a form of RAG and RAG can be nuanced (chunk size matters a lot and retrieval is far from perfect).

I’m still waiting to find the tool you’re describing…

Thank you I’ll check that out

Would relays be kicking me off or just decide to stop responding if I'm subscribed for too long? Is there any way I could detect that if it happens?

I should say that what I'm trying to do is run a python program that forever listens for DVM related events. I've been using a version of nostr_client() from here: https://github.com/believethehype/nostrdvm/blob/main/tests/test_dvm_client.py

It does look like this has been updated since I last started using it, maybe I should adapt this most recent code.

I did pass part of the source code through an LLM and it guessed that the timeout is being used for how long to listen for events, rather than the window of time in which events were published.

my rust is rusty 😇

nostr:npub1l2vyh47mk2p0qlsku7hg0vn29faehy9hy34ygaclpn66ukqp3afqutajft nostr:npub1nxa4tywfz9nqp7z9zp7nr7d4nchhclsf58lcqt5y782rmf2hefjquaa6q8 nostr:npub1drvpzev3syqt0kjrls50050uzf25gehpz9vgdw08hvex7e0vgfeq0eseet if I have questions about nostr_sdk bindings for python, is there a chat / telegram / etc where people hang out?

I'm trying to figure out what client.get_events_of() does, specifically what the timeout does.

I'd also be willing to contribute to documentation for the python bindings.

Sitting in a coffee shop working on a #nostr project. Life couldn't be better.

nostr:npub1n0sturny6w9zn2wwexju3m6asu7zh7jnv2jt2kx6tlmfhs7thq0qnflahe loving the notifications when specific people post!

Real willpower is to enjoy that which is uncomfortable and unpleasant.

I wonder if the infinity of symbolic state spaces like PDDL is bigger than the infinity of learned feature representations, like what you get out of an embedding model.

Of course I’m talking about PDDL without floats because that’s obviously equal to or bigger.

Whether this is true or not has an impact on the fundamental generalizability of symbolic vs statistical models

Thinking of starting an LLM disappointment support group. Have you been affected by an LLM hallucination recently? Come join us in our journey to recover from the negative effects of LLMs.

If these renaissance artists were alive today, Nostr’s the only network where they could post

Started using it because people on Nostr kept talking about, can’t go back. It’s so much better.

Thanks, I didn't realize I needed to add relays - I did get it to work!