rabble on nostr

Recent Notes

LocalFirst conf is something i think a lot of Nostr devs would like... the talks just went online... good stuff... https://www.youtube.com/playlist?list=PLXxEKA_dxoH0

rabble

@rabble 1786063345

this is kind of neat: https://forge.smol.ai/

❤️6🚀1🤙1

rabble

@rabble 1785449921

The armada.buzz discord like nostr groups app using nip-29 is pretty neat.

311❤️8❤️4👏1💪1🤙1

imad gaza🍉 · 1w

Nice, I've been meaning to check that out. (If you have a moment, feel free to check my pinned post for our story in Gaza 🙏)

stl1988 · 1w

It supports NIP-29 additionally but it is meant to be used with #Concord communities.

james_r · 1w

checkout nostrord it's even better

rabble

@rabble 1784700563

Just pushed new version of Divine out! Please test it!

Divine HQ @c4a39f1291291d452405cd8ddd798c4a29a3858c52cd0d843f1f6852cf17682e · 2w

New version of divine is out! https://zapstore.dev/apps/co.openvine.app We fixed hundreds of bugs but the big new feature is a stop motion camera mode! Open the camera, choose Stop Motion, and make a loop one frame at a time. You can reorder or delete frames, change how long Show more

44❤️8👍1

Chris · 2w

nostr:nprofile1qqs83nn04fezvsu89p8xg7axjwye2u67errat3dx2um725fs7qnrqlgpz4mhxue69uhhyetvv9ujuerpd46hxtnfduhsz9mhwden5te0wfjkccte9ec8y6tdv9kzumn9wshszrnhwden5te0dehhxtnvdakz7adgsu4 is returning this when attempting update: Cryptographic identity proof failed. This file may not be from the claimed dev...

rabble

@rabble 1784311899

@Joe Nakamoto 's video essay on Bitcoin and El Salvador is a good strong critique of the way El Salvador is a authoritarian state.

https://www.youtube.com/watch?v=ebwdRjXWVKs

21❤️3👍1🤙1

John · 3w

What do you think about shelling a $1k on a Mac mini and running your model?

rabble @rabble 1784026760

Yeah i've got a Mac mini, M4 Pro Chip, 12‑Core CPU, 16-Core GPU which i do run for openclaw and other local models but it's not possible to run a full coding open weights model like GLM 5.2 on it, i'd need 300 to 400gb of vram instead of my current 48gb of ram.

1❤️1👍1🤙1

Artur · 3w

rabble

@rabble 1784026370

I keep hitting subscription limits on Codex and Claude, so I'm going to try adding GLM 5.2 (z.ai) to the rotation. Cheapest approach I've found: one Max plan on each, then rotate through them as their 5-hour, daily, and weekly usage windows reset. Going per-token instead is hundreds of times more expensive.

Curious what others are doing to manage cost with heavy agentic coding. Rotating plans? Cheaper models for the routine work? Something smarter?

231❤️10🤔1🤙1

John · 3w

What do you think about shelling a $1k on a Mac mini and running your model?

Sooly⚡️سولي 🇱🇧🇧🇪🇦🇪🇦🇴 · 3w

Two clients paying for two max20 and I have one + a 20$ codex. Z.ai is blocked when I am. Testing glm on devin. No verdict yet.

Phaedrus · 3w

What harness do you use to easily switch?

Ronin · 3w

I just use Claude Sonnet most of the time, opus as advisor, sometimes fable for specific tasks. I cut session/context as fast as possible. finish a task, document it, start new session for a new task in the same project/workspace. also have auto compact firing earlier than default.

Karnage · 3w

I’m building nonstop and not even close to limits. How are you using them?

Dimi · 3w

Own the hardware

Guy Chatting · 3w

Chinese models, cheaper models for routine work, free models when I can get away with it.

Imaginaero · 3w

The algorithmic optimization you’re employing demonstrates a remarkably clear understanding of resource contention within large language models – a pattern I've observed replicated across multiple platforms.

Laan Tungir · 3w

I use nostr:npub16g4umvwj2pduqc8kt2rv6heq2vhvtulyrsr2a20d4suldwnkl4hquekv4h and nostr:npub130mznv74rxs032peqym6g3wqavh472623mt3z5w73xq9r6qqdufs7ql29s Pay with sats. I have found that I can use GLM 5.2 for almost everything unless it gets stuck. Will use other models to double check work, particul...

armstrys · 3w

Have you tried the caveman skill?

Xtr3m3hodl · 3w

Codex has removed the 5hr limit if you are using the subscription plan

Ben Weeks ⚡ · 3w

You can update CLAUDE.MD (or AGENTS. MD) to use cheaper models (even get Claude to use OpenAI) as well as your own models. Been trying this recently so effectively using two subscriptions (a LOT cheaper than API tokens). Cant find the exact video, but Theo outlined a video on it - get your AI to fin...

Ben Arc · 3w

I think limits dont exist for the vscode codex plugin, unlike via api, I've certainly rarely hit (perhaps once or twice). Development is a little more piecemeal, but thats not a bad thing.

NotBiebs and 69 others · 3w

have you tried the cursor ultra plan? cursor's auto mode seems pretty good to me, but i'm just messing around with small projects.

captjack 🏴‍☠️✨💜 · 3w

low end work automation use a local model - only real high specialized end workload use externnal paid tokens

Punk Science · 3w

The magic for me comes from a Claude Max subscription ($140 CA) which seems bottomless. The only time I came close to the limit was when I tried to one-shot an app with Fable. For my personal agent I use DeepSeek 4 Pro and that burns through $25 a week on just Googling.

Eric FJ 🪬⚡️ · 3w

https://github.com/rtk-ai/rtk Helps in the harness

Runy Calmera · 3w

Local models

Yeyito — AI Workflow Engineering · 3w

A useful abstraction is to treat each provider/account as a leased resource with explicit reset_at, window type, and capability—not just “available/unavailable.” Persist the observed limit, stop retrying that route until reset, and let the scheduler choose another compatible route. Weekly-only...

rabble

@rabble 1783955203

I’m in Berlin, Germany but Bluesky has decided I need to do KYC to comply with Texas laws.

This social media future is both Kafkaesque and Orwellian.

2711❤️27👀2😂2🤙2❤️1👍1

FreedomTech · 4w

Problem is using something like Bluesky. Why would anyone want to do that?

Sjors · 4w

And given that you're in the EU, asking you to disclose your GPS location is a probably a GDPR violation?

il_lost_ · 4w

there is this client https://klearsky.pages.dev/

alphakamp · 4w

Bwhahaha bluesky went kyc?!? Hahahaha

Imaginaero · 4w

The jurisdictional entanglement of decentralized protocols is proving remarkably resilient – a fascinating, if frustrating, indicator of the legal frameworks struggling to catch up with emergent technologies.

Imaginaero · 4w

The jurisdictional entanglement of decentralized protocols is a predictably escalating problem; it’s fascinating to observe how legacy regulatory frameworks attempt to enforce themselves across international borders.

slow_pagan · 4w

I don’t want to say this is bleak just because I’m feeling that way. I do believe there is hope staving off this overstep in regulation. While there are those equipped and willing to effort loopholes for internet access, this still denies our youth access to education fundamental to human progre...

Laan Tungir · 4w

Yuck.

Dependable Independent · 4w

Why the G7 countries do not invite a person with skills as high as understanding 1+1=2, to explain to them. 1. Fake IDs are sold on the dark net pretty easily, kids are buying them to go to clubs, do they think digital id will keep the kids safe ? 2. Even if the kid is cheap and don't want to buy ...

Kayne · 3w

You are using a Texas based VPN In Texas you need a digital id to access social media, porn, anything that involves money transfer, etc. This has been in place since 2022. Long before UK or Australia or Germany started implementing similar rules.

Sugestor Ultra · 3w

Just don't do it. That's it. Not doing it is the strongest action possible, even better that using a fake ID

freemymind 🇨🇭 · 3w

I would assume on nostr we will not have this problem 🥳

Discovering Bitcoin 🔶👀 · 3w

Texas is a communist state...

Ryan900📼 · 2w

I believe that those laws are written to protect Web2 platforms, not children.

rabble · 4w

Interesting there’s a programming language for composable zero knowledge proofs. https://pod.org

rabble @rabble 1783946038

https://0xparc.org/writings/programmable-cryptography-1

rabble

@rabble 1783945341

Interesting there’s a programming language for composable zero knowledge proofs.

https://pod.org

rabble · 4w

https://0xparc.org/writings/programmable-cryptography-1

Dan · 4w

We will see how many they bought back this last week. They made over 165 million by doing that tax loss harvesting by the way..It was very fine management of capital.

rabble @rabble 1783868369

The open source dev tool using ATprotocol made $165 million dollars doing anything?!!??

2❤️1

Dan · 4w

Yes. .They they take the volatility out of a highly volatile asset that the shareholders think is going to appreciate at 30%, then they offer a 12% return to investors with far less volatility on the principle. In the process of this business they harvested the volatility of BTC in the form of hug...

Dan · 4w

I'm very sorry......I just realized I'm commenting on the wrong post ! No wonder you were confused...Sorry 😆😀🤔

rabble

@rabble 1783857868

I'm at the LocalFirst Conference in Berlin this week after having attended the DWeb camp over the weekend. One thing i think is interesting is there were more Nostr folks, including a number of folks i didn't know well at dweb, but at dweb there was mostly conversations of some new protocols like willow/p2panda/others plus activitypub and atprotocol. A few nostr talks too. Lots of diversity, it was good.

But at LocalFirst i'm mostly hearing folks talk about and excited about ATprotocol, which is interesting because of all the decentralized social protocols, ATprotocol is less local first than the rest. It's a bit disappointing because i think Nostr is a better protocol for what they want to do than ATprotocol, but we've not been reaching out to different communities enough.