melvincarvalho
Melvin Carvalho

I tried it last week. The mobile client is solid but still a bit glitchy — about on par with Nostr mobile, in my experience. I’d expect a public release in June once they’ve ironed out the bugs.
As for Nostr’s censorship-resistance claims — they’re pretty weak. When real Bitcoin-level censorship resistance was on the table, there wasn’t much interest. The focus seems to be on building a Twitter clone, which makes sense given that most of the ecosystem is funded by Twitter’s founder.

2025-03-31 18:14:02

- reply

#pubky is like your npub with DNS. So from your public key, you go to the bittorrent network and run a query, then you get out a DNS record. So instead of domain_tld you have pubky to create a decentralized web. It can also store files, run relays, serve media and do all the things the web can currently do (including nostr).

2025-03-30 13:26:53

- reply

What you describe commonly known as "bike shedding"
Yes, nostr+torrents are definitely something we should so more of.
https://dtan.xyz/
But that work probably needs to be funded to get anywhere, in the current environment

2025-03-30 12:18:28

What if nostr accounts were tradable?

2025-03-29 12:59:56

- reply

No, dont put art on-chain. It was not designed for that. Node runners have every right to complain about that. Put a reference to the art, and store it with nostr, or blossom, or nosdav. It's a win-win.

2025-03-29 07:06:58

Two levels of censorship-resistance are emerging via open protocols. One is the dns-censorship-resistance model (nostr et al 2020) which is the same model as the current web w/ replication, and subject to take downs.
The second model is the bittorrent-censorship reistance model of the web (pkarr, pubky et al c. 2025) which seems to be holding its own, even against take down requests.
It's trivial to upgrade users to offer both level of censorshp-resistance for nostr devs. And it will be interesting to see which clients are the first upgrade their censorship resistance, and which are the last.
I think we're moving from a first gen nostr which was a proof of concept. To a second gen which actually is verifiably censorship resistant and is scale free. Exciting!

2025-03-29 01:04:42

They said it was impossible! But, as of today we have de-facto solved a long-standing problem in computer science. And that is of decentralized global naming, that is human readable, and fair.
1. Register a name on the bitcoin block chain, on a first-come, first-served basis
2. Name is in an OP_RETURN and human readable
3. That name has an npub associated with it.
4. The npub publishes the name to nostr relays, and also has a pubky
5. The name is globally resolvable using pubky protocol, and has bittorrent level censorship resistance
A truly censorship-resistant, human-readable, naming system that includes decentralized DNS. Quite an achievment!

2025-03-28 20:43:55

Another reason we need git without github.
#nevent1q…neh2

2025-03-28 20:36:05

[Question] Is there social tagging in Nostr?
https://github.com/nostr-protocol/nips/issues/1860

2025-03-28 18:53:55

- reply

consider that nostr should integrate on-chain bitcoin because they are both taproot, and how powerful that could be

2025-03-28 17:15:21

Hell threads are users attacking relay operators, who give up their time, to work for free. It's like going to a soup kitchen and throwing soup in the chef's face. Because they can. I understand tha aim is that the chefs, who work without pay, will improve their defenses. But a more likely outcome is that, when users throw soup in their face, they will just do other things with their free time.

2025-03-28 13:17:14

First impressions of #pubky vs #nostr
mobile experience: pubkey is better than nostr
desktop experience: nostr is better than pubky
key management: nostr nsec is easier than 12 word seeds
on boarding: pubky feels new and glitchy, but they are fixing things, nostr is more polished, glitches take longer to fix
Still only day 1, but imho a great new addition to the open web. Cant wait to see how it evolves and for the public launch.

2025-03-28 10:53:56

In retrospect, the main value of did-nostr is its ability to unify Nostr and Bitcoin primitives within a single JSON object. This not only enables a 10x–100x architectural scaling potential for both systems, but also unlocks advanced use cases like on-chain zaps, trustless atomic swaps, decentralized contracts, consistent profiles and agentic Nostr.
That said, I’m not entirely satisfied with the current DID spec—there are a couple of tweaks that would make a real difference. Still, it’s a significant improvement over what we have today, and those edge cases can be fixed later.
🔗 https://nostrcg.github.io/did-nostr/
https://media.ditto.pub/51900b2bca88ab374c2abe54bbfb9c93e02553b10303a521da81fe8736de1a6e.png

2025-03-26 07:17:20

- reply

Maybe this is the solution (timestamp set)
https://www.youtube.com/watch?v=8SOQduoLgRw&t=687s

2025-03-26 06:46:27

This is an instance of the awesome jumble project by @nprofile…59qy . Consider supporing it. Source code:
https://github.com/CodyTseng/jumble
#nevent1q…2a6f

2025-03-25 21:22:16

https://nostr.rocks/

2025-03-23 16:54:23

- reply

Use it. Use on-chain bitcoin. Use layer 2 bitcoin, such as liquid. Block should create an elements side chain at massive scale. Where the cost of a tx is close to the cost of storage (about 1msat). Integrate it. Integrate it into X. Into nostr. Into google. Use on-chain zaps. Use bitcoin as legal tender. Use all the layers, without forking bitcoin contentiously which may cause a chain split. Use the programmable layers. Use #lnbits. Use merge mining. Use single-use seals. Use RGB. Use client side validation. Use testnet. All these things are here today, and just need to be used. Once we see native currency levels, we can think about increasing the capacity of the mempool. Let's use bitcoin for commerce.

2025-03-22 10:45:46

Nostr is not a social network with payments. It's a payments network with social. At the heart of nostr is bitcoin taproot for every identity. Every npub is a taproot address with a different prefix and checksum. That means that every nostr account is a wallet and vice versa. Whether or not this proves to be a good design decision, time will tell. But it opens many unexpored doors.

2025-03-22 06:11:46

NIP-XX: Decentralized Social Verification for Sybil Resistance aka Orange Check
https://github.com/nostr-protocol/nips/issues/1851#issue-2940115890

2025-03-21 10:46:34

Zaps are great. But on-chain is 1000x bigger than LN. While zaps are needed, on-chain zaps would work far more robustly than zaps. And at far bigger scale, with more tips for creators. Also at higher layers such as our own side chains, it scales commerce almost infinitely.

2025-03-01 22:36:55

- reply

That's not Jack. Wrong T-shirt.

2025-03-01 17:25:01

I dont think The Economist helping here!
https://media.ditto.pub/087c5bf628d8aa2127675426f99a9b5f0501dd1797ec0ad188c4631208cdc390.png

2025-03-01 15:05:37

- reply

By the way I think techniques like this will become invaluable for projects like bitcoin. Especially if we can learn to subtract biases introduced by the group think. If bitcoin development could return to its roots of logical arguments to optimize the system, rather than, advocacy and politics. We could pave a cow-path towards ossification, while mitigating major risks, such as a chain-split.

2025-02-28 20:34:01

- reply

Any way I can persuade you not to put 3 things in a tag? It tends to make data less portable ...

2025-02-28 20:00:02

- reply

yours?

2025-02-27 08:57:56

Contrarly to economic narrative, money didn’t start as barter—it started as a ledger of trust.
Before coins, people tracked debts & favors, not trades.
Sumerian clay tablets, tally sticks—ledgers built civilizations.
Money isn’t a thing of value. It’s a system of memory and trust.

2025-02-26 23:21:04

- reply

The best thing the devs have done is resisted the urge to change bitcoin. That is not easy, and get's a lot of hate. But it's why we have bitcoin.

2025-02-26 20:52:45

- reply

One reason nostr was made was for a new git workflow. Which is why I wrote the original NIP-17. But we've nt fully realized that yet. What workflow do you need?

2025-02-25 18:52:56

- reply

LNBits 1.0 could help with TPS.

2025-02-25 18:26:45

- reply

Does it still have annual renewal?

2025-02-24 19:42:04

Claude 3.7 Sonnet is phenomenal. It scores 70% on SWE-Bench which is some of the hardest coding problems around. To put into context 20% was state of the art in open source a few months ago. Elon made a huge mistake not open sourcing grok 3, or even grok 2. Their massive data centers are already ancient history. FOSS is the new competitive advantage.
https://media.ditto.pub/69f263c666271a5b469cc5b6dc67735ee3a669a1dc6e2e11e90f9699b807ac62.png

2025-02-24 19:37:00

- reply

Class! If only they had the same front end too 🤣

2025-02-24 19:25:07

Nice gimmick here, where two Agents talk using a modem type language. This is how the internet worked in the early days using modems and BAUD rate. Faster today, to use a web socket, like nostr does. But it is nice to imagine agents having their own laguage,which a few of us are currently working on.
https://www.youtube.com/watch?v=EtNagNezo8w

2025-02-23 10:58:12

"Tbh I went on and just a lot of laser eyed btc maxi retards"
I was trying to explain nostr to a real cypherpunk, and this was his reaction. It's a common reaction but slightly dismissive of the whole of nostr. How do we make a better impression to technologists and cypherpunks?

2025-02-22 20:34:48

- reply

I am unsure outbox will fix everything. It seems to me that the best thing to fix is profiles, and consistent profiles accross all the clients. As that ties everything else together. And if a good way can be found to do that, the same technique can be applied to other things. One simple idea would be for the most recent profile sent by a user to be updated and linked to from your NIP-05. But there are a few different ways to solve the problem.

2025-02-16 16:13:53

- reply

Way easier than that. Treasury prints a trillion dollar coin.

2025-02-11 11:31:23

- reply

Nice work. #agenticnostr is going to grow and grow! We should form a group.

2025-02-09 14:23:42

Agentic Nostr is the future #agenticnostr

2025-02-08 20:20:35

- reply

Decent price!

2025-02-08 20:17:16

#USAID wanted to fund a twitter clone to overthrow governments

2025-02-08 20:12:26

#wikileaks exposes: The murderous history of USAID, the US Government agency behind Cuba's fake Twitter clone
https://web.archive.org/web/20210120110125/https://www.pando.com/2014/04/08/the-murderous-history-of-usaid-the-us-government-agency-behind-cubas-fake-twitter-clone/

2025-02-01 14:08:37

- reply

Nostr is over half a decade old. Corner cases are not the issue. The main architecture and setup is the issue, that it would seem to have devs and the network hitting walls. I was nostr's first daily user, and right now I will only use one client because more often than not my profile gets nerfed if I switch. Other things like performance and scalability have shown bluesky go from 100x smaller than nostr to 100x bigger than nostr in 2 years, while nostr stays the same size. The setup is quite centralized, and funds dont trickle down to the best projects (like chachi chat). It's more than just growing pains, structurally there are big problems, that no one wants to admit. I secretly hope things will get better but I'm OK with nostr as it is. I was the first daily user, and probably will be the last daily user. Enjoying the ride, but it's OK sometimes to tell the truth ...

2025-01-28 08:32:20

The game has changed.
Berkeley Researchers Replicate DeepSeek R1's Core Tech for Just $30: A Small Model RL Revolution
https://xyzlabs.substack.com/p/berkeley-researchers-replicate-deepseek

2025-01-27 16:54:25

I know it's a long shot, but does anyone know what a Trebuchet is?

2025-01-27 12:30:18

- reply

+1 Learn JavaScript

2025-01-27 03:14:11

Bitcoin strategic reserve probability plummets after ripple lobbies against it. I hope that people realize soon that ripple is 100% money printed security, and that bitcon is 0% printed. Massive difference. Additionally, everything that can be done on ripple, could be done on bitcoin+nostr.
https://image.nostr.build/96e38f327f931efc8225fa70551bb78ddbe381a7aeda3ed655589274b1d24574.jpg

2025-01-24 20:09:44

Here I Sit
https://video.nostr.build/ff5611a077ca1bf53a6a6fa72715815fb6628dba79ecde92577e40f1c37f5ace.mp4

2025-01-24 05:44:26

Chat with JFK Files
https://www.chatwithjfkfiles.com/

2025-01-18 06:48:51

Mix and match different voices to create your perfect text-to-speech voice.
https://huggingface.co/spaces/ysharma/Make_Custom_Voices_With_KokoroTTS

2025-01-18 06:07:00

- reply

Good thing to learn. Dont forget the point at infinity is Identity in the Group.

2025-01-17 17:32:10

Chachi.chat probably the best project in nostr. Sad.
https://image.nostr.build/9a6e583ffe1944e9927a951f5abbe6e87ecc7835465c09d29db90d206f09a6d4.jpg

2025-01-17 07:09:01

- reply

It makes sense, but decentralization isnt a binary thing both bluesky and nostr have different types of centralization. Nostr right now relies on subsidies. Bluesky has grown from smaller than nostr to 100x the size last year. And probably 1000x the size next year. While nostr has gone sideways in the last year. If nostr had a bigger user base e.g. 10x or even 2x, it would be less dependent on subsidies (which also causes competition) and be more sustainable. I hope nostr is open to learning from others who have achieved success.

2025-01-17 07:01:07

- reply

I get the point. But nostr should be a protocol, not a cult. It should be for everyone.

2025-01-17 06:59:00

Kokoro.js: a new JavaScript library for running Kokoro TTS (82M) locally in the browser w/ WASM
https://huggingface.co/spaces/webml-community/kokoro-web

2025-01-17 05:09:13

- reply

Not sure the rules are impossible to change. If enough people are persuaded to change a rule, then you would get a chain split, with different rules. That has already happened in the past. And it is probably the biggest threat in the short term.

2025-01-17 04:46:26

Bluesky growing fast. I hope that nostr can capture and retain some of these users, in future.
https://bcounter.nat.vg/

2025-01-15 05:52:49

Enemy
https://video.nostr.build/3b8a7bc9dfa388a9c7bb08d1e92971141309f15112f547e0373e470cde0f4540.mp4

2025-01-14 05:59:23

- reply

People are working on it

2025-01-14 05:15:35

- reply

The one thing that I really like about this, is that they are thinking about infrastructure from the start. Create capacity to grow, which can handle periodic spikes in user interest. Then grow that into a self-sustaining model.

2025-01-14 05:12:57

- reply

The large user base is the key to the at protocol.
Still a bit centralized, but looks like a step in the right direction.

2025-01-14 02:55:49

Fight
https://video.nostr.build/e58fb78a43afedbb4ab616320524012cb9ffccee296038187003e2c63915a4ec.mp4

2025-01-13 02:41:31

https://image.nostr.build/9cc229a53288c30d41197d1886aa2d3d7dcefacb32d150d8efc2924d4c3582a7.png

2025-01-12 23:19:40

SWE-Fixer achieves state-of-the-art performance among open-source solutions with open-source models achieving:
🔹 23.3% on SWE-Bench Lite
🔹 30.2% on SWE-Bench Verified
https://github.com/InternLM/SWE-Fixer

2025-01-11 14:52:54

- reply

They report T+1 so they were actually probably buying

2025-01-11 11:00:13

- reply

You may like this too:
https://www.reddit.com/r/sleep/comments/1h6vsov/i_live_streamed_my_sleep_and_applied_21_sleep/

2025-01-10 21:18:51

- reply

Good friend of mine wrote this. I've not read it, but he's quite good
"Sleepyhead: The Neuroscience of a Good Night's Rest"
https://www.amazon.com/Sleepyhead-Neuroscience-Good-Nights-Rest/dp/1541672577

2025-01-10 12:16:16

First try with minimax character references
https://video.nostr.build/448d1142f9e098a55a1795bb97b059c23c454caa026fbf107ff72efa5c1b81a6.mp4

2025-01-08 11:23:50

What could go wrong?
https://video.nostr.build/52f5f4abe002549493288f9d0c31ebec5751eda9274b27cc292b3b269651abdf.mp4

2025-01-06 21:45:43

She
https://video.nostr.build/873f8079385d1988b246070b04fee83651bd2841553fdb0095b1bb330ddfdc6b.mp4

2025-01-06 18:09:54

Magus
https://video.nostr.build/24e3aeacd501911e49ddbe5b6cecf6b3e64177d06dac3f8bf4f667f8dca3ff1b.mp4

2025-01-05 18:49:09

The Animals
https://video.nostr.build/ac99dd5493ff7bcdefa3c32afe126c75e9b9edf2e74f24533fcbbbe4558a13c4.mp4

2025-01-05 12:11:33

- reply

https://video.nostr.build/937f088bada35637f27fd210b566eb5f860fa9e2ef0900d81c5ebb45a13e35c0.mp4

2025-01-04 11:53:26

Running
https://video.nostr.build/30e73a72203dca91d931c13e471b3aa8b7921b6dffa1ccac01d63c8e4483a84b.mp4

2025-01-03 22:22:39

https://image.nostr.build/80aa1ae9832b5c792190003f76a2c4ff1048da4e6a354942c0a7e60f9f050abc.png

2025-01-03 22:20:49

2025 will be agentic
https://image.nostr.build/5fcdaeb8b53c10f6020fc509060c297e164e50400ec02426826dfeb817f5bb82.jpg

2025-01-03 22:18:57

This would be great on nostr
https://oasis.camel-ai.org/
https://image.nostr.build/7eddbd8e7963bb6a50edbaf10be2d9c1094870d11d6d4bef98ddc4ebec360c3b.png

2025-01-03 22:16:41

OASIS: Open Agents Social Interaction Simulations on One Million Agents
There has been a growing interest in enhancing rule-based agent-based models (ABMs) for social media platforms (i.e., X, Reddit) with more realistic large language model(LLM) agents, thereby allowing for a more nuanced study of complex systems. As a result, several LLM-based ABMs have been proposed in the past year. While they hold promise, each simulator is specifically designed to study a particular scenario, making it time-consuming and resource-intensive to explore other phenomena using the same ABM. Additionally, these models simulate only a limited number of agents, whereas real-world social media platforms involve millions of users. To this end, we propose OASIS, a generalizable and scalable social media simulator. OASIS is designed based on real-world social media platforms, incorporating dynamically updated environments(i.e., dynamic social networks and post information), diverse action spaces(i.e., following, commenting), and recommendation systems(i.e., interest-based and hot-score-based). Additionally, OASIS supports large-scale user simulations, capable of modeling up to one million users. With these features, OASIS can be easily extended to different social media platforms to study large-scale group phenomena and behaviors. We replicate various social phenomena, including information spreading, group polarization, and herd effects across X and Reddit platforms. Moreover, we provide observations of social phenomena at different agent group scales. we observe that the larger agent group scale leads to more enhanced group dynamics and more diverse and helpful agents' opinions.
https://oasis.camel-ai.org/

2025-01-03 15:10:47

Shiger?
https://video.nostr.build/505ff931863bd3021df177ae1ca13af829ee9aa09b3492098d5988c91e3f25ee.mp4

2025-01-03 14:45:12

https://video.nostr.build/1ef69164903a6b4778cc0d358179f8d708ca517a1c7ae43472ee1c42f2e9bc76.mp4

2025-01-03 11:27:19

- reply

I dont know. I havent seen it. We implemented OIDC in solid, and you can add your nostr account to solid now, too.

2025-01-03 11:23:46

Green = Open. Red = Closed.
https://image.nostr.build/114c63fc15bbfbd1ec598b3e0dcedec6ae1d324bdbb7a8c28e798e2881cff281.png

2025-01-02 18:29:33

luxury
https://video.nostr.build/1d031d3b0fc532abce5b37783e9819ad60598e5cf2c9cfbccc4198e0fff3dfe7.mp4

2025-01-02 18:16:52

Open Source Wins
https://image.nostr.build/248d1576c27721d561ccc90219988bd26732861b0967017987fcf747a71269ca.png

2025-01-01 13:31:36

Outgunned
https://video.nostr.build/e1e5d2f662a04f1fc55096736709ccd267baf60c6ea6f7e210ac370efc5b38ff.mp4

2024-12-31 11:56:09

- reply

Happy new moon, frens!
https://image.nostr.build/1e16f4931607bb4151e6074ccf289acf1b8f5d0b64929af3d8de2f68235f2e19.png

2024-12-30 15:01:07

- reply

I dont think so, because it's 24 words. So even if I were to publish them you could not get the privkey because they are sufficiently scrambled (24!) combinations.

2024-12-30 14:24:56

meow
https://video.nostr.build/764cf084ca5550c896d55fff8c63bac0e8b4c6a1fe3d2b465f4cf4f0c4fd926e.mp4

2024-12-30 14:11:47

- reply

I put a random seed phrase into AI image gen, and it spat out this
https://image.nostr.build/639e05316c55fc74bd6b4e4b302cc92dcf696e54f407e8400e713f77e6a3166e.png

2024-12-30 11:07:32

#prompt An artistic close-up shot of a perfume bottle, placed delicately on a snow crystal pedestal surrounded by twinkling Christmas lights. The sleek glass reflects the warm golden and red hues of festive candles flickering nearby. Snow-dusted roses and delicate pinecones frame the scene, adding a touch of winter sophistication. The Nostr logo is subtly illuminated by soft, natural light filtering through the scene. DSLR, moody and atmospheric lighting
https://image.nostr.build/87fbdf685784e148e6cafbf94a6cdf71821be0db322728c94cd3cf418822e468.png

2024-12-29 12:50:02

#AI Timeline
https://nhlocal.github.io/AiTimeline/#2024

2024-12-28 10:48:44

https://image.nostr.build/62c42c3b33fa8dd39a09ff583d45e7c678038c3ee09bd2033b01140c16e7ae28.png

2024-12-27 16:49:21

- reply

Great interview, Peter. The man who beat Roy Jenkins, still as eloquent as ever. The two brains at that table could solve most problems.

2024-12-27 05:35:28

Girl and Bear
https://video.nostr.build/df80e3c0b19a181a05bb195ea36816629cb96fd278932823f64a3fea980b4559.mp4

2024-12-26 13:38:02

- reply

Try phi 3.5 and 4 is out soon

2024-12-26 13:24:12

Beyond the Sum: Unlocking AI Agents Potential Through Market Forces
https://github.com/Fewsats/beyond_the_sum

2024-12-26 12:58:03

Crazy advances in open agentic AI. Deepseek v3 just raised the bar.
https://image.nostr.build/89d1f232e8f91e2219eb1d4d3d78a143e7f4f6f27955a0c8bb6aa4fa257ac71e.png

2024-12-26 10:02:47

- reply

OSHash is even better for big files—super fast since it doesn’t read the whole thing. Leading to a popular UX that works For most media uses, that speed is a win! Unreliable servers? They won’t last anyway. 😊
https://image.nostr.build/0fa891044b8531d3fa94fe4a407db0d71375f689fbaade20e88ef8aaafdd2f16.png

2024-12-25 23:54:54

- reply

Sounds great! I’ve always been a fan of Oshash for media—it integrates so well with the OpenSubtitles ecosystem, which has most of the world’s content indexed. I remember it was quite a challenge and even a bit controversial when they switched to a better hashing algorithm. I think the resistance came from the sheer scale of the ecosystem and the compatibility issues involved, but it’s impressive how they made it work in the end!
https://image.nostr.build/415b57b9e7d0fe4ea73b0e403f46d5b3fb8f191c061f105e46a1fa26a1a50e09.png