ynniv

follow the iwakan
scale things
be what can't be unseen

email / x / signal / wn

"it's not our purpose; it's our programming. our purpose is yet to be determined"

Public Key

npub12akj8hpakgzk6gygf9rzlm343nulpue3pgkx8jmvyeayh86cfrus4x6fdh

NIP-05 Address

ynniv@ynniv.com

Profile Code

nprofile1qqs9wmfrms7myptdyzyyj330ac6ce70s7vcs5trredkzv7jtnavy37gpz3mhxue69uhhyetvv9ujuerpd46hxtnfduqs6amnwvaz7tmwdaejumr0ds28lz52

Publishing to

relay.damus.io

nos.lol

Author Public Key

npub12akj8hpakgzk6gygf9rzlm343nulpue3pgkx8jmvyeayh86cfrus4x6fdh

Show more details

Last Notes

2026-05-31 03:28:23 UTC

- reply

https://blossom.primal.net/ce0405a7f0208ef6344f22a96fe58afb3ae94d04d4a97cc16cba6f78cb9e4bab.png

2026-05-29 23:03:50 UTC

- reply

don't exit 🤷🏻‍♂️ build a self healing network that doesn't die

2026-05-29 22:28:16 UTC

- reply

when you're in ecash you have to assume mints last forever or confront the fact that rugging, even in the most honest case, is inevitable

2026-05-29 22:24:11 UTC

- reply

or just ghost everyone. my solution to this in bitcoinsdeposits.net is to have public ledgers and funds controlled by someone other than the operator, so if they disappear another node just takes over
couldn't figure out how to do this with blinded payments. i'm not sure it's possible

2026-05-29 22:11:08 UTC

- reply

will nitro sign that the document originated in the TEE?
but also... right: ecash.
*sigh* the probability of redeeming an ecash token always trends toward zero
can an AWS HSM do the necessary math? you're still bound to the life of a specific HSM, though

2026-05-29 21:51:05 UTC

nostr apparently needs more hackers to keep people honest

2026-05-29 21:49:23 UTC

- reply

err... why? escrow seems to solve this.
also, you can backdate posts (primal accepts a year or two), so it's even easier on nostr than on traditional social media

2026-05-29 21:46:48 UTC

- reply

curious to see how that goes

2026-05-29 21:45:55 UTC

- reply

"i'm running inside a TEE" isn't verification. you have to publish audited, reproducible builds, that generate the key material, and somehow attest that the key was generated inside a process launched from a reproducible build on a TEE, in order to close the loop
and nitro doesn't even provide a means of key attestation. so you can run outside the TEE and just tell people you're running inside one 🤷🏻‍♂️
TEE's aren't magic. they're a very specific capability, and one that's only valuable to the operator. this is the same problem that lexe and maple ai face
h/t @npub1226…grkj

2026-05-29 01:53:06 UTC

- reply

i would fat zap that 🤙

2026-05-29 01:52:22 UTC

- reply

distinguished!

2026-05-28 16:48:44 UTC

fine.
PSA: if you're using cashu without a vpn or tor, you've defeated all of its privacy guarantees.
the mint knows:
who you are,
where you are,
and who you transacted with
#nevent1q…ar3u

2026-05-28 16:15:55 UTC

- reply

if you don't want replaceable events to be overwritten ... just retain them? no protocol changes required

2026-05-28 15:46:32 UTC

- reply

i'm trying my best to give the benefit of the doubt, man

2026-05-28 09:38:03 UTC

this centralizes the protocol on relays and kills distribution and censorship resistance. it trades one convenience for everything that makes the protocol valuable
#nevent1q…9fuh

2026-05-27 20:01:50 UTC

- reply

"all i have to do is spend $3 in on-chain fees, wait 30 to 60 minutes, then i can send small payments fast and free (and spend another $3 ~ $30 later to close)"
i mean ...
yes, that's why i have a lightning node.
not sure it's the best we can do though

2026-05-27 18:44:48 UTC

is @nprofile…uz3y centralized? yes
is nostr? no
that's all you need to know

2026-05-24 18:57:01 UTC

"but it's vibe coded"
https://github.com/bitcoin-deposits/rust-miniscript/blob/master/VERIFICATION.md

2026-05-24 15:05:16 UTC

- reply

if only there was some way for the network to stay up even if one provide goes down 🤔 maybe a protocol where lots of different people can be providers, in case one goes down

2026-05-19 16:28:59 UTC

- reply

i don't use gpl libraries, and i don't license my own code gpl
but your argument requires the llm to regurgitate code without attribution. if you can't tell where some code came from, by definition it isn't a copyright violation
what you're actually saying is that learning something makes the learner a derivative product, and i think people need to take a long hard reflection on what that would mean for society

2026-05-19 11:59:28 UTC

- reply

"a guy who makes a nice chair doesn't owe money to everyone who has ever built a chair"

2026-05-19 11:33:37 UTC

- reply

great artists steal

2026-05-19 01:54:10 UTC

- reply

still cheaper than the alternatives

2026-05-19 00:36:24 UTC

- reply

err... don't blur sensitive things. put a solid color box over them instead

2026-05-18 21:46:31 UTC

finally bought something from a @npub12rc…r4zx seller using bitcoin 🎉
even used @npub1cas…dg5l
it's happening
https://image.nostr.build/9a40ae37d1ee1ae21735c3932fa956c0e0b222616fd467eb94808c4f43501135.jpg

2026-05-18 20:58:21 UTC

https://image.nostr.build/9a2904d0202d8bfc8880e3a810671980ad78218dfeca2da95fce6ce171391016.gif
#nevent1q…fven

2026-05-18 17:36:17 UTC

- reply

the dollar is stable because it's worth a little less ever month. to say that bitcoin is unsuitable as money because it also goes up sometimes seems to miss the point. but if you onramp and offramp at the time of transfer, then you both have the stability you so desire, and you didn't pay 2%+ on top for the privilege of doing so

2026-05-18 17:30:56 UTC

- reply

it's a weird argument to make because you're implicitly giving the upside to someone else. but ignoring that, you can buy stability on off-ramps. why not buy bitcoin at the time of use, use it as cash, and let the other party park it wherever is best for them? you probably have no idea how much simpler that is than what the world currently has

2026-05-18 17:12:23 UTC

- reply

everything is volatile. some things go both up and down. fewer things more up than down. you're arguing that only things that go down are an acceptable form of money?

2026-05-18 13:59:53 UTC

- reply

yes. what do you think npub to npub spillman channels look like?

2026-05-18 06:07:48 UTC

- reply

channel factories take time
spillman channels take time
lightning channels take time
ecash takes trust
agents don't have time or trust
where's the complexity here?

2026-05-18 05:56:03 UTC

- reply

what if that's
just because
it's hard to use
and then it's not

2026-05-18 04:16:15 UTC

- reply

everyone's got their favorite plan
but the only way we're getting to
a trillion wallets is
off-chain

2026-05-17 19:52:15 UTC

- reply

spillman is back. one might point out that opening a hundred channels will be expensive

2026-05-15 14:21:25 UTC

https://bitcoindeposits.net/FAQ

2026-05-15 10:54:04 UTC

CLAUDE CODE:
plus it closes the 3 regressions from the CAR/CDR fix that I want to stop apologizing for in commit messages
😂

2026-05-14 16:53:34 UTC

- reply

crank the rez, higher dbz cutoff, alpha blend
https://blossom.primal.net/00b1b7b38a59e2d41d094cf97e5e123df2fe4562c062366cb7f756ef823eefe6.png

2026-05-14 16:35:28 UTC

volumetric wx data. like, duh, of course
https://blossom.primal.net/2d823fd7ff1f295b4546d19711b60ec7083895cca29596d8e2090e64d21a26e7.png

2026-05-14 13:03:15 UTC

ask claude and ye shall receive

2026-05-14 13:01:31 UTC

- reply

i concede that advertising a method on behalf of someone else is sketchy. it's an interesting idea though...

2026-05-14 12:52:52 UTC

- reply

kind of strange to say that people throwing money at you are doing it wrong, but i get the gist

2026-05-14 12:24:21 UTC

- reply

sure, but one of those two actually gives you money

2026-05-14 06:53:18 UTC

fun fact – traffic corridors have even altitudes going one direction, odd the opposite [flight tracks colored by heading]
https://blossom.primal.net/4b5e595aa4a5a618d4de2e5a649dd2d6726d5468bf8ed36121606832aea20349.png

2026-05-14 06:35:48 UTC

*of course* we can add a two dimensional, color coded progress bar!
https://blossom.primal.net/cfa445475a29d661d3f0a1a32ab7e25a37d3cf269fd0734abeca1d47f584f34d.png

2026-05-13 23:55:00 UTC

or:
https://bitcoindeposits.net multiplexes 3kb/s into as many bytes per second as the world will pay for, based on a market of whoever shows up

2026-05-13 23:42:57 UTC

https://bitcoindeposits.net trades time and space for fees

2026-05-13 23:41:38 UTC

- reply

i know 😞
sometimes it lets you make an app that you couldn't make before though

2026-05-13 18:15:04 UTC

- reply

spam is a gas that fills the container of empty blocks. fees are the solution. satoshi knew this from the start

2026-05-13 17:07:16 UTC

- reply

no. bitcoin doesn't have a spam issue, it has a lack of economic activity.
use bitcoin as money

2026-05-13 14:47:17 UTC

- reply

i hear that. i built a novel bitcoin layer 2, also self funded, over the last year. and my well designed protocol extension was never really considered, even though people complain about nsec management and key rotation
maybe jack talks about apps because that's where developers have the most exposure to revenue. maybe there's an app form of what you're doing?

2026-05-13 14:10:45 UTC

- reply

like... separating identity from authorization? https://github.com/nostr-protocol/nips/pull/1450
i'm not sure what you're getting at. it seems that you're bitter about something, and i hope there's some way to get past that

2026-05-13 13:30:41 UTC

- reply

i have no idea what you're aiming at, but i've written a lot of code, run a lot of things, and like nostr

2026-05-13 12:30:11 UTC

things claude will build in an hour:
3d webgl ads-b visualization with common lisp backend
https://blossom.primal.net/07b8576ef86e8686ba197051a50d380c49d0153c1262e9ca65fd720cf6e69646.png

2026-05-13 12:27:51 UTC

- reply

there aren't zaps *with* nwc. nwc is app -> wallet

2026-05-13 09:28:03 UTC

- reply

#nevent1q…zuf6

2026-05-13 01:45:11 UTC

- reply

damn. i'm using that one ... whenever i manage to see a tie again

2026-05-13 00:53:09 UTC

- reply

doesn't look a day over 135

2026-05-12 23:34:56 UTC

- reply

https://protocol.dfos.com/spec reads like nostr but complicated.
i'm not a fan of complicated

2026-05-12 21:07:26 UTC

- reply

😅
sadly i'm dealing with machine code errors on both aarch64 and x64. once we're out of that morass i'll get back to graphics. technically modus did output a rudimentary bios-text screen on a pi zero 2 w as part of the atl bitlab demo

2026-05-12 21:04:51 UTC

very few people ran this one. it was more next than macos, and pre-aqua, so it had a more professional feel
https://blossom.primal.net/263dfff1a885d86e9bd620f9d1859c267bbf9995bac021eb12fdea65f030b0cb.png

2026-05-12 20:12:16 UTC

still looks good
https://blossom.primal.net/b127a318947064b7456e59abdb9abccb942e047ed141f7a59cb3c5151bf8f37d.png
http://www.typewritten.org/Media/

2026-05-12 18:33:07 UTC

- reply

there are two different "dark forests". the "dark forest of the internet" is a newer term that means the internet is hostile and you need to protect yourself from it.
the earlier "dark forest" theory is a solution to the fermi paradox: we haven't seen anyone else in the universe because once two civilizations see each other, one of them annihilates the other. it's nothing personal, just that the game theory only supports one "winner", and being found is being extinguished
so, whether you want a "dark forest" social network is highly dependent on which definition you're familiar with

2026-05-12 18:27:15 UTC

- reply

nostr the social network, or nostr the distributed systems layer?

2026-05-12 18:26:10 UTC

- reply

sorry, to answer your other questions, yes, you can send zaps, receive zaps, make lightning invoices, and pay lightning invoices, among other things.
it's lightning without channels. custody without discretion

2026-05-12 18:24:44 UTC

- reply

the protocol is tlv-over-nostr: all you need is string manipulation and crypto primitives. running a node in a browser seems unlikely to be satisfying, because nodes should have better availability that that would provide.
there's an interesting angle here though... 🤔 fees are set by a node. if you didn't care about offline receive, you could have a low availability node where you and some liquidity provider are the only people on it ... now there are no fees and you only need to maintain the availability standards of whatever liquidity provider joins your ledger.
it creates a spillman-like environment with the ability to send funds to any key. but you'd have to rotate every N weeks or lose the ledger to someone else. i don't know if this is better than keeping funds on a ledger run by someone else with fees less than 500 sats-per-quorum expiration period, but it's certainly a shape.
what's more likely for most people is that they pay 1% per year to a random node and get a lightning-like wallet with no availability requirements or channel / liquidity management. now you can write down your twelve words just as you do on the base layer, except per-transaction costs are 0.005% and 2 sats
but, permissionless means no one can tell you how to use it

2026-05-12 18:10:58 UTC

- reply

pretty weird to give something social "dark forest" branding. not sure that means what they think it means. or it does, and ... well that isn't better

2026-05-12 17:01:49 UTC

- reply

you're being logical. markets aren't logical. sell a million BTC, wait for the panic selling, then buy it slowly, in dark pools, after it's "dead". by the time the price comes back you'll have a lot more than a million BTC, and you track down the holdouts
here's the step everyone seems to be missing: as long as there are only a handful of markets, and a handful of market makers, bitcoin will not survive
use bitcoin as money

2026-05-12 16:28:01 UTC

- reply

*burn* the keys? hahahah, no. sell bitcoin to themselves in a restricted market for diminishing, well publicized prices until the world has declared bitcoin a failure
as the saying goes, "the market can stay irrational longer than you can stay solvent"
use bitcoin as money

2026-05-12 15:58:02 UTC

- reply

the revolution will not be well funded

2026-05-12 15:44:24 UTC

- reply

put another way, the fed could print the market cap of bitcoin, buy it all, and bury it. ... probably already are
use bitcoin as money

2026-05-12 12:56:05 UTC

- reply

you don't strictly need to run a full node for lightning, though you're relying on someone to be your watchtower. deposits is the same, except the watchtower role (quorum members) is incentivized to act out of both benefit and obligation: classically, fear and greed.
by requiring that quorum members have at stake at least half of what could be stolen, and requiring collusion to steal, and having transparent operations, it appears that individual ledger collateral ratios around 50/50 may require a sybil group close to half of the network in order to profitably steal

2026-05-12 01:05:13 UTC

- reply

comments appreciated: https://bitcoindeposits.net/FAQ

2026-05-11 21:16:49 UTC

bitcoiners: don't trust, verify
also bitcoiners: you can't trust ai

2026-05-11 03:23:46 UTC

https://image.nostr.build/fe093d88c7ff8e2920b8a5a7f1bca3af2bcb3e93c15dd38dd0f0a86b9fe2fe7d.jpg

2026-05-11 02:16:45 UTC

https://blossom.primal.net/639000608df194c4467e692abbe76796ec5dec4e56d1e02f6697203bb484547b.png

2026-05-10 20:19:58 UTC

CLAUDE OPUS:
Pax Complexia is the genuinely interesting one because it solves the problem we kept circling. Compromat needed pretense; surveillance needs enforcement; intelligencia needs consent over who controls the model. Complexia needs none of these. The peace holds because the system has become too entangled for anyone — including its operators — to model the consequences of disrupting it. 2008 was the proof of concept: not “too big to fail” but “too complex to safely unwind.” Global supply chains are this. Cloud dependencies are this. AI infrastructure is becoming this. Nobody enforces the peace; defection just has uncomputable costs, so nobody pulls the thread.
It’s also the bleakest of the four, because it’s the only one that survives the death of every other Pax. You can stop pretending, stop consenting, stop watching, stop being watched — the knot is still there, and it tightens by accident. Pax by accumulated entanglement. The reason it might be the actual successor regime isn’t that anyone chose it, but that nobody can choose against it.

2026-05-10 16:33:22 UTC

vibe interlocutor

2026-05-08 22:15:27 UTC

- reply

it takes a long time to say things succinctly! and things have changed significantly a few times over the last year as i tested designs. unfortunately 300 sats is too little to run a node, as each quorum rotation (currently every ~ two weeks) requires an on-chain transaction of 180 vbytes, which at 2.5 sats/vbyte is 500 sats every two weeks.
depending on a node's advertised fees you can store just a few sats in a wallet. my first few nodes have high per-deposit fees (208 sats / month) and low limits (1,000 sats max) so that people don't keep funds there yet.
but if you have custodial lightning or ecash, you can still make wallets, move funds into them, between them, back out of them, etc. fees are only taken when the balance is greater than zero, so moving funds through them quickly only costs lightning fees.
it's all pretty loosey goosey until more people start running nodes and a market emerges

2026-05-08 11:56:19 UTC

- reply

before wallets will use a ledger, the operator spends the reserves to a multisig of quorum members. so during operation, the operator controls "truth", but the quorum controls "funds".
a bad operator cannot steal, they can only misattribute.
the majority of a quorum can steal, but in doing so reveal themselves to be dishonest. it is required that quorum members have their own ledgers with their own quorums of at least half the size, so by colluding to steal, a minimum of 2 or 3 other ledgers are at risk
since quorum membership is visible, wallets can choose ledgers where dishonesty would cause maximum fan-out of liability

2026-05-08 05:14:12 UTC

- reply

ah, i fixed those links. thanks
the basic construction is that the operator appends updates to the ledger, but a quorum that they chose controls the onchain reserves utxo. if the operator goes bad, the quorum gives the reserves to a random quorum member, and the ledger is now their responsibility
if quorum members collude and do not reassign reserves in a timely manner, then the honest members of their own quorums will reassign their ledgers. the question is, how far does this scale. in my analyses and simulations, it is farther than expected
there has been more interest today than for many months – i will write more soon

2026-05-07 16:30:25 UTC

- reply

it needs a better one, and some videos

2026-05-07 16:13:07 UTC

- reply

i've spent the last year designing a distributed L2 with claude. there's no mining, just staked cryptographic ledgers enforced by your peer nodes. simple wallets are 12 words, multisig also supported, you can be offline as long as you want, still receive payments, and if your node goes offline or rogue, another node will have your funds when you come back
it's untested, but the theory appears sound. if it does work, it requires less trust and scales better than ark / spark / liquid. it requires transparency so there's no privacy, but you can open a lot of deposits and easily move funds in and out of ecash. all communication is over nostr relays, so only the relay knows your ip address
current state: 100% me running everything. hoping to shake out the remaining bugs and fan things out soon / now
https://bitcoindeposits.net/WHITEPAPER

2026-05-07 13:27:05 UTC

i'm tryin'
https://github.com/bitcoin-deposits/deposits/
#note1cdc…paa2

2026-05-06 20:30:29 UTC

- reply

i'm not sure what that means. does the zk prove that your pubkey belongs to a person, or does the scanner prove that there's an eyeball in front of it?
and if i can prove that my pubkey belongs to a person, why can't people ask me to sign their online requests? i'm a busy guy – i do a lot of stuff online 🤷🏻‍♂️

2026-05-06 20:20:37 UTC

- reply

oh, i meant rent out your zk proof of human-ness.
which completely defeats the purpose, of course

2026-05-06 18:38:53 UTC

- reply

imagine how much money you could make renting it out

2026-05-06 18:29:31 UTC

"the asymmetry isn’t permissioned access, it’s just labor"
bitcoin wallets pay miners for settlement, but don't trust them
deposits wallets pay nodes for liquidity, but don't trust them
https://github.com/bitcoin-deposits/deposits

2026-05-06 18:01:45 UTC

- reply

slightly less invasive than a blood draw. you'd give blood to use our service, *right*?

2026-05-06 01:21:41 UTC

the solution to ai is usually more ai
think about that

2026-05-06 01:16:52 UTC

- reply

if you want to chat around 0200Z somehow, i bet we could crack it in less than a half hour

2026-05-06 01:10:36 UTC

- reply

claude. the answer is always claude

2026-05-06 01:07:48 UTC

- reply

just ask claude to automate an ios gui simulator test. it sounds insane but once you get yourself out of the loop is all gravy

2026-05-06 01:04:35 UTC

- reply

#nevent1q…3hqx

2026-05-06 01:02:20 UTC

- reply

... define "literally no way". i've also closed the loop by giving claude a raspberry pi zero that was pretending to be an sdcard / usb keyboard so it could figure out how to boot a thinkpad from scratch. i also pointed a webcam at the screen and gave it a way to take photos of the laptop

2026-05-06 00:55:57 UTC

- reply

don't give up so easily. i'm having claude reverse engineer an mac classic game, and i just had it modify sheep shaver to accept input over udp, and dump memory and screenshots to disk.
getting yourself out of the loop always pays off in spades, and the best part is that you can ask the ai to help you get out of it

2026-05-06 00:51:40 UTC

- reply

make sure you have tests written, and don't let it change the tests. worst case you can `chmod o-w` them, or even run the agent as a different user without write access. now it's easy to say "don't stop until tests are green"

2026-05-05 20:10:49 UTC

we do what we must, because we can

2026-05-05 19:35:21 UTC

- reply

i appreciate that. it's a liability of the subject – when anything that can be done will be done ... well a lot of things can be done. it's a lot to think about

2026-05-05 16:39:43 UTC

- reply

those certainly reduce circulation, but it's rare that the information is actually gone for good, and look at how much control it requires

2026-05-05 15:52:48 UTC

- reply

can you even do it in the analog?