matt
Matt Corallo

https://bluematt.bitcoin.ninja

10th known contributor to Bitcoin Core. Now Full-Time Open-Source Bitcoin+Lightning Projects at Spiral (Part of Block).

Public Key

npub185h9z5yxn8uc7retm0n6gkm88358lejzparxms5kmy9epr236k2qcswrdp

NIP-05 Address

Profile Code

nprofile1qqsr6tj32zrfn7v0pu4aheaytdnnc6rluepq73ndc2tdjzus34gat9qpp4mhxue69uhkummn9ekx7mqpz3mhxue69uhhyetvv9ujuerpd46hxtnfduq37amnwvaz7tmwdaehgu3dwfjkccte9e3xjarrda5kutnwd9hx5cgpzamhxue69uhhyetvv9ujucm4wfex2mn59en8j6gpzamhxue69uhkummnw3ezuendwsh8w6t69e3xj7sprpmhxue69uhhyetvv9ujuumwdae8gtnnda3kjctv4k0zu7

Publishing to

nos.lol

relay.damus.io

nostr-relay.bitcoin.ninja

relay.current.fyi

nostr.fmt.wiz.biz

relay.snort.social

Author Public Key

npub185h9z5yxn8uc7retm0n6gkm88358lejzparxms5kmy9epr236k2qcswrdp

Show more details

Last Notes

2026-07-20 00:53:17 UTC

- reply

They all just had fun.

2026-07-19 21:12:16 UTC

- reply

Hell yea. Congrats.

2026-07-18 01:01:55 UTC

- reply

It seems a better answer is to probe some (or download results from others). That way you can discover these channels have capacity before you need to use them.

2026-07-18 01:00:21 UTC

- reply

Huh? No of course you should store a history of received payments, but that’s very different from DoSing yourself by storing generated invoices.

2026-07-17 16:44:49 UTC

- reply

LDK Server…

2026-07-17 16:39:37 UTC

- reply

Payment metadata is more than sufficiently deployed to use it in production today.

2026-07-17 11:42:10 UTC

This idea that lightning nodes, by default, store every single invoice or offer you generate over RPC is insane. There’s no reason to bloat the DB and expose users to DoS risk if they generate invoices automatically, the damn things can be stateless and additional metadata you need can be stored in the invoice and retrieved on payment, rather than stored.

2026-07-12 20:51:41 UTC

- reply

Worst. “Professional” is code for “something a business would buy”, which also means “you’re not the one using this”.

2026-07-08 21:25:33 UTC

- reply

Also heuristic-based replacements please. Strict diagram comparison is nuts and enables pinning attacks!

2026-07-08 19:35:28 UTC

- reply

Honestly why bother? No one uses Spark on the assumption that they can exit or that it’s somehow trustless, the point is that it’s better than custodial and Lightspark is taking on the legal risk instead of someone else.
If you actually care about trustless something you’ve gotta do a graduated wallet or LN+Ark.

2026-07-08 19:33:41 UTC

- reply

Yes…

2026-07-08 17:19:18 UTC

- reply

@npub1vad…nuu7 and @npub1ke4…8jfm since damus doesn’t carry over RT tags.

2026-07-08 17:18:27 UTC

- reply

Problem is any time you transact (and sometimes other times!) the “recovery bundle” changes. So you can’t back it up/download it, your wallet needs to support this natively (and I don’t believe any do).

2026-07-08 12:09:19 UTC

- reply

Zero spark clients today implement unilateral exit without asking a server for their list of coins to exit. Makes it…. Not very unilateral.

2026-07-07 21:15:31 UTC

- reply

Or on top of. L-ARK seems like kinda the best of both worlds if you’re worried about low-balance user onboarding.

2026-07-07 20:08:03 UTC

- reply

I disagree, you could do a graduated wallet, or you could have an LSP that locks up less liquidity and charges comparatively higher fees. Of course in that case the LSP would never need to charge more than on-chain - you still get *some* “compression”.

2026-07-07 18:41:12 UTC

- reply

None of the constraints you raise are required for lighting, they’re just things people did to make simplified versions of the economics of an LSP work out. An LSP putting in a bit more effort, or conjoining the wallet with a graduated wallet design, makes it way more doable.

2026-07-07 15:19:35 UTC

- reply

lol no I mean a mobile app with lightning built in.

2026-07-07 14:59:12 UTC

- reply

Who said anything about hosting a routing node? That’s not what I was talking about at all…

2026-07-07 14:44:30 UTC

- reply

Yea, I think this is the common experience these days. It’s certainly not evenly distributed - there’s lots of terrible lightning experiences out there. But the fact is it’s very clearly possible to make better, so it’s time to put pressure on the developers shipping bad experiences, understand where their applications are going wrong, and help them get to where we know they can be.

2026-07-07 14:42:04 UTC

- reply

Not with halfway competent self-custodial either. I agree there is lots of incompetent self-custodial tho.

2026-07-07 14:41:01 UTC

I fully do not understand the desire of lightning people to build protocols to “buy a channel”. No one has ever wanted to “buy a channel”, they want to buy reliable HTLC delivery. That might imply multiple channels, splicing, and probably implies just-in-time channels, but it definitely doesn’t imply just buying a channel and calling it a day.

2026-07-07 11:11:49 UTC

- reply

> Lightning fails in about 30% of the time
Holy shit what? Let’s debug this. Which issues have you opened to the Lightning implementations and providers you’re using? This is not the experience most have.

2026-07-07 11:10:00 UTC

- reply

Yes, *lighting* has good sender privacy. Some people prefer to use custodial services which don’t. Not entirely sure you can blame LN for that? Though even for such services you can do client-side pathfinding, which admittedly isn’t all *that* common but is very doable and some protocols support.

2026-06-30 22:06:42 UTC

- reply

For now, for the most part, the explicit censorship is done at the hosting provider level and model devs are focused on quality not manipulating output. I imagine that won’t last forever.

2026-06-27 16:30:30 UTC

- reply

Now do inflation-adjusted.

2026-06-24 22:18:41 UTC

- reply

Learn to read.

2026-06-24 11:50:07 UTC

- reply

*on our forge. If someone wants to make a bot that streams moderation decisions onto X or nostr-social-media and people want to discuss those decisions, great! They should! But distracting people trying to focus and work with harassment and brigading on the platform where people work isn’t an option.

2026-06-24 11:44:15 UTC

- reply

Obviously it’s just a question of thresholds. In your specific scenario obviously that’s fine. I may be missing something but AFAIU there’s some kind of nostr standard for how to do git forge on nostr, with multiple clients implementing it. Today, none of those clients support moderation (or, equivalently, restricting events to specific relays per repo). In the future, if all the common clients support that, I don’t see why that wouldn’t suffice. Given the narratives in nostr I assume that won’t be the case, but I could be wrong!

2026-06-24 10:52:22 UTC

- reply

I’m not sure, there’s plenty of bitcoiners with a following who choose to spend their time shitting on open-source devs without any factual basis…

2026-06-24 10:46:50 UTC

- reply

#nevent1q…56ha

2026-06-24 10:46:16 UTC

- reply

I imagine most clients will pull from multiple (public) relays, but I could be wrong.

2026-06-23 13:25:52 UTC

- reply

No I think you misunderstood the requirement. A client-selected moderation option isn’t sufficient, nor is a client-software-specific one. A repo forge is our place of work. People will (and have) taken the fact that allow anyone to comment as a free pass to harass people, throw baseless accusations, and generally be disruptive. They can of course go do that on social media but we can’t have our place of work disrupted by people using it as a bulletin board to shit on us. Having the ability to hide it for ourselves doesn’t suffice - when anyone else looks at our work and sees a stream of “pedos!” and “compromised!” conspiracy theory nonsense that makes it our problem. A nostr-based solution where every client hides crap that we removed would be fine, but I assume it’s quite likely clients would be built that would ignore such moderation and we’d be back to square one.

2026-06-22 23:03:44 UTC

- reply

For now we’re moving to self-hosted forgejo. The current nostr solutions aren’t an option for us as they lack moderation features.

2026-06-20 12:39:19 UTC

- reply

If they’re polarized the right direction, use sunglasses - 50% less brightness from the rest of the world, no chance to the laptop brightness.

2026-06-17 18:23:30 UTC

If you can’t kill it by regulation, kill it by taxation.
#nevent1q…h0ey

2026-06-13 13:34:36 UTC

- reply

Yea, props to Nvidia for investing in remote attestation. The high cost of compute still sucks for everyone, though, even remotely.
3N billion parameter models have gotten much better and most can run them, but the intermediary sizes where hardware is expensive but practical for some ($5k, ish) have seen a drop off in investment as more compute goes into the frontier sizes.

2026-06-13 13:23:27 UTC

- reply

(btw Sparks are way overpriced for what they offer, really shouldn’t buy those at this point!)

2026-06-13 13:22:59 UTC

- reply

I’ve got local GPUs with almost as much memory as a Spark (but way more BW). I was mostly tongue-in-cheek referring to the annoying reduction in 120B-class models lately and the increasing reality that not many can afford to access frontier models lately.

2026-06-13 13:02:48 UTC

- reply

Will HRF give me a grant for $50k so I can buy the GPUs I’d need to run them? :)

2026-06-10 16:58:53 UTC

Two weeks of crazy travel and suddenly I can barely walk out of the gym after my workout. Blood sugar is rough.

2026-06-08 21:17:05 UTC

- reply

Usually when you hear about a custodian storing money with another custodian it’s in the context of fraction reserve or a custodian-failure :)

2026-06-08 21:02:47 UTC

- reply

I mean sure, it’s pretty weird for your custodian to store your money with another custodian, though?

2026-06-08 20:48:56 UTC

- reply

We figured there wasn’t enough custodian so we added a custodian so you can have a custodian inside your custodian.
Max fractional reserve lfg.
#nevent1q…y6q3

2026-06-08 20:47:02 UTC

- reply

lol wtf

2026-06-07 11:10:46 UTC

- reply

I mean that is a really huge portion of “real” trading volume (ie not counting market makers making short-term bets or rotations from self-custody to ETFs).

2026-05-30 19:33:25 UTC

- reply

To support failover you can copy the keys from one TEE to another in a secure manner such that they still cannot leave an equivalent TEE running the same software. The issue of course is if the operator can replay withdraws across the TEEs.

2026-05-30 14:03:45 UTC

- reply

Sure if you’re using something that can do remote attestation you know the other TEE is running the same software and you can copy it. Tho I guess maybe replay is an issue?

2026-05-30 06:28:58 UTC

- reply

Shouldn’t you have redundant TEEs that, using remote attestation, copy the keys so that you have a backup?

2026-05-29 08:52:19 UTC

Lighting Airdrop: Apparently Coming Soon.
Ask @npub179e…lz4s for your free money.

2026-05-20 02:09:08 UTC

Wow Massie lost. This country man…

2026-05-06 11:43:32 UTC

Strategy selling is, in fact, good for Bitcoin.

2026-05-01 10:27:09 UTC

- reply

Ah, my recollection was that only a fixed (large) set of parties could challenge the withdraw with the fraud proof, so it was also a security assumption, but maybe that was BitVM v1? The nice thing about the BitVM machinery even in that model is you get 1-of-N liveness *and* 1-of-N security which isn’t possible with a multisig.

2026-04-29 01:06:13 UTC

- reply

My understanding is that you were still limited to a 1-of-N for the fraud proofs anyway.

2026-04-28 16:58:18 UTC

- reply

IIRC the version of BitVM they shipped already had a 1-of-N assumption with relatively small N so it’s not a change in trust model. Even latest BitVM has that but N can be bigger.

2026-04-23 16:58:58 UTC

Every stablecoin loves freezing tokens. These tokens will never be redeemed. They are now forever assets tether can invest as they wish and get returns on.
#nevent1q…6vfx

2026-04-23 10:36:21 UTC

- reply

IOW, personally, I came to the conclusion ECC should be disabled (in conjunction with a commit-reveal scheme to allow pre-BIP-32 wallets to retain ownership without revealing they still have the keys) precisely because it is *less property-rights-violating than the alternative*.
Yes, it’s “us” violating some property rights instead of “them” violating all property rights, but I think the distinction is really weird? Like, it feels like shoving our heads in the sand to pretend reality doesn’t exist (once a CRQC is undeniable) instead of facing a tough decision.

2026-04-23 10:30:21 UTC

- reply

People are going to say what they say. They might also reasonably say that cause Mark Karpeles was proposing a fork to reclaim the mtgox funds. But as much as you try to argue these are similar situations, they’re just really not?
Fundamentally, in a world where a CRQC exists and more will later, your fundamentally cannot have a concept of property rights for coins secured by ECC. “If you have the key” is nonsense when everyone has the key!
More generally, I posted this yesterday:
I think its actually an interesting question as to whether allowing a BIP 32+seedphrase-based recovery scheme that also necessarily freezes some coins as a result is more or less property-rights violating than doing nothing.
My view of Bitcoin's property rights has always been that someone who bought bitcoin, took self custody and wrote down the seedphrase 10 years ago then promptly forgot about bitcoin and went and lived their life should be able to retain access to their coins no matter what.
For that user specifically, a recovery path actually *retains* more property rights than it loses. But for patoshi its a different question. Obviously you can do a commit-before-Q-day-reveal-later scheme for patoshi to retain ownership without revealing whether they still have the private keys, but they do have to do something.
Its tricky but ultimately I don't think the "freezing coins is violating property rights so its bad" view is fully thought out - freezing some coins actually allows owners of other coins to retain the coins! Its far from black-and-white.

2026-04-20 14:11:44 UTC

- reply

There does seem to still be a lot of people talking past each other. There’s a big difference between “this is bad and I don’t support it” and “I don’t think the market will support this”. I’ve seen a lot of arguments to the first but relatively few of the second. Ultimately, the view of quite a few people I’ve spoken to (and myself) is “this is bad but it’s not clear that it’s worse than the alternatives and the market will very likely support this”. Of course “this” is not some premature seizure but rather only freezing coins when it’s very obvious to all involved that anything else will result in theft.

2026-04-17 13:09:13 UTC

- reply

To be fair, there would likely be a futures market into which people could sell pre-fork.

2026-04-17 13:08:25 UTC

- reply

Yes and I can also get a rotating line of credit.

2026-04-15 15:23:40 UTC

Y’all remember Long Island Blockchain Corp?
https://www.cnbc.com/2026/04/15/allbirds-bird-stock-shoes-ai.html

2026-04-15 15:09:09 UTC

lol this is a good point. Is literally cheaper for me to take a personal, non-collateralized loan to buy bitcoin than for strategy to.
#nevent1q…8xqg

2026-04-10 22:40:43 UTC

- reply

Sonnet, not Opus. Supposedly GLM is competitive but you ain’t running that thing locally (unless you spent $50k on GPUs…assuming you can even get them, which you probably can’t).

2026-04-07 10:07:52 UTC

- reply

Holy shit thank God.

2026-04-04 10:11:03 UTC

- reply

Eh, shove an option into a tap tree leaf and let wallets embed a backup key there for zero cost - don’t need it? Don’t use it. If/when Q-day comes (or secp gets weakened by traditional computers) you can start using the backup path and we can work on adding something cheaper based on what’s available at the time.

2026-03-31 13:47:09 UTC

New Quantum News.
Bitcoin Flat.
BuT pEoPlE aRe SeLlInG bEcAuSe QuAnTuM.
Not the biggest reason for price action, stop. Yes we should add validation for hash based signatures in a soft fork, obviously. Market seems to assume that will happen (pretty reasonable!)

2026-03-31 13:45:01 UTC

- reply

This was one of the more expensive ones! (Nordic Naturals). I’m honestly surprised any do.

2026-03-31 13:44:17 UTC

- reply

I assume that’s true yes but my point was more that if someone wants a supplement for 3 it’s probably because they have a bad 3:6 ratio (like the vast majority of Americans…. Because of sunflower oil)

2026-03-31 00:50:35 UTC

An Omega-3 supplement with sunflower oil? I mean whatever you think about seed oils why would you include omega 6 in an omega 3 supplement???

2026-03-30 23:15:09 UTC

- reply

Except they’re not fucking yup my life. They’ve been doing it for twenty years now. Chill man.

2026-03-30 23:14:41 UTC

- reply

I’ve gotten it several times!

2026-03-30 22:09:17 UTC

One day I should like to meet the person who’s been using my social since before I was 18. I mean I appreciate that they don’t use it for banking or credit, I’d just buy them a beer, really.
https://image.nostr.build/acaae3d112405de1f8bc565a7be0acdb9a4fe8bb9e072b87cb2812b3f54d0282.jpg

2026-03-27 02:08:16 UTC

https://www.techdirt.com/2026/03/26/everyone-cheering-the-social-media-addiction-verdicts-against-meta-should-understand-what-theyre-actually-cheering-for/
#nevent1q…fewe

2026-03-27 02:00:26 UTC

Oooffffff
https://image.nostr.build/a0b088ac4e96c11c75784e8e421baee7d0e88c37e4097c9aaca7a0de7ab3b155.jpg

2026-03-25 10:21:39 UTC

- reply

Why do people think any of this stuff is going to impact price? Isn’t the whole reason “Bitcoiners” hate the payments usecase is cause the merchant usually sells the Bitcoin so it hurts price?

2026-03-21 13:07:58 UTC

lol Iran is winning the war
https://image.nostr.build/1d507ebbca4e6ac4ac1035b81187e8090f87aad46297aa5161dd1580a338eac9.jpg

2026-03-17 01:18:34 UTC

https://image.nostr.build/636d2b7662fd8f05ef5afe63f2c53aaf9d1ad985f0bd55bfe1cafc3212df0b1b.jpg

2026-03-09 14:19:06 UTC

It’s time to build. Everyone who cares about Bitcoin has to chip in if we want a chance at competing with the big players. The time to build momentum is now.
#nevent1q…fl8l

2026-03-07 16:47:50 UTC

- reply

🤮

2026-03-07 14:44:09 UTC

- reply

Lemons always have some sugar. “Sweet” they are not but I find them more than sweet enough to drink, was my point.

2026-03-07 13:06:04 UTC

- reply

Organic lemons.

2026-03-07 12:46:07 UTC

I don’t understand why people add sugar to lemonade. Fresh squeezed lemon juice is plenty sweet.

2026-03-06 00:34:36 UTC

There have been more posts about OP_RETURN than there were > 80b OP_RETURNS

2026-03-04 12:46:32 UTC

- reply

Ubuntu was always shit.

2026-03-02 15:11:36 UTC

- reply

I don’t care about up/down, I care about correlation.

2026-03-02 14:52:56 UTC

- reply

It was originally included in AP2, but then x402 was created and then coinbase kinda stopped accepting PRs and seems focused entirely on USDC…not sure it’s worth bothering with vs just using L402.

2026-03-02 00:46:37 UTC

S&P Futures down 1.5%, Bitcoin up since Friday. Wild world.

2026-03-01 21:22:44 UTC

AI gives sound money another shot on goal. Payments are going to be disrupted, we’ve got a great shot at making sound money a big part of payments if we build.
#nevent1q…s50q

2026-02-28 22:28:59 UTC

A reminder that Ayatollah Khamenei was far from the most hard-liner of Iran’s ruling class. As with so many other regime-change missions the replacement may well be worse. For the sake of the Iranian people I certainly hope not.

2026-02-25 20:02:19 UTC

- reply

Meanwhile the Bitcoin community is focused on driving devs out as best they can 😭

2026-02-25 14:47:01 UTC

Now is Bitcoin’s time to build.
#nevent1q…rlvl

2026-02-25 14:46:41 UTC

https://spiralbtc.substack.com/p/open-source-ai-needs-to-get-serious

2026-02-25 14:26:07 UTC

- reply

Yea, guess I’m just lacking on specific use cases. Would have to think about it more…crypto seems a long ways off though.

2026-02-25 13:52:04 UTC

- reply

Also AFAIU we really need “multi-input”/function-hiding functional encryption, just “just” FE.

2026-02-24 22:36:05 UTC

- reply

Sorry, I meant 321, not 353 🤦‍♂️

2026-02-24 21:49:18 UTC

- reply

Well why not? You can now encode Cashu payment requests in a BIP 353, and it would be nice to have one standard so that every Bitcoin wallet can pay every other Bitcoin wallet and people don’t have to care about “Lightning” or “Cashu” or other nonsense they don’t want to understand.

2026-02-24 19:43:13 UTC

- reply

I assume this is just communicating a BIP 353 over NFC (so that it’s also compatible with Phoenix and others)?

2026-02-24 15:16:17 UTC

- reply

Yes. Bad photoshops and at least one pointed to a segwit wallet and claimed it predated the existence of segwit by almost a decade.

2026-02-24 13:12:53 UTC

There appears to be a bunch of “news” accounts on X shitposting about huge whales dumping 100k coins at a time/selling their entire 2011 positions with faked screenshots and are reliably getting 50k RTs and not corrected. 🤦‍♂️