Peter Todd's avatar
Peter Todd
pete@petertodd.org
npub1ej49...ndrm
Peter Todd's avatar
Peter Todd 2 years ago
tl;dr: I have a Libre Relay prototype available for testing. It does three things: 1. Removes the OP_Return limits. 2. Connects to an additional four Libre Relay nodes, to ensure transactions propagate. 2. Implements pure replace-by-fee-rate (RBFR) and full-rbf, to solve Rule #3 transaction pinning. I'm already running it on a few mainnet and testnet nodes. The combination of #1 and #3 is interesting, because an objection to RBFR is it makes possible certain expensive DoS attacks. And #1 should piss off some people... So we'll see if the DoS attacks are serious enough that those pissed off people actually attack it. 😁 Background on RBFR:
Peter Todd's avatar
Peter Todd 2 years ago
“Edmund Burke drew a distinction between the sublime and the beautiful in the 1750s. The beautiful, such as well-watered flatlands, are conducive to human flourishing. The sublime, such as the Alps, can kill you. the sublime was only beginning to come into fashion in Romantic Age.”
Peter Todd's avatar
Peter Todd 2 years ago
Full-RBF Peering Bitcoin Core v26.0 Released Available from: eg: git clone -b full-rbf-v26.0 What is this? It's Bitcoin Core v26.0, with Antoine Riard's full-rbf peering code, and some additional minor updates to it. This does two things for full-rbf nodes: 1) Advertises a FULL_RBF service bit when mempoolfullrbf=1 is set. 2) Connects to four additional FULL_RBF peers. Doing this ensures that a core group of nodes are reliably propagating full-rbf replacements. We don't need everyone to run this. But it'd be helpful if more people did. As for why you should run full-rbf, see my blog post: I'm already running v26.0 on a few nodes with v2transport=1 enabled. You should too!
Peter Todd's avatar
Peter Todd 2 years ago
One-Shot Replace-By-Fee-Rate tl;dr: we can solve Rule #3 transaction pinning attacks by simply replacing based on fee-rate, if the old transaction wasn't going to be mined in the next block or two, and the new transaction will be. This _is_ incentive compatible, because it makes more sense to make money now with a high fee rate transaction than to favor a transaction of lower fee rate that may never even be mined. I'm thinking of releasing a simplified implementation of this in the Libre Relay fork I'm doing. I have at least one mining pool interested in running it. Will be interesting to see if people even bother attacking it.
Peter Todd's avatar
Peter Todd 2 years ago
What did the physicist say to the guy in the faraday cage? “You're in charge.”
Peter Todd's avatar
Peter Todd 2 years ago
https://blog.wasabiwallet.io/friends-and-plebs-dont-pay-w/ A neat thing about this is that “Friends Don’t Pay” is actually the a only possible way Wasabi coinjoins could work, given that they've decided to make remixing free. Why? Because Wasabi coinjoins are private, so it's impossible¹ for Wasabi to know if a UTXO came from the same user, or their friend. And yes, I've taken advantage of this inherent feature multiple times, getting paid by clients who also used Wasabi. 1) Technically they could limit free remixing to amounts equal to that of the Wasabi standardized amounts. But even if they did that you could just send your friend one of those standardized amounts.
Peter Todd's avatar
Peter Todd 2 years ago
I'm starting a new business in the chaotic world of farm equipment. My first product: "Lorenz, A Tractor”