Going the long way around is faster: Erlay's road to development. Bitcoin Tech Talk Issue #178

In a world where websites get launched in a matter of days and new features rolled out on SaaS applications on a weekly basis, Bitcoin looks like it’s going too slow. This is often the critique of many altcoins which argue that their coin adds features faster and that they’re more willing to “experiment”.

What a lot of people used to the web-speed of development forget is that security is a critical element of Bitcoin and has far bigger consequences when it’s breached. An error on a website just leads to an error page. An error in Bitcoin may take down the entire network, lead to inflation, or even centralize the currency. One need only look at the Windows ecosystem in the late 90’s to the early 2000’s to see how drastic the consequences of lax security can do to critical systems.

Image result for move fast break things comic

Erlay is an idea that Gleb Naumenko started thinking about 2 years ago. The idea was born out of necessity as a large percentage of Bitcoin network traffic was tied up figuring out which nodes needed what transactions. The current way in which nodes figure out which transactions need to be transmitted to whom is done inefficiently. Each node broadcasts a list of transactions it has putting the onus on receiving nodes to figure out which transactions they need. As transaction ids are 32-bytes, multiplied over many thousands of transactions and many thousands of nodes thousands of times per day, this takes up a lot of the bandwidth.

Erlay changes the equation by making the broadcasting node give a summary of the transactions it has instead of each transaction id. The specification is in BIP330 and involves a technique called set reconciliation. This article from June of last year explains in non-technical terms the tech behind Erlay. Essentially, Erlay adds network messages (NOT protocol changes) to figure out which transactions any pair of connected nodes need to transfer in order for each to have the same set efficiently. This saves bandwidth requirements by around 40% according to simulations while keeping the propagation of new transactions the same.

So after 2 years, here is the pull request implementing Erlay. Kudos to Gleb and the many reviewers of the proposal for getting this change in!


Lloyd Fournier has a post the the dev list about the security assumptions around Taproot. Schnorr assumes that SHA256 is cryptographically secure, that is, finding a hash collision requires brute-forcing. Taproot turns out to require a slightly more general collision resistance assumption. The assumption is that finding two hashes with a fixed difference between them also requires brute-forcing. The assumption seems intuitive, but is slightly more general than the ones Bitcoin has thus far used. He’s created a poster to summarize the research. Warning that this is highly mathematical and not easy to parse, but it’s worth taking into account new security assumptions.

Image result for hardware cartoon

Stepan Snigirev has created a video tutorial for making a hardware wallet from commodity parts! I love that we’ve progressed to a point where something like this is possible. Note that the wallet is still experimental, but if you want to play with this technology, a few hundred dollars is not much to spend to strengthen the security of being your own bank.

Square continues to fund the Bitcoin ecosystem. If you are working on an open source project that benefits Bitcoin, consider applying for a grant! Grantees will be expected to present an execution plan for their project and justification for why they can achieve said plan. Interested individuals or groups should email grants@squarecrypto.org with a proposal outlining how the project benefits Bitcoin, how the individual or team is suited to accomplish the project and how the project fits the criteria from the post.

Image result for funding comic

Speaking of funding developers, Goldman Sats has a really thoughtful article on the problem of developer funding. As he points out, there are lots of “Bitcoin” companies that are free-riding off of time-consuming work to the protocol while actively promoting altcoins and the like. Some sort of market punishment mechanism for bad actors and a market reward mechanism for good actors is definitely worth considering.

Luke Dashjr has written a guide on verifying Bitcoin Core software from scratch. This is a step too few people actually follow through with, but verifying signatures, sha256sums and fingerprints is an essential part of any node setup and this guide is a good start to getting used to doing those things.

Exodus has come out with a Bitcoin full-node router. It’s hard to know exactly what’s running and how much privacy such a node preserves and if you’re a cypherpunk, you’d install your own node on your own hardware, but this is a good first step.


Bitcoin Magazine has published the results of some research using LNTrustChain. LNTrustChain was a lightning torch passing event back in January of this year. The trustchain LN transactions were what Tony Sanak of Bitcoin Magazine analyzed. One of the main findings is that 2/3rds of Lightning node operators from the trust chain torch passing were using TOR. The article also points out that backing up lightning channel states is not on by default in any lightning node setup (e.g. LND and c-lightning) and implies that it should given the ubiquity of TOR use.

Image result for research comic

From a privacy perspective, the on-chain channel opening transactions leak some data, so the article argues for some coinjoin transaction before channel opening. Overall, I found the article very insightful and useful for establishing best-practices on running your own Lightning node.

Economics, Engineering, Etc

Cory Klippensten has written on the intransigent minority in Bitcoin. He argues that societal change happens through convicted people, often a tiny minority of the population. He estimates that 2-3% is around the tipping point at which Bitcoin becomes the dominant currency. The article is well worth reading and if true, we would be almost half way to the tipping point.

Dergigi has an interesting article on where Bitcoin lives. Is it in code, software, computers, people? He argues that Bitcoin is in the intersection of all of these things and that it’s more akin to a living organism than anything else.

If you’re wondering about the price dump in the past 24 hours or so, you might be interested in this story from Chainalysis. Apparently, the PlusToken scammers have been dumping bitcoins during every price rise since December and there’s decent evidence that this is what caused the dump in the past 24 hours. It’s nothing conclusive, obviously, but it’s worth watching the story.

Image result for authority cartoon

Remember how Justin Sun bought Steem? It turns out that wasn’t the Steem network itself, but the Steem corporation. He’s apparently used the currency of Steem to use its voting power by cooperating with exchanges to essentially take over the network and prevent a hard fork. These are the perils of a distributed Proof-of-Stake scheme, or any PoS scheme, for that matter. Exchanges have an outsized say in what the real chain is because most people store their altcoins on exchanges. It’s hard enough to secure your own bitcoin, imagine trying to secure your own Steem or IOTA!

In a first, the US government is suing 113 Virtual Currency accounts. Most of these are BTC and ETH addresses that were alleged to have been used by North Korean hackers to launder money that they stole form South Korean exchanges. This is a really odd lawsuit as it’s targeting the accounts themselves as presumably, it will be difficult to actually prosecute the alleged perpetrators. The only people that are being charged are two Chinese nationals who essentially converted the BTC/ETH to CNY for the alleged North Korean hackers.

Image result for product market fit comic

OpenBazaar the company has essentially laid off one of its founders. This was a company who bet on Bitcoin becoming popular as a method of payment. They haven’t had much traction in 5 years of operation and it wasn’t a surprise when they started supporting altcoins, particularly BCH after Segwit (one of their developers is a BCH core developer). The sad reality is that there isn’t much demand for crypto as a payments technology and if it was, this company would have been a lot more successful than it has been.

Podcasts, YouTube, Speaking Gigs

I talked about a lot of the Bitcoin developments coming in 2020 in this interview with Nugget’s News. I talk about eltoo, dandelion, Schnorr, Taproot, cross-input signature aggregation and carnivory.

I read through my last newsletter and talked a bit about Trace Mayer on my show on Monday. I spoke about MWC and the scam that it is on Tone Vays’s show as well. Lastly, I read through the updates to BIP340 on Whitepaper Wednesday.

Two of the conferences I was scheduled to speak at in March and April have been cancelled. The Bitcoin 2020 conference in San Francisco and The Bitcoin Reformation conference in Antwerp. As such, my seminars in both venues have also been cancelled.

While you’re at home, it might be a good time to read or listen to my books.

Fiat delenda est.