No silver bullet for Bitcoin's engineering

Leo Zhang

By Leo Zhang

"There is no single development, in either technology or management technique, which by itself promises even one order-of-magnitude improvement within a decade of productivity, in reliability, in simplicity." -- Frederick P. Brooks, Jr.

Two weeks ago Bitcoin Core fixed a critical bug by issuing a quick software update. In the past decade Bitcoin as an ecosystem has experienced a tremendous amount of growth, and every day there are new endorsements from institutions and high-net-worths. The media attention might create the impression that Bitcoin is a mature and highly developed product. The emergency update is a sober reminder that it's not.

Cryptocurrency systems are incredibly complex, comprising concepts from distributed systems, networking, cryptography, and psychology. Nobody has a clear understanding of how these systems will evolve, and nobody can confidentally guarantee these systems are fully secure. If someone claims they can, they're selling something.

In Frederick Brooks's famous paper No silver bullet - Essence and accidents of software engineering, he separates "accidental difficulties" from "essential difficulties". The former arise from the production process; and the latter from the inherent complexity of the tasks at hand. In Bitcoin's case, both "accidental difficulties" and "essential difficulties" are prevalent. Take it from Matt Corallo, one of the most active Bitcoin contributors:

From pizza to lambos: charting Bitcoin's first decade
(Bloomberg, by Eric Lam and Lauren Leatherby)

"Here’s a look at how the purchasing power of that original transaction has changed over the past eight years, running the gamut from iPads to Lamborghinis (or Lambos, in industry parlance) as early adopters, true believers, Wall Street financiers and retail investors all jumped on the Bitcoin rollercoaster."

Technical Updates

Code reusability vs. accidental commonness
(Hugo Nguyen)

"In software there is this grand vision of software components being perfectly modular as their physical engineering counterparts. Like that laptop charger or USB wire, you don’t have to carry a different charger or USB wire everywhere you go.

So there’s always a strong push for code reusability. Writing redundant code is often frowned upon. Why do the same work twice when you can do it once?"

Argument against scaling the block size

Digital signatures
(Scaling Bitcoin Tokyo 2018, by John Newbery)

"The prover signs a message m as follows: set z as the leftmost bits of H(m) (the hash of a message). He then picks a random nonce scalar k. He sets K = kG and r as the x coordinate of K. So K is a point on the curve and he just picks the x coordinate, and then he sets s as the inverse of k multiplied by (z + rx). The ECDSA signature is the pair (r, s)."

Prof. Andrew Miller on how cryptographers model attacks

Creating money out of thin ether
(Decrypt Media, by Tim Copeland)

"Etherdig, the mining pool capitalizing on the loophole, has mined over 1,250 blocks in the last three months, without validating a single transaction. As a result, it’s received 3,750 ETH ($862,500) in mining rewards. Instead of gathering transactions, confirming them and including them in blocks, it has been creating blocks that just contain the phrase, 'Interim Global Authority,' a reference that appears to be related to the popular computer game, Colony. Etherdig did not reply to queries from Decrypt seeking comment."

News & Commentary

Big institutional investors are buying large amounts of Bitcoin in OTC market
(NewsBTC, by Joseph Young)

"While some cryptocurrency market data providers estimate the daily trading volume of Bitcoin to be around $4 billion, according to ShapeShift’s, which filters out inflated volumes and cryptocurrency exchanges suspected of utilizing bot trading, the real volume of Bitcoin hovers at around $2.7 billion."

With rambling Clinton keynote, Ripple is sending a clear message
(CoinDesk, by David Floyd)

"Not only does it seemingly vault Ripple into a small circle of establishment firms that book Clintons: Goldman Sachs, Moran Stanley and Deutsche Bank. It simultaneously distances Ripple even more emphatically from the world of fly-by-night ICOs, dark web markets, Ponzi schemes, tokens, sectarian hard forks and meat-only diets – the 'crypto' scene that Ripple clearly wants nothing to do with."

Why I'm worried about Google
(Slate, by Matthew Green)

"In short, I fear Google is well on the way to becoming a different kind of company, and it worries me. This is not because I inherently love Google—it’s a profit-making entity, and its shareholders will always come before me. But I worry that it is increasingly trading away my trust for short-term benefits. Even worse, this course change indicates that companies’ self-interest in maintaining user trust may not be a match for the business pressures that drive them to become more intrusive."