After it became apparent that Adam Back had been unable to “broker peace,” with Blockstream’s president stating he could not tell developers what to do, the conversation turned towards trying to understand why some Bitcoin Core developers strongly hold the view that bitcoin should be turned into a settlement system.
We began by advancing the other side’s point, arguing Back wants “a settlement system and that’s it.” “No, I would like an ecash system.” – Back said. “A settlement only system is not so interesting. I think the interesting use cases that actually benefit from permissionlessness unseizability etc are only in a range of values. If it’s truly bank level settlement it doesn’t need mining, won’t use bitcoin etc. but bitcoin is extremely optimised, algorithmically, and micro-optimised. We are just at the limits.”
We asked if he knows Hal Finney, a former PGP developer who was the first to receive a bitcoin transaction and publicly expressed his view that bitcoin should be turned into a settlement system.
“Yes, fairly well for many years online, collaborated with him on a number of things and discussions.” – Back replied. “Did that include digital currencies?” – we asked. “Yes, I talked with him about a number of systems over time, including b-money and rpow and probably digicash and variants also, definitely hashcash.”
We stated that we mean no judgment, but we thought it uncontroversial to say they failed. “Why do you think they failed?” – we asked.
“They were not decentralised. Digicash failed because it was centralised. The others were unimplemented ideas (b-money, bit-gold) or demos (rpow) or not directly ecash (hashcash).”
We asked why they were not decentralized – whether it never crossed their mind or for some other reason.
“Because architecturally it was not obvious how to make them so. B-money and bit-gold proposed to be to varying degrees decentralised but were not implemented and a bit fuzzy in places how it would work. Hashcash is very decentralised but not directly respendable. So, the concept that it would be good if they were decentralised was there, just not a good concrete mechanism to achieve it.”
Gregory Maxwell and Peter Todd have previously publicly stated they worked with Hal Finney. We asked: “Hal Finney expressed his vision, back in 2010, of a settlement network that now his apparent students, Gregory Maxwell and Peter Todd, are seemingly strongly pushing, was this established doctrine after all your attempts? The concept comes up time and time again instantly after Nakamoto’s [announcement of bitcoin’s whitepaper in 2008].”
Back replied: “I am not sure I would call bitcoin’s use alongside lightning a settlement network. Retail and micropayments have stronger demand for instant finality which lightning is better placed to provide. Bitcoin can not easily do that on chain.
Retail and micropayments going to lightning creates spare capacity on chain and reduces fees. In the mean time, on chain scale is improved. Seems like a reasonable set of onchain and layer 2 steps. More ideas to improve scale further are in the pipeline.”
Returning to the question, we said “the layer two concept was there before Nakamoto’s announcement, were you close to how an internet currency could or should work?”
“It seems like Szabo and Hal maybe had that view. I was more going for single layer, I more work bottom up.” – Back replied. After we asked for further elaboration, Back stated:
“I like to make things work first and optimise from there vs thinking too speculative architectural directions up front. It’s a systems programmer view of architectural strategy. I think bottom up works better, leads to tighter results. Top down is kind of fluffy. You can get to things where the bottom is not even possible via top down thinking.”
“That’s interesting because top down Finney took up Nakamoto’s idea, while you bottom up, didn’t, or were you busy at the time with your own things?” – we asked.
“I am more of an ideas or implement person, Hal was more of hands on lets install/try. Once I read how it works, that was enough. I often don’t run software if I know how it works.” – Back stated.
“Did you know Gregory Maxwell or Peter Todd before 2013-14?” – we asked.
“I interacted with Peter Todd way back in 2001” – Back said, linking to a Reddit comment by Todd. “And Greg?” “Not sure, I might have. I didn’t recall the Peter Todd discussions until someone found the email threads.” – Back replied.
“How did Blockstream come to be?” – we asked. “Oh, well, I met the bitcoin developers on IRC and got to know them that way.” “Why didn’t you include Peter Todd?” – we asked. “Can’t hire all the developers. We were concerned about there being multiple independent developers. Similar Jeff Garzik as he was at BitPay. I don’t think we hired any developers who were already employed doing bitcoin development as a job.” – Back said.
“You know, Garzik’s commit rights were removed, with no discussion, how you feel about that?” – we asked. “He wasn’t using them, I think. You can and most do contribute without being committers, the ratio is like 20:1 or higher of contributors to committers.” – Back said.
We pressed by asking “how you feel about it.” To which Back replied: “Yeah, I am staying out of that one. Not much to do with me.”
We turned the conversation back towards trying to understand the origins of the settlement system. We asked, “when you opened Nakamoto’s email, what did you think?”
“I think I wrote about my impressions on the crypto list. Weak fungibility and also why didn’t he use sander and ta shma’s “auditable anonymous electronic cash”? Also, security model is weak relative to previous ecash systems. However, it’s the price to get a decentralised system so then I took an interest in improving privacy and fungibility. Hence confidential transactions and other ideas.
(and I recognised the value of a decentralised system because that was the underlying interest for b-money and bit-gold… many of those interested in ecash having seen digicash fail from being centralised)”
We asked: “Do you know when this idea of second layers was decided and why? It seems, looking back, well entrenched by the point Nakamoto announced his paper, like given knowledge, like an a-priori.”
“I think Wei used it in b-money or in some discussion of b-money. It can be an artefact of the base system being uncertain or potentially clunky/inefficient and then people imagining things from it.
I wasn’t actually a fan of what Hal said about 2 layer in 2010. He was talking about bitcoin banks. You don’t want bank money you want real ecash that only you have the keys for in my opinion. So, lightning is more interesting because each LN transaction is a real bitcoin transaction.”
We said that it “feels like an old generation is trying to impose their ideas.” To which Back said:
“No, I don’t think so. Views vary on layer2. I think most would say scale them both and let the market decide. It’s largely different people working on layer2 anyway. So, it’s not like there’s anyone that’s stopped working on chain scale to focus on layer2 instead.”
It’s a view shared by many as most appear to prefer a solution that involves both on-chain scaling and higher scalability through the Lightning Network for micro-transactions. In fact, this specific point appears to be the main contention between the two sides.
Bitcoin Unlimited developers have put forward a proposal to implement segwit as a hardfork. If the client does attract the majority of miners and the wider ecosystem, it appears probable that the Lightning Network would also be deployed. Giving us both on-chain scalability and layer2 scalability.
On the other hand, a number of Bitcoin Core developers, including Blockstream employees and/or contractors, have stated that the 1MB limit is already too high or that full blocks are preferable. Giving us layer two scalability, but no on-chain scalability that operates above demand as long as technological capabilities allow it.
The majority of miners who have decided appear to be preferring a scaling of both. It is interesting that it seems Adam Back does too. However, as stated in the previous article, he appears to lack any actual say, thus, somewhat understandably, seems limited to backing the views of his employees despite of what he personally may think.