Maybe We Can All Get Along After All – Even in Bitcoin

Increasing blocksize so large that it can no longer be independently verified by nodes or run by nodes behind tor is not a viable solution. Decentralization must be protected at all costs.

This is one of the constraints, I agree. My question for you as a fellow node operator is where you feel the limit is for you in terms of running a node? Storage is cheap. I wouldn’t be stressed by the additional load at all, but I may not be representative.

BU does try to establish a mechanism for creating this sort of balance through the protocol, but its by no means certain to strike the right balance. How would you propose to balance it instead, other than relying on One Core to Rule them All to dictate changes when it sees fit? That is just a different and potentially more dangerous centralization, isn’t it?

The real concern, however, is that by limiting capacity on the main chain and pushing transactions to side chains you are drastically increasing the risk of centralization. In most markets, players consolidate through mergers & acquisitions until there are only a few left. Two or three U.S. automakers; 2 or 3 U.S. stock exchanges, etc. In the end, there will only be two or three major side chains if ordinary market principles apply.

Core is only a partially decentralized team and certainly not diverse in terms of approaches, as the politics of the matter have shown.

The more significant problem is that having one dominant client is a monoculture that presents a real risk to the network. I agree with you that bcoin and Parity are more responsible implementations and perhaps more technically sound. The bcoin Extended Blocks implementation is particularly intriguing.

“It’s not about space it’s about bandwidth.”

Yes, quite right. I’m fairly unusual being in a university town with fiber-to-the-house. I take it the challenge is upload speeds on asymmetrical lines? Or is it data caps and costing in some areas?

“Regarding side chains what you are saying is pretty much like saying there will only be a few websites on the internet.”

No, not really. “Anyone will be able to create one” is not a guarantee of decentralization, because the issue is access and convenience. Anyone can create an online store, but that hasn’t stopped the online shopping market from being dominated by Amazon and Alibaba. That’s before lots of mergers & acquisitions, which are likely to further consolidate the space. And we all know that regulation is eventually coming to cryptocurrencies, which will increase barriers to entry for new players.

Segwit would appear to be a technically sound albeit somewhat complex solution that hasn’t really thought through the broader changes it would create in the ecosystem. By contrast, simply dealing with the hard-coded blocksize parameter algorithmically poses fewer fundamental risks to the great bitcoin experiment.

That’s not to say that Segwit or Lightning shouldn’t be implemented as optional improvements, but they should be implemented only alongside a permanent fix to the fundamental blocksize problem. That seems to have been the position of many miners, despite the fact that they are raking in huge additional fees from the currently constrained blocksize.

We’re getting into a bit of a Gish Gallop here.

I think what we have to face is that there really isn’t consensus of the bitcoin community on Segwit activation (or BU). If consensus is what you value, then the proper course forward is to engage productively with the other parts of the community and move forward to a new solution.

I’m afraid we’ll just have to disagree about the security and survivability of monocultures. From my perspective, you would not want 90+% of the internet to be run on Cisco routers, because of the risk of an exploitable flaw in Cisco routers.

In bitcoin terms, we should not want more than 50% of the network running on Core nodes for reasons similar to why we don’t want 50% of the hash rate associated with one mining pool. Yes, that may be the choice of individual miners to be part of that pool. Yes, the pool may be composed of geographically diverse miners. Yes, the pool may be acting like a good citizen and supporting the network impartially even with 90% of the hash rate. BUT it enables the possibility of attacks, and it risks incentivizing bad behavior or the feeling that the pool IS bitcoin and what it’s doing is right and everyone should just agree.

No different with Core.

Well, OK, let’s say that it’s clear there isn’t consensus now for either Segwit or BU, and the tenor of the “conversation” is such that consensus is unlikely to emerge.

I appreciate the thoughtful conversation, but I think you lose the thread of your own argument in the last paragraph. Miners are just individuals who choose voluntarily to run mining equipment, no different than developers are individuals who choose to contribute code. Miners contribute their hashpower to a pool of their choice. Developers up until very recently could contribute only to Core, and their work can only be included if approved by a tiny group with commit permissions. Now there are a small handful of different clients (Core, bcoin, parity), but Core still has 90% of the node count. That’s like one tightly controlled mining pool having 90% of the hash rate.

Monopolies can do real damage to the network (or any society). Too large a mining pool is hazardous, though I’d worry more about F2Pool on LiteCoin than anything on bitcoin. Antminers have too large a share of the mining hardware, I agree. There should be more competition, and it does make the network vulnerable to bugs like Antbleed. Core has WAY too large a share of nodes, which makes the entire network vulnerable to zero-day bugs in Core, and contributes to some of the antisocial behaviors when anyone dares to offer alternate viewpoints (and, conversely, generates resistance to even sound technical solutions because people feel they’re being dictated to by the monopoly).

The implication that you’re missing is that consolidation in mining hardware is an inevitable feature of markets. People buy equipment or use services by finding the best equipment and services they can. They don’t buy worse gear because of abstract commitment to decentralization. The popular products & services generate more income and in turn improve their products faster than competitors.

Exactly the same pressures will lead to consolidation in side chains under Segwit/Lightning. Without concurrent on-chain scaling, it’s a guaranteed way to privatize bitcoin.

I agree that Bitmain is producing the fastest miners at the moment, and since I’ve done ASIC development I’m pretty aware of the barriers to entry. I agree that there should be more competition in the space. Keep in mind, folks could probably just buy the ASIC chips from the fab house and build their own boards and miners. I wish someone would!

If this sort of thing worries you, though, then you should be just as concerned about Segwit/Lightning. What started as lots of independent miners ended up gradually consolidating. The same thing will happen to side chains. There will be a lot, initially. It will seem great. Then over time it will gradually consolidate.

When I was talking about different clients, I was really referring to clients on different codebases with different development teams. Bcoin (javascript) and Parity (Rust) match that criteria. That’s what helps protect the network.

If you really feel that having a few hundred people working a project guarantees “anti-fragile” code, then I’m afraid we’ll just have to disagree. Microsoft has way more developers on its projects worldwide, and yet the codebase is often a mess. Open and ad-hoc doesn’t always/often make for tight code. In fact, it’s frequently the case that small teams or even single coders (Satoshi) make for more coherent and secure development.

The risk, as you point out, is that for debugging you need broader review. BU closing the source in response to bad actors was understandable in the short-term toxic environment where people would rather attack than contribute, but it was a very foolish and unhealthy move from a long-term perspective.

At the same time, a smallish community of a couple of hundred are likely to get into group-think over time working together on a project. True “peer review” requires genuine outsiders, not people on a diversified team.

I always get a little bit concerned when an individual or group starts claiming to be “the backbone” of any particular community.

I think the point being made is that replacing one hard-coded parameter at 1MB with another hard-coded parameter at 2MB is not a scaling solution. And it’s not. This should not be a hard-coded parameter. Seamless scaling should be imbedded in the protocol, so that moving transactions to side chains is an option rather than a requirement, and so that the network doesn’t have to rely on (potentially private/proprietary) side-chains.

Actually, we have no data at all about whether increasing block size will increase or decrease “centralization” (by which I assume you mean node count for Core nodes?). It’s never been done, so all we have is speculation. Side-chains enabled by Segwit/Lightning which represent the majority of the possible capacity increase would be private/proprietary.

Rather than spin for political reasons it’s good to think through the real arguments being made by different people. I get your “why not just do Segwit first and see what happens?” argument. It does make sense. I just think it fails to address the concerns about broader impacts on the bitcoin ecosystem expressed by many members of the community.

We can either try to address those concerns, which is admittedly annoying extra work, or we can fort up and start lobbing turd-bombs at each other and try to shout people down with Trumpian pejoratives. I’d humbly suggest that is also annoying, it’s also extra work, and it has the disadvantage of getting nowhere.

What rudeness, has the audacity to slander vortex without even backing up his claim with any other logic than “that’s not a scaling solution!” How about you back up your words with logic and fact and keep your small minded bashing comments to yourself sir.