bitcoinThroughputAnalysis VS libbitcoin-system

I did an analysis a while back that concluded that we should be able to safely increase the block size to a degree where we can see a 5-10X increase in transactions per second in the next 15-20 years because of proposed scaling improvements and increased average machine resources.

I agree. But that's unrelated to the points I was making.

Hileman is completely wrong here. Pretty sad to see the head of research at blockchain.com get this one so wrong. The choice of consensus protocol does not affect transaction throughput or "capacity" ("capacity" as far as I can tell is not different from throughput in this context). What affects transaction throughput is the size of your transaction data and the limits placed on the blocks that hold that data. You can increase the size of your blocks to increase throughput, but as a result you inherently increase the computing resources required to participate in the network. The larger your blocks, the fewer people will be able to participate, which is one kind of centralization pressure a blockchain can create for itself. Fundamentally, distributed consensus is limited by the speed of light and physical size of the network. There are many ways to increase throughput of a blockchain like Bitcoin's, but they all boil down to reducing the data per transaction or making it more acceptable to increase the amount of data allowed in a block.

I wrote a paper a while back on bitcoin throughput. Basically, we need as many people to run public full nodes as possible. Probably around 5 million would be enough. However currently we have orders of magnitude less than that. So until we have enough, we need to keep it easy to run a full node - as easy as possible.

Some really good points in there. Bottle necks are basically whack a mole. I wrote a related paper on scalability in the context of bitcoin: https://github.com/fresheneesz/bitcoinThroughputAnalysis

If you want to discuss that, I'd be happy to. You might be interesting in reading this paper I wrote on the subject of blocksize and throughput: https://github.com/fresheneesz/bitcoinThroughputAnalysis

The number of public full nodes. Bitcoin still has only about 10,000 public nodes. An attacker would only have to spend $1 million per year to eat up all the public node bandwidth that is likely available. That's only $160k per month. If someone wanted to disrupt bitcoin, that's the way to do it.

Last release 2019, master branch marked as non-functional. Doesn't seem to be very active. Their homepage https://libbitcoin.info/ looks polished, but no signs of recent life. This vulnerability seems unmentioned. Well, they claim it's not a vulnerability but a demo implementation that should not be used to store anything of value. But there is no prominent note about that and some users did exactly that.

"fix block reward", my god why won't the stupidity ever end. block subsidy decreases censorship resistance. https://github.com/libbitcoin/libbitcoin-system/wiki/Censorship-Resistance-Property