Oddbean

1. I haven't had a sync time take a month even on a small raspberri pi. It's about a week, maybe a little bit more. Are you sure these people aren't using old HDD drives instead of an SSD? Or maybe the internet connection is a limiting factor here? Those make a huge difference 2. You are correct that a larger UTXO set makes it take quite a bit longer because of the I/O from disk, but this would be no different if it was literal adoption growth or some BRC20 garbage being shoved into the chain. Which means the only solution is literally to drastically lower the blocksize, or to alter the UTXO model entirely because of this inescapable reality of the system. 3. The baremultisig problem can be gotten around at the node level, by being instantly disregarded after creation. The only UTXOs that contribute to this problem are ones that get spent again, and therefore are involved in future transactions. 4. I completely agree we should be doing everything possible to make it easier to run a node, but the big problem with falling/static node counts isn't the difficulty of running one, its the fact that no one is incentivized to and everyone coming in isn't thinking about network stewardship, they just want their app to work. 5. High fees and ignoring stupid projects that do nothing but shit in the public park that is #Bitcoin are still the only real solution to this problem, imo. We cannot control how people use Bitcoin, and we cannot change the UTXO model (assuming there is even something that could "fix" this issue), which means we need to think about and implement ways to compress the validation of the UTXO set. This was why I mentioned UTREEXO. I also saw you mentioned assumevalid as well. I suspect this can be largely mitigated with a database mechanism rather than any sort of block control or consensus alteration. Would be nice to see people working on this problem more seriously, but I don't think it's our #1 concern. In fact, I think if we could use UTREEXO to get the UTXO set from other nodes immediately and start working with a full node within minutes, while the full validation happens in the background, I think this isn't a drastic concern. (also we get the lucky benefit that Ai is currently causing an *aggressive* funding and focus on an increase in the size of, and lowering of cost for RAM, so at least that is working in our favor) A legit concern ≠ existential crisis