Also I'd argue that if nodes aren't able to IBD (in a few days) or even sync after IBD behind Tor / limited bandwidth environments then decentralization is substantially hurt

Monero looks to be going in this direction.

Reply to this note

Please Login to reply.

Discussion

You're right about bandwidth that's the major concern. FCMP does make that worse, there will certainly be a time in the future when a scaling solution will need to be invented. I disagree time to download the blockchain will matter much its already multiple days, not much difference to add more days at that point IMO. Plus it's one and done.

I disagree that Tor nodes help much for decentralization. Mostly they're for client privacy. It makes it too easy to spin up thousands of unique looking onion addresses if your node is malicious. It used to be officially supported to open up the p2p port for onion addresses now it's just used for relaying transactions over hidden services https://docs.getmonero.org/running-node/monerod-tori2p/

The key to node decentralization is to make it difficult for malicious entities to spin up thousands. Ipv4 only and limited reselection of /24 subnets is huge for this https://monero.observer/monero-v0.18.4.3-fluorine-fermi-released/ this is much more effective than trying to limit block sizes to hopefully get more normal people running nodes.

Also mining centralization is the most dangerous centralization force for a L1 which we've obviously seen recently with the Qubic attack. P2pool adoption is pretty high at around 5% hopefully this increases. Node decentralization is inconsequential tot he attacks of mining centralization.