15 sats \ 2 replies \ @l0k18 OP 17 Jan 2023 \ parent \ on: Indranet's consensus bitcoin
I don't think average 10 minutes to start a node broadcasting is that long a delay... Keep in mind it's for relays not for clients.
Large locked UTXOs are not for anything other than further differentiation candidates for introduction to new clients sooner. The age also matters, as does the frequency of their shoutouts from clients.
https://medium.com/@SDWouters/erlay-an-awesome-performance-improvement-to-the-bitcoin-network-f59d757b2d14
Erlay is a scheme that improves latency by focusing on nodes nearby denser amounts of traffic? Best I can make out from a brief reading of that article.
In a way similar principles are at play in the rules I laid out above. age and frequency, these are two things that also define a "good customer" in retail, how long they keep showing up and how often they show up, and yes, how much they spend.
Yes, it's probably even a good way to explain the security model.
So erlay is something like full nodes sharing BIP 157 style filters instead of the transactions themselves, and only sending tx's that the peer doesn't match out of the filter set. I'm not sure how much this is actually more efficient or not...
I'm also not sure how this fits in to the idea of clients giving belated shout outs to a sprinkling of their nodes based on their reliability in order to shape the peer list to reduce bad nodes getting free advertising.
Anyhoo, the OP is just to note down what I figured out last night thinking about the problem of sybils trying to capture sessions and not deliver on them.
reply
Lmao sorry for wasting your time. Sounds like the people I asked didn't give very useful feedback
reply