Pay-per-query-hop and Ripple Distributed Protocol scalability revisited
42 views
Skip to first unread message
Johan Nygren
Sep 5, 2022, 7:27:17 PM9/5/22
to Ripple Project
To reach any person by multi-hop, you kind of have to reach _every person_ each request (when taken to an extreme. )
Query requests can be decreased from that, to the 2*sqrt(every_person), if it is bidirectional.
They can be further decreased, when people are closer than "six degrees of separation", maybe 5 degrees, or 3 degrees.
To limit query request reach (to avoid them bouncing forever, or if people have a filter to not accept twice, once per person), a pay-per-query-hop mechanism can be used.
To query 2*sqrt(every_person), total cost could be 1 cent. The cost to query every_person would be sqrt(every_person)/2 cent. If every_person is 8 billion, 450 dollars.
The data passed via each person (assuming requests follow trust lines), also increases and decreases in the same way. If every_person is 8 billion, 45000x more data has to pass via each person. A difference equal to going from 25 kb to 2 gb.
Query requests can be decreased from that, to the 2*sqrt(every_person), if it is bidirectional.
They can be further decreased, when people are closer than "six degrees of separation", maybe 5 degrees, or 3 degrees.
To limit query request reach (to avoid them bouncing forever, or if people have a filter to not accept twice, once per person), a pay-per-query-hop mechanism can be used.
To query 2*sqrt(every_person), total cost could be 1 cent. The cost to query every_person would be sqrt(every_person)/2 cent. If every_person is 8 billion, 450 dollars.
The data passed via each person (assuming requests follow trust lines), also increases and decreases in the same way. If every_person is 8 billion, 45000x more data has to pass via each person. A difference equal to going from 25 kb to 2 gb.
Johan Nygren
Nov 30, 2022, 6:18:45 AM11/30/22
to Ripple Project
This idea is probably overkill. Using equivalent to "Time To Live" (TTL) in IP header in TCP/IP is probably enough. But, Ryan's Ripple somewhat uniquely seems like it could also relatively easily use a payment-based "anti spam" mechanism. But TTL probably easier short term and maybe better overall long term, spam in message routing networks isn't a new problem.
Reply all
Reply to author
Forward
0 new messages