--- Log opened Fri Feb 04 00:00:49 2022
00:49 < kcalvinalvin> Ok proof overhead sounds good
00:49 < kcalvinalvin> I'll do that
02:28 < kcalvinalvin> So I ran the numbers for the mainnet multi-block proof. Proving at 10 block intervals reduced the proof to 289G
02:28 < kcalvinalvin> This is vs ~380GB for 1 block intervals
02:29 < kcalvinalvin> But we're skipping proofs for only about 20% of the total dels
02:30 < kcalvinalvin> total that'd need to be proven is 1,594,139,231 and we're proving 1,303,511,228
02:31 < kcalvinalvin> Ideally we'd skip a lot more
02:31 -!- sloorush [~sloorush@2405:201:301b:9018:6b92:da2d:b3f:4135] has quit [Ping timeout: 240 seconds]
02:32 < kcalvinalvin> My SerializeSize() functions were wrong and I need to re-run to get the accurate numbers for how much each (leaf data, proof data, target data) take up.
02:33 < kcalvinalvin> With 10 block interval proofs, I think the multi-block proofs take up ~75% of the 289GB
02:34 < kcalvinalvin> I'll also re-calculate with the proof overhead algorithm to see if multi-block proofs are more efficient. Seems like with 10 block intervals, it is more efficient on mainnet
02:34 < kcalvinalvin> Not on testnet and signet though
04:31 -!- dunxen [~dunxen@gateway/tor-sasl/dunxen] has quit [Quit: Bye!]
05:30 -!- sloorush [~sloorush@2405:201:301b:9018:6b92:da2d:b3f:4135] has joined #utreexo
11:40 -!- dunxen [~dunxen@gateway/tor-sasl/dunxen] has joined #utreexo
13:13 -!- sloorush [~sloorush@2405:201:301b:9018:6b92:da2d:b3f:4135] has quit [Ping timeout: 240 seconds]
19:33 -!- dunxen [~dunxen@gateway/tor-sasl/dunxen] has quit [Remote host closed the connection]
22:36 -!- sloorush [~sloorush@2405:201:301b:9018:6b92:da2d:b3f:4135] has joined #utreexo
--- Log closed Sat Feb 05 00:00:50 2022