--- Day changed Tue Dec 17 2019 01:10 -!- Netsplit *.net <-> *.split quits: instagibbs 01:11 -!- Netsplit over, joins: instagibbs 04:06 < kcalvinalvin> Heya. I push the new stuff on github 04:07 < kcalvinalvin> But... it's around 20~25 percent slower... 04:07 < kcalvinalvin> GetRawBlockFromFile() is sloooooooow 05:51 < adiabat> kcalvinalvin: cool will check it out 05:52 < adiabat> also utreexo call in 10 min at meet.jit.si/utreexo 06:01 < adiabat> (on jitsi now) 06:58 < adiabat> paste of tree example: 06:58 < adiabat> 14:b67d 06:58 < adiabat> |-------------------------------\ 06:59 < adiabat> 12:1276 13:728b 06:59 < adiabat> |---------------\ |---------------\ 06:59 < adiabat> 08:ca37 09:e7bf 10:6acc 11:d3a0 06:59 < adiabat> |-------\ |-------\ |-------\ |-------\ 06:59 < adiabat> 01:0600 04:0200 05:0700 06:0800 07:0900 07:01 < adiabat> better example 07:02 < adiabat> better example: 07:02 < adiabat> 14:b67d 07:02 < adiabat> |-------------------------------\ 07:02 < adiabat> 12:1276 13:728b 07:02 < adiabat> |---------------\ |---------------\ 07:02 < adiabat> 08:ca37 09:e7bf 10:6acc 07:02 < adiabat> |-------\ |-------\ |-------\ |-------\ 07:02 < adiabat> 00:0200 01:0600 06:0800 07:0900 08:04 < adiabat> cpp version: https://github.com/theuni/libutreexo 08:04 < adiabat> (theuni == cfields_) 08:21 < instagibbs> is ibdsim just running through "IBD" of the proofs against block data, in other words just testing it utxos exist for claimed inputs 11:16 < adiabat> instagibbs: yes; right now it doesn't do any signature / script checking 11:18 < adiabat> the idea (which isn't implemented yet but in works) is to have the utreexo library: 11:18 < adiabat> 1) return a bool about if the block is OK proof-wise 11:18 < adiabat> 2) return a serialized utxo for all inputs in the block 11:19 < adiabat> 3) wait while script-checking / consensus happens and then only modify if that returns OK and block is accepted 11:20 < adiabat> IBD sim only does 1 so far 11:21 < instagibbs> makes the most sense was just checking :) 21:49 < kcalvinalvin> huh ok I'll check out the cpp lib 23:55 < kcalvinalvin> I've merged blockparser and genproofs on the PR 23:55 < kcalvinalvin> Performance is around the same