--- Day changed Thu Feb 06 2020 00:03 -!- nick_freeman [~nick_free@92.116.142.143] has joined #utreexo 00:06 -!- nick_freeman [~nick_free@92.116.142.143] has quit [Remote host closed the connection] 00:10 -!- nick_freeman [~nick_free@2001:16b8:307e:c600:dd40:69a2:a5e6:733a] has joined #utreexo 00:12 -!- nick_fre_ [~nick_free@2001:16b8:307e:c600:dc78:d903:4115:d27c] has joined #utreexo 00:15 -!- nick_freeman [~nick_free@2001:16b8:307e:c600:dd40:69a2:a5e6:733a] has quit [Ping timeout: 260 seconds] 00:17 -!- nick_fre_ [~nick_free@2001:16b8:307e:c600:dc78:d903:4115:d27c] has quit [Remote host closed the connection] 00:17 -!- nick_freeman [~nick_free@2001:16b8:307e:c600:dc78:d903:4115:d27c] has joined #utreexo 00:24 -!- nick_freeman [~nick_free@2001:16b8:307e:c600:dc78:d903:4115:d27c] has quit [Remote host closed the connection] 00:24 -!- nick_freeman [~nick_free@2001:16b8:307e:c600:dc78:d903:4115:d27c] has joined #utreexo 00:31 -!- nick_freeman [~nick_free@2001:16b8:307e:c600:dc78:d903:4115:d27c] has quit [Remote host closed the connection] 01:02 -!- nick_freeman [~nick_free@2001:16b8:307e:c600:dc78:d903:4115:d27c] has joined #utreexo 01:11 -!- nick_freeman [~nick_free@2001:16b8:307e:c600:dc78:d903:4115:d27c] has quit [Remote host closed the connection] 01:11 -!- nick_freeman [~nick_free@2001:16b8:307e:c600:dc78:d903:4115:d27c] has joined #utreexo 01:14 -!- nick_freeman [~nick_free@2001:16b8:307e:c600:dc78:d903:4115:d27c] has quit [Remote host closed the connection] 01:15 -!- nick_freeman [~nick_free@2001:16b8:307e:c600:dc78:d903:4115:d27c] has joined #utreexo 01:21 < kcalvinalvin> So tested it out. FullPollard isn't any faster than forest 01:21 -!- nick_freeman [~nick_free@2001:16b8:307e:c600:dc78:d903:4115:d27c] has quit [Remote host closed the connection] 01:21 < kcalvinalvin> ./simcmd genproofs -testnet=1 > log 69.47s user 27.26s system 199% cpu 48.365 total (FullPollard) 01:22 < kcalvinalvin> ./simcmd genproofs -testnet=1 63.42s user 27.76s system 199% cpu 45.647 total (Forest on ram) 01:22 -!- nick_freeman [~nick_free@2001:16b8:307e:c600:dc78:d903:4115:d27c] has joined #utreexo 01:22 -!- nick_freeman [~nick_free@2001:16b8:307e:c600:dc78:d903:4115:d27c] has quit [Remote host closed the connection] 01:22 -!- nick_freeman [~nick_free@2001:16b8:307e:c600:dc78:d903:4115:d27c] has joined #utreexo 01:22 < kcalvinalvin> I guess we need both forest and pollard 01:28 -!- nick_freeman [~nick_free@2001:16b8:307e:c600:dc78:d903:4115:d27c] has quit [Remote host closed the connection] 01:33 < kcalvinalvin> Oh, another goal for May before the Utreexo workshop should be to make godoc nice and pretty so people can just look at that 01:34 -!- nick_freeman [~nick_free@92.116.142.143] has joined #utreexo 02:53 -!- nick_freeman [~nick_free@92.116.142.143] has quit [Read error: Connection reset by peer] 07:05 < adiabat> hm... but not really any slower though. Like pretty much the same speed 08:16 -!- rafalcpp [~racalcppp@ip-178-211.ists.pl] has joined #utreexo 08:57 -!- rafalcpp [~racalcppp@ip-178-211.ists.pl] has quit [Ping timeout: 265 seconds] 09:09 < kcalvinalvin> adiabat Do you think worth persuing? 09:33 -!- kcalvinalvin [~kcalvinal@ec2-52-79-199-97.ap-northeast-2.compute.amazonaws.com] has quit [Ping timeout: 265 seconds] 09:34 -!- kcalvinalvin [~kcalvinal@ec2-52-79-199-97.ap-northeast-2.compute.amazonaws.com] has joined #utreexo 09:43 < adiabat> persuing as in... what exactly? 09:44 < adiabat> like it works, so... could use it 09:44 < adiabat> I guess the only difference would be deleting all the forest code. That's probably not great 09:44 < kcalvinalvin> I guess. But forest is easy to get it on disk 09:45 < adiabat> right, for on disk, forest is great 09:45 < adiabat> I'm a little surprised full pollard isn't faster for huge numbers of nodes with large swaps 09:45 < kcalvinalvin> btw, I decided to just rip off stuff from btcd and make a separate node. I like writing code more than reading code too much :) 09:46 < adiabat> because all pollard needs to do for a 12-high swap is change 4 pointers 09:46 < adiabat> but forest has to change, like, 8000 things 09:46 < adiabat> heh, cool... I think the function that would be cool to have 09:46 < adiabat> would be like : 09:46 < kcalvinalvin> I only tested it on blk0000{0..1}.dat files on testnet3 09:47 < kcalvinalvin> so I could run a longer test to see if full pollard is faster 09:47 < adiabat> func CheckBlock(utxos []txo, block *wire.MsgBlock) bool {} 09:47 < adiabat> well probably (bool, error) 09:47 < adiabat> or could just return false if there's any error 09:47 < adiabat> but probably want to show errors while getting the code to work 09:48 < adiabat> if we have an isolated function like that to do all the block checks, that would be easy to plug in to the current IBDsim 09:48 < adiabat> and then it's not really "sim" anymore, then it's legit IBD 09:50 < kcalvinalvin> could you clarify what you mean by "block check" 09:50 < kcalvinalvin> The signatures? 09:50 < adiabat> everything 09:50 < adiabat> like, blocck size 09:50 < adiabat> header is OK 09:50 < adiabat> proof of work 09:51 < adiabat> all signatures, everything 09:51 < adiabat> amounts aren't too high 09:51 < adiabat> fees add up in coinbase 09:51 < adiabat> basically every consensus rule bitcoin enforces SHOULD be decidable from those function arguments 09:53 < kcalvinalvin> ah ok 09:53 < kcalvinalvin> sounds fun 10:00 < adiabat> yeah... cfields_ did mention writing something like this in c++ from bitcoin core code 10:00 < adiabat> but if we get a btcd golang version first, can test with that 10:01 < adiabat> then it's a super easy drop-in replacement to call the c++ version 10:03 < kcalvinalvin> I looked at some c++ project that I did a while ago today 10:03 < kcalvinalvin> I remembered why I don't like c++ 10:03 < kcalvinalvin> heh 10:08 < adiabat> yeah I'm not really looking forward to porting everything to c++ 10:09 < adiabat> most of it should be not that bad though 12:12 < adiabat> .. oh wait it still has a printf in there! 12:12 < adiabat> line 97 of pollard.go 12:13 < adiabat> I'll take that out. maybe it actually is faster, and printfing a zillion times was slowing it down 12:14 < adiabat> yeah maybe try it again? It's was printing a line every utxo added. Which is kindof a lot. But maybe not enough to really slow it down that much 19:22 < kcalvinalvin> adiabat I commented on the commit that you did but yeah I did have the printf commented out already :) 19:22 < kcalvinalvin> It's even slower with the printf 22:40 -!- jb55 [~jb55@gateway/tor-sasl/jb55] has quit [Ping timeout: 240 seconds] 22:54 -!- jb55 [~jb55@gateway/tor-sasl/jb55] has joined #utreexo