1 2016-10-28 03:16:13 <achow101> can someone explain exactly how segwit makes sighashing O(n)? I can' wrap my mind around how it is any different from traditional sighashing
 2 2016-10-28 03:33:44 <instagibbs> achow101 look at SignatureHash in interpreter.cpp.
 3 2016-10-28 03:37:03 <achow101> instagibbs: I understand _how_ it is hashed. I don't understand _why_ it makes a difference.
 4 2016-10-28 07:38:46 <arubi> achow101, I think it's because in the old style, say tx with 1 input -> 1 output is 100kb, and a 2 input is 110kb, you'll be hashing the 1->1 one time (100kb) but the 2->1 twice which comes to 220kb and so on, and you'll be flailing around partial transactions as you rebuild each of the signed messages
 5 2016-10-28 07:39:03 <arubi> in segwit you're hashing all the inputs, outputs and sequences once, and end up with the skeleton of hashed stuff, and just replace the the outpoint for each input so as more inputs are used, the total size of the stuff you hash only grows the size of that input.  100kb->110kb->120kb, where 100kb is done once and used for each 10kb input
 6 2016-10-28 07:40:07 <arubi> if you have multisig inputs it's especially noticeable
 7 2016-10-28 07:45:30 <arubi> (numbers are made up here of course)
 8 2016-10-28 07:49:11 <arubi> I guess what I'm trying to say, old: each input is the whole message, each checksig is the whole message, so it grows a lot quickly.  new: each input is just its size, each checksig is just its outpoint :)
 9 2016-10-28 09:06:05 <ap4lmtree> hi
10 2016-10-28 09:06:17 <ap4lmtree> does bitcoin need help or more people writing or looking at bitcoin code?
11 2016-10-28 09:06:35 <ap4lmtree> like with core do you guys neeed people looking for bugs or do you need people to add new features or what?
12 2016-10-28 09:07:16 <arubi> ap4lmtree, there are 399 open issues in github :)
13 2016-10-28 09:14:57 <jonasschnelli> ap4lmtree: the bottleneck is mostly reviewing code
14 2016-10-28 09:15:28 <ap4lmtree> okay
15 2016-10-28 09:15:45 <jonasschnelli> ap4lmtree: if you are willing to help and if you have a decent skill in c++ or python (test framework) and if you are willing to learn how bitcoin-core works internally (or already know it), join #bitcoin-core-dev and help reviewing pull requests
16 2016-10-28 09:16:14 <ap4lmtree> thanks, ill take a look,  i am having more monitors now, so i could put the coding window in a monitor dedicated to it, and it will help me stay focus more on coding, i would like to get into coding more
17 2016-10-28 09:16:45 <ap4lmtree> i completed a c++ and java classes, but i dont really code, but i want to get into it, and bitcoin is the only thing that is smart to get involved in
18 2016-10-28 09:19:09 <ap4lmtree> im going to have five monitors , i had three and i recently bought two more
19 2016-10-28 09:23:13 <ap4lmtree> i am watching dark net right now, tv series,  and in japan ,  people have girlfriends who are nintendo ds  loveplus's rinko ,  and these people are sincere  and actually love them
20 2016-10-28 09:31:17 <arubi> does bitcoin scale to 5 x 4096 x 2160?  only ap4lmtree can tell
21 2016-10-28 11:36:27 <sturles> The balance of my main account was significantly reduced after upgrading to 0.13.1, and the sum of all account values is negative.  It should be the same as getbalance reports.  See http://0bin.net/paste/0FL35YiW078pNsF0#Hj2qAvE1Kr8keMBxgwxTjJqfp-1Qmn4tGhqMjVzUvGJ
22 2016-10-28 11:38:52 <sturles> I hope a downgrade will fix it..
23 2016-10-28 11:44:43 <sturles> I know that an account can have a negative amount, of course, but the sum of all accounts should always be equal to the total balance.
24 2016-10-28 12:03:11 <sturles> Restarted an older version with the same wallet, and it reports the correct account balances.
25 2016-10-28 12:13:21 <instagibbs> sturles please submit an issue
26 2016-10-28 12:13:53 <instagibbs> Balance stuff is a mess but i don't think it should have changed
27 2016-10-28 12:27:15 <sturles> This is very strange.  I just tested more or less the same wallet (old backup) on an armv7l CPU, and the balances sum up correcty there.
28 2016-10-28 12:30:56 <sturles> Is it possible to downgrade my wallet to libdb 4.8 without losing history?  I have tried various dump/load combinations in the past with no luck.
29 2016-10-28 12:31:30 <sturles> I should probably test this with the officially released Bitcoin Core, not my own compiled version from git.
30 2016-10-28 12:35:56 <luke-jr> instagibbs: accounts are deprecated though, it may have
31 2016-10-28 12:36:32 <instagibbs> I see.
32 2016-10-28 12:36:36 <luke-jr> sturles: I'd bisect it :p
33 2016-10-28 12:44:01 <sturles> luke-jr: sendmany still requires an account as first argument.  I use sendmany a lot.
34 2016-10-28 12:44:13 <sturles> Depercated or not.
35 2016-10-28 12:49:16 <sturles> luke-jr: I'll try it with bitcoinknots, and send the bug report to you. :-P
36 2016-10-28 12:53:05 <sturles> make check fails if you have both libdb 4.8 and 5.x installed, and configure with --with-incompatible-bdb.  test_bitcoin will link with libdb4.8.
37 2016-10-28 13:06:13 <sturles> Ah, it is a searh path problem.  My libdb_cxx-4.8.so is in /usr/lib while libdb_cxx.so is in /usr/lib/x86_64-linux-gnu/.
38 2016-10-28 13:27:31 <sturles> I'm recompiling with libdb_cxx-4.8.so completely out of the way.
39 2016-10-28 22:03:30 <musalbas> So is there any SPV client implementation of Bitcoin apart from bitcoinj? Genuinely can't find any other functioning SPV API
40 2016-10-28 22:09:26 <dgenr8> breadwallet-core
41 2016-10-28 22:19:34 <musalbas> looks nice, but with no documentation it is useless.
42 2016-10-28 22:31:20 <musalbas> looks like im going to have to write java :-(