Transcript for #bitcoin-dev 2017/03/07

00:20 ruid what is the proper way to only remake a certain binary in bitcoin?
00:20 praxeology ruid? you mean when compiling w/ make?
00:22 ruid praxeology, yes, is the default to do a clobber, or an incremental?
00:23 praxeology um, when you run make, it only compiles for the architecture target you are doing AFAIK, and then it also only re-compiles the files you change or the files that depend on those changes
00:25 praxeology So right now when I change the file I am working on, it recompiles the .o for that file, then like 6 executables and a .a
00:25 ruid ok. That is not always the case. Make can imply clean > all depending on how the target is defined
00:26 ruid thank you
00:32 praxeology Where is the segwit witness data stored? Given that it is not part of a block... If I load a segwit block from disk, and I iterate and read through all of the transactions, does it give me the witness data too?
00:33 praxeology If I make a new "CCoins" with a segwit tx, does the witness data get stored in the "CCoins"?
00:37 praxeology or wait... the witness data is proof that the spending of an output is permitted... never mind I think I just answered my question
00:37 praxeology CCoins doesn't store witness data, never did, and won't after segwit either
00:39 praxeology I still wonder where/how the witness data is stored though. Guess I can look another day unless somebody knows off the top of their head
04:55 rusty OK, dumb q. In regtest mode, I want to test a zero-fee tx. Is there a way to avoid "insufficient priority" ?
04:57 achow101 try setting minrelaytxfee to 0 in your bitcoin.conf
04:58 rusty achow101: Error: Invalid amount for -minrelaytxfee=<amount>: '0'
05:01 achow101 huh. didn't know that that wasn't allowed
05:01 rusty Yeah, me neither. That's why I wondered if I was missing something...
05:02 praxeology zero fee tx are pretty much blocked now to prevent spam
05:02 achow101 but regtest...
05:07 rusty OK, I'll file a bug.
05:13 rusty Let me first upgrade from version v0.13.99.0-9346f84 :)
05:14 achow101 just remove the check for 0 and don't forget to put it back when you go to mainnet :)
05:15 rusty achow101: well, this is to test the lightning test vectors, so would be nice if everyone could run it without hacking bitcoind :(
10:33 bit7 can some kind soul please help me with a small section of C# Bitcoin Nbitcoin code ? :)
18:47 RainMan28 Attempting to send bitcoin using bitcoin core 0.13.2 to a valid (and previously used) address. Getting the error message that it is an invalid address.
19:43 coin_trader question on coin-UTXO selection .... example: i send bitcoins to wallet in these sizes: 1, 1, 1, 2, 2, 2, 10 -- if i then go to send 7, i should get a transaction that is just one or two inputs and 2 outputs. but i am instead seeing behavior where it wants to group together many of the smaller chunks which makes an inefficient & large byte size transaction.... i am seeing
19:43 coin_trader behavior where it will group 5-6 inputs when the node has viable and ready-to-spend outputs larger that would make same transaction with 1 input and 2 outputs or 2 inputs and 2 outputs...
19:43 coin_trader this did not happen in previous versions of core - i'm running 13.2 and have been noticing this type of behavior more and more...
21:05 praxeology Detailed description of my proposed Balances Commitment Data Structure: "". Feedback welcomed!
21:06 praxeology Balances <-> UTXO set
21:11 luke-jr * [new tag] v0.14.0.knots20170307 -> v0.14.0.knots20170307