Transcript for #bitcoin-dev 2017/05/05

02:02 btcdrak This looks like a great replacement for my addrindex fork https://github.com/dcousens/indexd
15:02 dgenr8 luke-jr: bips test is failing on absence of optional Comments-Summary header
15:10 luke-jr dgenr8: because it's absent, duh
15:10 luke-jr it's not optional
15:11 luke-jr hmm, or is it supposed to be?
15:12 luke-jr that's annoyign
15:12 dgenr8 * Comments-Summary: <summary tone>
15:12 dgenr8 just following the rules
15:13 luke-jr nitpicking* :P
15:14 dgenr8 it will always be empty at the start, that might be why?
15:15 luke-jr well, BIP 2 does specify "Comments-Summary: No comments yet." as the initial state also
15:15 luke-jr anyhow, fixed the script, now it fails on README.mediawiki: https://travis-ci.org/bitcoin/bips/builds/229089899?utm_source=github_status&utm_medium=notification
15:22 luke-jr dgenr8: "If the block height is encoded at the start of the coinbase scriptSig, as per BIP34," <-- why the "if"?
15:26 luke-jr dgenr8: "rounded down to the nearest byte" is self-contradicting
15:28 dgenr8 implementation efficiency. bip34 doesn't have a defined starting height and we need to extract the vote before the bip34 check
15:30 dgenr8 1.05^4 is not an integer
15:30 dgenr8 1.05^4 * 1000000
15:44 luke-jr dgenr8: but "rounded down" means truncating any fractional part, and "nearest" means rounding up when the fraction is >=.5
16:24 dgenr8 Since we say that raise limit is denominated in bytes, I agree "rounded down" alone is sufficient. Ditto for lower limit. I'll add to a fixups commit in process.
17:19 luke-jr dgenr8: I don't post those questions, to be answered directly (where nobody will see the answer later); they should be clarified in the BIP so others have the answer right there
17:25 dgenr8 ok, they will make it better, sure, mostly. i don't plan to add anything about lowering block size below 1MB though
17:41 dgenr8 I don't know where you get 1164% annual growth. i get 1.05^26 - 1 = 256%
17:44 dgenr8 oh, it looks like you used 52 periods. nope only about 26 :P
18:08 luke-jr oh, right
18:08 luke-jr still, this stuff should be documented at least
20:31 ProfMac any guess on the grand total of share submissions per minute to the mining pools?
21:31 CryptAxe Can anyone explain the comments about fee sniping in CWallet::CreateTransaction to me? I understand that setting the locktime discourages fee sniping by forcing the blockchain to move forward in order to include the tx. I don't understand the second comment though "// Secondly occasionally randomly pick a nLockTime even further back..." Doesn't this part cause transactions to occasionally encourage reorgs / fee sniping? And how does it help with priv