Transcript for #bitcoin-dev 2017/07/30

02:45 Chicago Mad7Scientist, the most efficient reliable way to build for Windows is to use the Gitian instructions. https://github.com/bitcoin/bitcoin/blob/v0.14.2/doc/gitian-building.md ( from your question about 12 hours ago )
03:57 Mad7Scientist thanks
04:48 molz i've been using this guide: https://github.com/bitcoin/bitcoin/blob/v0.14.2/doc/build-windows.md
07:00 eck achow101: thanks <3
16:15 johntosin when is PoW change ready?
16:15 johntosin or you gonna keep that all private so nobody but blockstream sees it
16:26 johntosin did you know that right now BCC is the 4th most valuable coin by market cap in existence? and it doesnt even exist yet.
17:39 johntosin Bitcoin Cash has implemented the necessary protection against replay attacks, meaning TREZOR can support the currency safely.
17:42 arubi bullshit, it's only implemented for scriptpubkeys with a checksig
17:44 arubi lots of crap wrong with their bip143 ripoff. good for us :)
17:45 arubi funny how segwit is so bad, but no issue with using its sighash method
18:23 grubles https://twitter.com/slushcz/status/891599154941759488 <- looks like trezor will support it as BCH
18:24 arubi they will, but we're at the mercy of when exactly the HF block is mined
18:25 arubi if p2sh(segwit) transactions are live on the core network, these can be replayed as anyone-can-spend because of their redeemscript
18:25 arubi not sure if their code requires a hard fork block before segwit is live
18:26 arubi if it doesn't and they can ride the main chain until it's live and people, even trezor owners, use p2sh(segwit), then redeeming it on the abc chain is just an anyone-can-spend
18:27 arubi their code tests aren't very easy for me to follow, rules aren't well defined as in proper bips
18:31 arubi it /seems/ like a HF block has to be mined after some unix time passes, and then all transactions have to be signed with their fork byte in sighash. that makes transactions replay protected if the checksig is either in the bare script or the redeem script
18:34 arubi so if their rules say that the unix time has to be before segwit is active, then probably everybody should be protected. if they can change it to be after segwit is live, people could lose their coins on the HF chain
19:36 waxwing arubi: i have a vague memory of them requiring the first post-HF block being > 1MB? maybe
19:46 arubi yea that's right, but it's not important really. I'm not sure that the HF start time parameter has to be before segwit is active, and that's when the change to sighash is made
19:47 arubi so it's all okay if it happens before and all transactions paying to segwit scripts are protected, but if it happens after, then payments to segwit scripts will also exist in the abc chain and be spendable once the p2sh redeem script is revealed
19:48 arubi also bare scripts
19:49 arubi as soon as the unix time passes, replay protection should be active /and/ a HF block has to be mined where these transactions are allowed. it all rests on when that time happens
19:51 arubi their doc says "REQ-2 (configurable activation time)" and "The client shall allow a "activation time" to be configured by the user, with a default value of.." , and there also exists a bitcoind -uahfstarttime option
19:51 arubi so they can still "decide" to configure it for after segwit
19:57 arubi waxwing, we've seen to many bait-and-switch done by some mining pool ops, and abc folks were quoted to something like "the fork will activate if bip148 causes issues on aug 1st". it's all pretty hazy to me wrt what they're planning on doing
19:58 arubi these people being very vocal about segwit being insecure etc only adds to this
20:01 waxwing yeah
20:01 waxwing but we have quite a bit of time, no arubi ?
20:01 waxwing oh sorry didn't read backscroll
20:02 arubi right, segwit is still away. if they back out on august 1st, we can start being suspicious I guess
20:09 joakkinen Hi. Can someone tell me the url equivalent to 'https://www.bitstamp.net/api/v2/transactions/' in coinbase? I can't find it anywhere. Thanks.