< adam_> yeah same
< adam_> lol
< fanquake> adam_ asdfasdf Yes, this channel is used for discussion specific to core development, as well as dev meetings. See https://botbot.me/freenode/bitcoin-core-dev/ for logs.
< fanquake> Could someone send some testnet coins to 2MxhnRfjqjps8CSsUgp1US51q555Js5mUMr . Cheers.
< BlueMatt> fanquake: done
< fanquake> BlueMatt Thanks
< fanquake> It's an upstream issue with Qt, and won't affect running core. Are you running windows?
< BGL> yes
< mryandao> quick question, where can I find code on how AES encryption is applied to the encrypted private keys in wallet.dat?
< mryandao> alternatively, a grep regex will be sufficient ;)
< phantomcircuit> mryandao, maybe just in the code
< mryandao> yeah, I mean in the core source.
< phantomcircuit> mryandao, src/wallet/crypter.cpp
< mryandao> phantomcircuit: wonderful, thanks.
< echeveria> has there been any discussion recently about improving the methods for low-latency GBT updates?
< echeveria> the pool software that's publicly available at the moment seems to use a variety of methods for detecting when there's a new block, -blocknotify calling a script that sends an interrupt to send an RPC call, polling the RPC in a tight loop, or even attempting to build an empty block template from the ZMQ socket raw block output. the RPC interface has a long polling option but that's pretty hacky as well.
< echeveria> most ideally there'd be a mode for the ZMQ socket which emits a GetBlockTemplate response on UpdateTip, and at some configurable interval. eliminates any polling misses, any latency from spawming a new shell just to send an interrupt.
< bitcoin-git> [bitcoin] fanquake opened pull request #11936: [build] Warn that only libconsensus can be built without Boost (master...libcon-boost) https://github.com/bitcoin/bitcoin/pull/11936
< bitcoin-git> [bitcoin] fanquake closed pull request #11806: [Build]: Modify error if boost lib is not found (master...boost) https://github.com/bitcoin/bitcoin/pull/11806
< bitcoin-git> [bitcoin] fanquake closed pull request #11920: Trivial: Spelling fix, python → Python (master...patch-1) https://github.com/bitcoin/bitcoin/pull/11920
< echeveria> to exemplify that, CreateNewBlock() takes 26ms, sending it down RPC takes 90ms.
< jonasschnelli> echeveria: maybe have a look at https://github.com/bitcoin/bitcoin/pull/9919?
< jonasschnelli> Why is the longpolling hacky?
< echeveria> jonasschnelli: long polling is really difficult to deal with in a reliable way. there's software out there which calls GBT every 100ms, holding cs_main for who knows what portion of the time; as preferable to using long polling.
< promag> echeveria: jonasschnelli: both zmq notification and long polling have a problem: missed data
< promag> if you need reliability IMO you have to poll
< echeveria> promag: if you're losing ZMQ frames with a unix domain socket you've got other issues. there's sequence numbers in the messages that Bitcoin Core emits too, so you should be able to detect if you're getting messages lost.
< promag> I know that
< promag> just saying that generally ZMQ is not reliable, you have to work around the reliability
< echeveria> at any rate, the main uncomfortable feature is that JSON is still the most logical output.
< bitcoin-git> [bitcoin] hsjoberg opened pull request #11937: GUI: Setting for deciding address type (legacy, p2sh or bech32) (master...segwitwalletguioption) https://github.com/bitcoin/bitcoin/pull/11937
< bigboytimmyc> hey guys got a quick code question
< bigboytimmyc> if anyones around
< provoostenator> bigboytimmyc: this place is like New York City. No need to see hi, just ask :-)
< bitcoin-git> [bitcoin] Gerjan77 opened pull request #11939: Update build-windows.md (master...patch-1) https://github.com/bitcoin/bitcoin/pull/11939
< bitcoin-git> [bitcoin] Gerjan77 closed pull request #11939: Update build-windows.md (master...patch-1) https://github.com/bitcoin/bitcoin/pull/11939
< nedn> I am trying to build bitcoin on Mac, stuck at running "autogen.sh" step. Anyone knows what may be missing here? https://pastebin.com/4ndyXPz7
< sipa> looks like you don't have libtool
< nedn> thanks! "brew link libtool" fixed it for me
< jonasschnelli> In SignCompact (message signing), we do: vchSig[0] = 27 + rec + (fCompressed ? 4 : 0); can someone explain that? There is no specs?
< jonasschnelli> Is that just appending the recovery id with a flag if its compressed or not in 4 bits?
< sipa> yes
< sipa> the + 27 was just a random number to prevent it being confusable with public keys
< sipa> the +4 was added later for compressed pubkeys
< sipa> we'll probably need to extend it further for P2WPKH and P2SH-P2WPKH
< phantomcircuit> heyyy
< phantomcircuit> 500k blocks
< phantomcircuit> i feel this should involve cake
< andytoshi> ;;bc,blocks
< gribble> {"blockcount":500008}
< sipa> phantomcircuit: it's also the 4y anniversary of the term 'hodl'
< contrapumpkin> screw the price, that's the real milestone
< jonasschnelli> thanks sipa
< cluelessperson> phantomcircuit: but you already have lots of cake on your face.
< phantomcircuit> heh
< * jtimon> notices block height 500000 is a nice round number
< jtimon> for people who care about decimal anyway
< phantomcircuit> jtimon, i suggested cake
< jtimon> phantomcircuit: cake is poison, but I'm afraid my beer replacement isn't more healthy
< phantomcircuit> jtimon, delicious poison
< meshcollider> Happy half million \o/
< meshcollider> ryanofsky: thanks for all the reviews today!
< blaster> I wonder which block size limit we'd be at by 500k if satoshi's idea got added as per https://bitcointalk.org/index.php?topic=1347.msg15366#msg15366
< ryanofsky> meshcollider, no problem. nice easy reviews