Transcript for #bitcoin-dev 2017/08/09

12:05 Chicago dakk, Have you looked at any of the blocks on Testnet or on the blocks w/ SW on Litecoin's Main network? This links to one, https://twitter.com/SatoshiLite/status/862345830082138113 and this links to one from Bitcoin Test network, https://bitcointalk.org/index.php?topic=1758923.0
13:20 dermoth Hi... When I try to decode some of the largest transactions that made it into the chain using bitcoin-cli I get this error: _: line 12: bitcoin-cli: Argument list too long
13:26 dermoth this is os-level limitation I think, it there any other way to pass the tx, ex stdin?
13:50 dakk Chicago, thank you it's what I was looking for; I didn't known this explorer, the raw-hex feature is missing in almost all explorers
14:06 Chicago dakk, you're welcome. If you happen to want to parse the raw blk data manually, find this page http://james.lab6.com/2012/01/12/bitcoin-285-bytes-that-changed-the-world/ on archive.org. It goes a long way to making you read hex like Link, in The Matrix.
14:08 dermoth I worked around using rpc directly...
14:08 dermoth I think it'd be nice for bitcoin-cli to accept data from stdin
14:09 dermoth oh it does? :)
14:15 dakk Chicago, this how-to is awesome, thank you again
14:16 Chicago dakk, you're welcome - have fun. I recommend using a piece of paper with that Genesis block printed out so that you can take a ruler and underline the various sections -- and then label them for quick reference.
14:16 dakk nice trick
14:34 jouke Chicago, dakk: http://srv2.yogh.io/#block:last
14:42 Chicago jouke, well that makes it easy
14:42 Chicago so much for paper, pen and a ruler
14:53 Chris_Stewart_5 dakk: You are working on the OCaml implemenation right?
14:55 dakk Chris_Stewart_5, yep this is the library https://github.com/dakk/bitcoinml (and this is a full-node implementation wip: https://github.com/dakk/letchain/; just a toy atm)
14:56 Chris_Stewart_5 cool, I'm a OCaml fan glad to see it's coming along.
14:58 dakk Chris_Stewart_5, great, one potential user for my lib :) if you are an ocaml dev, I also wrote a secp256k1 wrapper: https://github.com/dakk/secp256k1-ml
16:50 Emcy_ What was the actual reasoning behind the 0.25bytes weight for segwits txns [giving a theoretical max hard cap of 4mb blocks]?
16:50 Emcy_ some people say its arbitrary, or has a reasoning other than technicals or something
19:02 ginseng anyone with experience using berkeley databases know how to explore the data in a wallet.dat? i have been using the bsddb python module to explore it as a btree, but i just find a single key "main" with value '\x00\x00\x00\x02'
19:04 arubi ginseng, I had success with the db-util package
19:13 arubi ginseng, fwiw my past try at reverse engineering it a bit, it might save you some time :) https://gist.github.com/fivepiece/74b6f8bfa8816ee7e283d25a36947629
19:14 arubi worth to mention that this is a regtest wallet
19:23 ginseng arubi: thanks for posting that. the comments you added are very helpful. i'm trying to replicate that with db_dump using -p to print out printable characters, it looks like you did more than this though to clean up the output
19:35 arubi np, it's helpful to have walletdb.cpp open when navigating the file. if you manage to figure out dumps with script stuff, transactions, watchonly.. I'd love to follow up
19:49 arubi oh crap, I forgot that this dump was from the patch where you can set your own derivation path
19:49 arubi I think that was it. if you read the logs ginseng, be advised
19:50 arubi I think that was it :\ so long ago
20:54 ptolemaus when I make some btc escrow transactions and take shares myself
20:54 ptolemaus should I create a new address for each share?
20:55 ptolemaus coz if I were to collect and transfer all these cockroach shares then I would pay like 50 ct for each coz of transaction fees
20:56 ptolemaus if I just put it on 1 address I could save a lot of dinero then
21:04 dermoth ptolemaus, no
21:05 ptolemaus then put it on multiple or really just one?
21:05 dermoth even if you have a single address, each output will end up a tx input
21:05 dermoth the transaction has to identify each previous transaction with unspent fund
21:05 ptolemaus oh right
21:06 ptolemaus inputs are utfo anyway
21:06 dermoth it doesn'T matter it it's the same address or different, the resulting TX will be the same size
21:06 dermoth what you may want is to use the key type that comes with the smallest txin size
21:07 dermoth (which I'd have to check which is...)
21:07 ptolemaus what might one be?
21:07 ptolemaus just one so I can search it and check what exists
21:09 dermoth I believe P2SH is what you want
21:10 dermoth when Segwit is activated, P2SH-P2WPKH will be even better (and clients doesn't have to support Segwit to sent to those addresses)
21:10 ptolemaus P2SH is used on inputs?
21:13 dermoth p2sh means you send to a script hash, then txin provides the script and its signature
21:13 dermoth https://en.bitcoin.it/wiki/Transaction#Input
21:18 dermoth oh... no bot spamming for every url here :)
21:33 ptolemaus dermoth, thank you