00:23:19c0rw|away:c0rw|away is now known as c0rw1n
01:29:57Greed`:Greed` is now known as Greed
06:24:44fanquake:fanquake has left #bitcoin-wizards
06:34:42justanotheruser:"If every transaction includes an nLockTime that's the current/next block, its fees can't be claimed without fully catching up to the blockchain. This means that rewinding 10 blocks in order to get all their fees requires re-mining all 10 blocks."
06:36:14justanotheruser:I don't see how that's any different from the current situation. You can't include tx already in a block anyways.
09:05:16wilhelm.freenode.net:topic is: This channel is not about short-term Bitcoin development | http://bitcoin.ninja/ | This channel is logged. | For logs and more information, visit http://bitcoin.ninja
09:05:16wilhelm.freenode.net:Users on #bitcoin-wizards: andy-logbot irclouis justanotheruser AaronvanW NewLiberty todaystomorrow Guyver2 jaekwon Meeh hastagg_ eristisk MoALTz [7] koshii dgenr8 copumpkin heath gnusha kanzure fenn LarsLarsen Aquent ebfull Logicwax jaromil weex helo fluffypony Alanius Greed roconnor DougieBot5000 c0rw1n_ Dr-G moa samson_ kefkius tdlfbx Guest86169 mkarrer_ tromp_ Keefe K1773R Iriez shesek Eliel jrayhawk crescendo mortale waxwing Baz___ PRab lnovy tacotime epscy
09:05:16wilhelm.freenode.net:Users on #bitcoin-wizards: Starduster Adlai prepost Luke-Jr coutts Fistful_of_Coins BananaLotus CodeShark DoctorBTC HaltingState maaku Grishnakh iddo null_radix Emcy PaulCapestany Cory eric hollandais warptangent altoz gmaxwell comboy Flyer33 [\\\] nuke1989 bobke_ sneak napedia huseby GnarSith stonecoldpat Anduck johndoe01_ phedny iambernie MRL-Relay berndj phantomcircuit spinza kyletorpey zenojis mmozeiko pi07r btcdrak midnightmagic sl01 nsh wizkid057 Muis Guest1930
09:05:16wilhelm.freenode.net:Users on #bitcoin-wizards: arowser go1111111 sipa poggy NikolaiToryzin cfields coryfields Sangheili mappum jbenet kjj21__000 Taek EasyAt Hunger- optimator_ kumavis andytoshi BrainOverfl0w fds4345 gazab bbrittain BigBitz Apocalyptic emsid Starsoccer throughnothing warren gavinandresen dansmith_btc Nightwolf AdrianG mr_burdell zibbo_ tromp SomeoneWeird kgk firepacket Dyaheon myeagleflies wumpus pigeons nanotube asoltys gribble Krellan kinlo a5m0 superobserver artifexd
09:05:16wilhelm.freenode.net:Users on #bitcoin-wizards: [d__d] LaptopZZ HM_ espes__ CryptOprah Graet livegnik _2539 hguux petertodd smooth BlueMatt @gwillen michagogo yoleaux amiller btc_ danneu catcow TD-Linux [Tristan] ryan-c roasbeef pajarillo Gnosis ahmed_ so @ChanServ lechuga_ abc56889 harrow
10:16:04gmaxwell:If anyone else is at IETF this wek and wants to get togeather please feel free to ping.
12:44:21K1773R_:K1773R_ is now known as K1773R
14:09:44instagibbs:justanotheruser: yeah it doesn't stop the attacker from simply cloning the last 10 blocks, taking all fees. The proposal appears to be a variant of "pay txn fees forward" with some smoothing to discourage sniping.
14:11:20instagibbs:stopping rewinding could be done by txns including a hash of the most recently seen block, committing it to that branch. Downside seems to be a re-org means all transactions after the re-org get "orphaned"
14:18:21instagibbs:I kind of like that latter solution, as assuming dark hashpower is quite liquid, you'd at least only let that dark hashing power double spend their inputs instead of fee sniping
14:29:41nsh:* nsh wonders if instgibbs has used another nickname
14:29:48nsh:*instagibbs
15:58:08kefkius_:kefkius_ is now known as kefkius
16:08:32rdponticelli:rdponticelli has left #bitcoin-wizards
16:12:54OP_NULL:instagibbs: "dark hash power" seems like a rather ill defined concept. the way I’ve seen it described is that people have huge farms of obsolete equipment who don’t mine anymore due to it not being profitable, and can suddenly spin it up to attack when there’s an opportunity.
16:13:20OP_NULL:I take issue with that being particularly probable as it’s normally described. imagine I have my 1% farm sitting around, I see a block that’s got a high fee, I’ve now got to fire up 30 megawatt of hardware and cooling, which is just impossible on a short timeframe.
16:13:57OP_NULL:what seems more likely is that a large farm alters their behaviour from the network norm to being rational but immoral. i.e., they have no problem aggressively trying to double spend fees, ditching all free transactions and do little tricks like trying to orphan blocks that orphan theirs (I’m surprised big pools don’t do this now, actually, they should be).
16:14:14kanzure:http://diyhpl.us/~bryan/papers2/bitcoin/Two-factor%20authentication%20for%20the%20Bitcoin%20protocol.pdf
16:14:17kanzure:http://diyhpl.us/~bryan/papers2/bitcoin/Realizing%20two-factor%20authentication%20for%20the%20Bitcoin%20protocol.pdf
16:18:32OP_NULL:kanzure: somebody should tell the author that BIP32 exists, and that they don't need a RNG on the android device at all past the initial setup.
16:24:32Eliel:OP_NULL: a lot of miner hardware is also such that you can improve efficiency by lowering your hashrate. In such a case I'd expect it to be much smaller of a task to switch the farm to full blast with lower efficiency.
16:26:47OP_NULL:kanzure: that system is quite cute, but it has different properties than P2SH multisig. I'm not sure people care enough about transaction size to end up using it, most wallets haven't got around to using compressed keys and they've existed for 3 years.
16:29:25kanzure:huh, "First of all, it would be visible on the blockchain that multiple signatures were used"
16:29:37kanzure:for some reason that did not occur to include in my threat models
16:30:20kanzure:seems reasonable enough: it's totally irrelevant if everyone else knows whether or not multiple signatures have to be used to sign, so it's best not to offer that information voluntarily by default
16:31:39OP_NULL:what effect does exposing that actually have though?
16:33:47kanzure:at the moment there are less multisig p2sh transactions than other types, so at the very least correlation
16:35:06instagibbs:nsh: long-time lurker; I post infrequently.
16:35:23nsh:* nsh nods
16:36:18instagibbs:OP_NULL: I think of dark hashing power as any machine which isn't being used actively. Presumably these days it means it's inefficient compared to alternatives, but not necessarily that way in the future
16:38:22instagibbs:assuming we approach some $/hash hardware that anyone can get at any amount(100% liquid I suppose), that doesn't mean everyone is going to turn it on at same time
16:40:06nsh:as long as we put everybody on the same electrical grid, we can inherit some pretty cool security properties about maximum hash power delta
16:40:09OP_NULL:instagibbs: if I am a "dark" miner who sees a chance of profit, I'm going to want to be getting mining as quickly as possible. your window is going to be quite small.
16:40:49instagibbs:OP_NULL: I'm assuming 100% liquidity of miner power, like magic. Clearly not even close to true today
16:41:02instagibbs:but let's say we get down to some reasonably small fabrication size
16:41:09instagibbs:gets commoditized
16:41:36instagibbs:it'll be a very different story
16:42:44instagibbs:this is more long-term though, not relevant today
16:43:34OP_NULL:commodity mining hardware is fairly scary as a thing, actually. if a lot of people have almost no stake in the network they are mining on, there's a very real chance they'll just sell their hashpower to whoever the highest bidder is.
16:43:46instagibbs:exactly
16:44:31instagibbs:im not sure what that does to consensus; people could fire up ASICs, rewrite some history, get fees/whatever. In the presence of interested mining pools, I guess the mining pool would blacklist such a fork
16:44:47instagibbs:seems like a MAD situation
16:45:07instagibbs:either people don't do it, or consensus will break for a while
16:45:14OP_NULL:kanzure: good point. 1.5% of all BTC are in P2SH, a lot more than I expected.
16:45:29kanzure:OP_NULL: http://p2sh.info/transactions-per-block
16:46:04instagibbs:cool chart
16:46:47OP_NULL:it's interesting that you can see weekends in that graphic.
16:49:37instagibbs:last point on long-term mining(been thinking about his for a while): The only slam-dunk scenario where dark hashing isn't too dangerous is people mining at a loss as a given. Mining as -EV game.
17:05:41justanotheruser:kanzure: it would be nice if they labeled their colors
17:50:47c0rw1n_:c0rw1n_ is now known as c0rw1n