Transcript for #bitcoin-dev 2017/07/18

14:11 SopaXorzTaker I have an idea to improve decentralization by preventing pooled mining.
14:12 SopaXorzTaker The idea is to add a signature field to the block structure, which would contribute to the Proof-of-Work hash
14:13 SopaXorzTaker The miner then would have to sign the block with a public key corresponding to the coinbase address before submitting it.
14:14 SopaXorzTaker This would make pooled mining infeasible, as every miner would have to have access to the pool's private key.
14:15 Emcy i think that would just make huge private farms supreme
14:15 SopaXorzTaker The signature field would contain a standard DER signature and would contribute to the header's hash, thus making PoW require signing the block
14:16 SopaXorzTaker This would also thwart ASICs to some degree, and the PoW function would become a bit computationally harder
14:50 lavenders what is the difference between segwit2.x and segwit bip9?
14:51 lavenders after bip91 activation is HFork on 1.8 possible?
15:28 goatpig BIP9 uses a 95% activation threshold, that's reduced to 80% in the process that is part of 2x, after signaling from another bit first.
15:28 goatpig not sure what 1.8 refers to
15:57 Murch goatpig: probably August 1st.
15:57 Murch lavenders: Segwit2x's hardfork is planned for 90 days after segwit activation.
15:58 Murch lavenders: The only faction planning a hard fork is the newly announced "Bitcoin Cash" championed by BitcoinABC.
15:59 Murch lavenders: I don't expect this to be of significant interest in the greater Bitcoin landscape. :p
16:01 goatpig oic, the . got me confused
16:02 Murch goatpig: I'm at an advantage, I'm familiar with the German address format (which in full would be 1.8.2017) ;)
16:02 goatpig hah, we use / in france!
16:18 notmike We should increase bitcoin's transaction capacity.
16:46 lavenders Murch, thank you
16:50 goatpig Murch: what about UASF forking?
16:50 Murch UASF is a soft fork, not a hard fork
16:51 goatpig if some portion of the hash rate does not signal SW and the BIP91 gang mines atop of them, UASF would fork
16:51 goatpig it will split since it only accepts signaling blocks
16:51 Murch yeah, it could introduce a chain split
16:51 Murch but that doesn't make it a hard fork
16:51 goatpig no it doesn't, but for all intents and purposes, people interchange HF and split
16:51 Murch It pains me, because they mean different things.
16:51 goatpig indeed
16:52 Murch You can lead a horse to water, …
16:52 Murch
16:54 notmike If UASF orphans valid blocks, that's a hard fork.
16:55 goatpig that sentence is too semantically blurry
16:55 notmike Ok, let me get my crayons real qwik
16:56 goatpig be my guest
16:59 notmike Everybody is on one bitcoin. On Aug 1st UASF will begin orphaning non-bit1 signaling nodes, presumably miners will cooperate. Some time later - perhaps September - miners (80%+ of hash) will hard fork to >2mb blocks. At that point, there will be two mutually exclusive bitcoins. However, prior to this (Aug 2st) UASF
17:00 notmike ... will have already split the chain by excluding BIP 91 blocks and anything other than BIP 148 conforming blocks.
17:01 goatpig UASF nodes do not reject BIP91 blocks per se
17:02 goatpig with the timeframe as outlined by 2x, there would be no more BIP91 signaling by August 1st, they'd already be pushing bit1
17:03 goatpig the only instance where UASF splits from the current chain is if the NYA miners do not orphan blocks without bit1 flagged
17:03 notmike So, UASF will accept segwit2x and anything else until they produce a non-bit1 signaling or >2mb block?
17:04 goatpig not exactly
17:04 goatpig if 2x mines a top a non bit1 flagged block
17:04 goatpig UASF splits
17:04 goatpig if they don't, the split is at 2x
17:04 goatpig ie the actual HF
17:04 notmike Right, this is all intentionally designed to produce a fork
17:04 goatpig a split actually
17:04 goatpig the fork is baked in to begin with
17:05 goatpig this would create a split before the HF would be more accurate
17:05 notmike Split and fork are awfully similar
17:05 goatpig not exactly, you can splits during orphan races
17:05 goatpig both sides of the split follow the same consensus
17:05 goatpig HF imposes a different consensus
17:06 notmike I understand that two blocks could be produced at the same time and that those could cause a split
17:07 goatpig as for the question is UASF effectively a HF under the current circumstances?
17:07 goatpig idk if it can be characterized as such
17:07 notmike Right, it's effectively a HF. But not technically because consensus layer is still compatible
17:07 goatpig well that distinction is kinda vague to me
17:08 goatpig UASF could wipe out the 2x chain if they get more hash rate pre 2x HF
17:08 notmike I was discussing this with gmaxwell in the mid-80s. It's a tough question.
17:08 goatpig but a HF without wipeout protection, like say BU, is susceptible to the same issue
17:08 notmike I don't see any way for UASF to do that. Opponents own 80% of hash!
17:08 goatpig hash rate follows the money
17:09 goatpig and a minute instance is not useful to a fundamental analysis
17:09 notmike That's true. Can LN and sidechains prove their value before miners set the house on fire?
17:10 goatpig well the idea of 2x under this schedule is precisely to prevent this
17:10 goatpig I would imagine it would take a year for the ecosystem to get really going with LN
17:10 goatpig for my part I wouldn't be able to deliver LN functionality in under 3 months starting August
17:10 notmike I get it. Politics is what I do. Bitcoin is particularly transparent. The miners are generating an artificial crises.
17:11 goatpig it's disruption but with an unclear endgame
17:11 notmike But they're also in control of most of the votes
17:11 goatpig as long as hash rate votes for forks, that is true
17:11 notmike User opinions don't matter in bitcoin
17:12 goatpig the network effect matters
17:12 notmike I mean, they matter, but I think you're picking up what I'm laying down here.
17:12 goatpig if valuation is an indicator, so far the market/users are happy with the status quo
17:12 notmike As long as the TXs get confirmed quickly, people will take what they get.
17:12 goatpig well that depends
17:13 goatpig people like me would never run 2x
17:13 goatpig if you lose a huge portion of your technical ecosystem over this kind of take over
17:13 goatpig you will be trouble in no time
17:13 notmike I don't mine bitcoin, that's crazy!
17:14 goatpig a non 2x split can survive just fine
17:14 goatpig take a 6-8 weeks of pain for the retarget
17:14 goatpig assuming some 20% hashrate
17:14 goatpig and you're in the clear
17:14 goatpig afterall etc is alive
17:15 goatpig and they don't even have the technical community on their side
17:16 notmike Nah, man, just set retarget to 288 blocks w/ reset to 2016 48hrs after that, and issue a very public oporder
17:16 goatpig sure that's acceptable too
17:16 goatpig if you are in the politics of no HF, the wait period is acceptable however
17:16 goatpig and you get to enforce some idea of integrity
17:17 notmike Yeah, definitely! One problem with bitcoin is everyone wants to rush shit.
17:17 goatpig im fine with the status quo really
17:17 notmike Need a new lambo!
17:17 goatpig there's eth ICOs for that
17:18 goatpig let's segregate duties
17:25 notmike Do you wanna bed friends now?
17:25 notmike Be* yikes
18:29 goatpig i dont care either way?
21:45 notmike goatpig: well that's just sad
22:09 goatpig such is life
22:49 eck what's the convention with these global variables prefixed with f? e.g. fRequestShutdown and fDumpMempoolLater