sipa changed the topic of #bitcoin-wizards to: This channel is for discussing theoretical ideas with regard to cryptocurrencies, not about short-term Bitcoin development | http://bitcoin.ninja/ | This channel is logged. | For logs and more information, visit http://bitcoin.ninja
tjader has quit [Ping timeout: 272 seconds]
tjader has joined #bitcoin-wizards
justanotheruser has joined #bitcoin-wizards
binaryFate has quit [Ping timeout: 255 seconds]
adlai has quit [Ping timeout: 264 seconds]
adlai has joined #bitcoin-wizards
Monthrect is now known as Piper-Off
Oizopower has joined #bitcoin-wizards
zookolaptop has joined #bitcoin-wizards
ThomasV has quit [Ping timeout: 256 seconds]
rusty has joined #bitcoin-wizards
frankenmint has joined #bitcoin-wizards
stevenroose_ has joined #bitcoin-wizards
gocrazy has quit [Ping timeout: 260 seconds]
Tiraspol has quit []
dnaleor_ has joined #bitcoin-wizards
dnaleor has quit [Read error: Connection reset by peer]
frankenmint has quit [Ping timeout: 260 seconds]
Tiraspol has joined #bitcoin-wizards
gocrazy has joined #bitcoin-wizards
voxelot has quit [Remote host closed the connection]
voxelot has joined #bitcoin-wizards
GAit has quit [Quit: Leaving.]
hazirafel has quit [Remote host closed the connection]
Tiraspol has quit [Remote host closed the connection]
Tiraspol has joined #bitcoin-wizards
Tiraspol has quit [Changing host]
Tiraspol has joined #bitcoin-wizards
phiche has joined #bitcoin-wizards
shen_noe has quit [Quit: Leaving]
frankenmint has joined #bitcoin-wizards
Tiraspol has quit [Remote host closed the connection]
mkarrer_ has quit [Remote host closed the connection]
Tiraspol has joined #bitcoin-wizards
veleiro has joined #bitcoin-wizards
frankenmint has quit [Remote host closed the connection]
GAit has quit [Quit: Leaving.]
GAit has joined #bitcoin-wizards
ThomasV has joined #bitcoin-wizards
phiche has quit [Quit: Leaving.]
rusty has joined #bitcoin-wizards
Tomiii has quit [Read error: Connection reset by peer]
Jeremy_Rand_2 has quit [Ping timeout: 240 seconds]
Tiraspol has quit [Remote host closed the connection]
Tiraspol has joined #bitcoin-wizards
Tiraspol has quit [Changing host]
Tiraspol has joined #bitcoin-wizards
phiche has joined #bitcoin-wizards
Crowley4k has quit [Quit: Connection closed for inactivity]
Tiraspol has quit [Remote host closed the connection]
Tiraspol has joined #bitcoin-wizards
Tiraspol has quit [Changing host]
Tiraspol has joined #bitcoin-wizards
dstadulis has joined #bitcoin-wizards
eudoxia has quit [Quit: Leaving]
Tomiii has joined #bitcoin-wizards
nonaTure has quit [Quit: Leaving.]
phiche has quit [Quit: Leaving.]
PeterL has left #bitcoin-wizards ["Leaving"]
stevenroose_ has quit [Ping timeout: 255 seconds]
Tiraspol has quit [Remote host closed the connection]
Tiraspol has joined #bitcoin-wizards
Tiraspol has quit [Changing host]
Tiraspol has joined #bitcoin-wizards
dstadulis has quit [Quit: ZZZzzz…]
tjader has quit [Ping timeout: 276 seconds]
Tiraspol has quit [Remote host closed the connection]
Tiraspol has joined #bitcoin-wizards
tjader has joined #bitcoin-wizards
phiche has joined #bitcoin-wizards
laurentmt has joined #bitcoin-wizards
tjader has quit [Ping timeout: 240 seconds]
priidu has quit [Ping timeout: 245 seconds]
GAit has quit [Read error: Connection reset by peer]
GAit has joined #bitcoin-wizards
phiche has quit [Quit: Leaving.]
Tiraspol has quit [Remote host closed the connection]
Tiraspol has joined #bitcoin-wizards
Tiraspol has quit [Changing host]
Tiraspol has joined #bitcoin-wizards
adlie has joined #bitcoin-wizards
conner has quit [Remote host closed the connection]
bramc has joined #bitcoin-wizards
bramc is now known as Guest65395
phiche has joined #bitcoin-wizards
conner has joined #bitcoin-wizards
dstadulis has joined #bitcoin-wizards
Tiraspol has quit [Remote host closed the connection]
Tiraspol has joined #bitcoin-wizards
laurentmt has quit [Quit: laurentmt]
tjader has joined #bitcoin-wizards
voxelot has quit [Ping timeout: 240 seconds]
adlie has quit [Quit: Not all who wonder for lust]
memymo has joined #bitcoin-wizards
cryzlr has joined #bitcoin-wizards
RoboTeddy has joined #bitcoin-wizards
memymo has quit [Quit: My Mac has gone to sleep. ZZZzzz…]
memymo has joined #bitcoin-wizards
RoboTedd_ has quit [Ping timeout: 260 seconds]
Tiraspol has quit [Remote host closed the connection]
Tiraspol has joined #bitcoin-wizards
erasmospunk has quit [Remote host closed the connection]
rusty has quit [Ping timeout: 265 seconds]
Guest65395 is now known as bramc
dnaleor_ has joined #bitcoin-wizards
phiche1 has joined #bitcoin-wizards
dnaleor has quit [Ping timeout: 240 seconds]
phiche has quit [Ping timeout: 250 seconds]
b-itcoinssg has joined #bitcoin-wizards
Tiraspol has quit [Remote host closed the connection]
Tiraspol has joined #bitcoin-wizards
markus-k has quit [Quit: My Mac has gone to sleep. ZZZzzz…]
Heliox has left #bitcoin-wizards ["Leaving"]
dEBRUYNE has quit [Ping timeout: 265 seconds]
belcher has joined #bitcoin-wizards
GAit has quit [Quit: Leaving.]
Erik_dc has joined #bitcoin-wizards
GAit has joined #bitcoin-wizards
dstadulis has quit [Quit: ZZZzzz…]
RoboTeddy has quit [Remote host closed the connection]
RoboTeddy has joined #bitcoin-wizards
RoboTeddy has quit [Ping timeout: 245 seconds]
priidu has joined #bitcoin-wizards
nivah has joined #bitcoin-wizards
bit2017 has quit [Ping timeout: 240 seconds]
Tiraspol has quit [Remote host closed the connection]
Tiraspol has joined #bitcoin-wizards
Tiraspol has quit [Changing host]
Tiraspol has joined #bitcoin-wizards
voxelot has joined #bitcoin-wizards
memymo has quit [Quit: My Mac has gone to sleep. ZZZzzz…]
<maaku>
~200 what?
memymo has joined #bitcoin-wizards
tjader has quit [Ping timeout: 276 seconds]
conner has quit [Remote host closed the connection]
Jeremy_Rand_2 has joined #bitcoin-wizards
Tiraspol has quit [Remote host closed the connection]
gielbier has quit [Quit: Leaving]
Tiraspol has joined #bitcoin-wizards
Tiraspol has quit [Changing host]
Tiraspol has joined #bitcoin-wizards
dEBRUYNE has joined #bitcoin-wizards
phiche has joined #bitcoin-wizards
shesek has quit [Ping timeout: 276 seconds]
tjader has joined #bitcoin-wizards
phiche1 has quit [Ping timeout: 264 seconds]
bramc has quit [Quit: This computer has gone to sleep]
conner has joined #bitcoin-wizards
<tromp_>
presumably 200 bits of private key
RoboTeddy has joined #bitcoin-wizards
conner has quit [Ping timeout: 245 seconds]
paci has quit [Ping timeout: 260 seconds]
paci has joined #bitcoin-wizards
bramc has joined #bitcoin-wizards
bramc is now known as Guest1739
paveljanik has quit [Quit: Leaving]
frankenmint has joined #bitcoin-wizards
bitcoin-wizards2 has joined #bitcoin-wizards
<bitcoin-wizards2>
l
bitcoin-wizards2 has quit [Client Quit]
qawap_ is now known as qawap
frankenmint has quit [Remote host closed the connection]
phiche has quit [Quit: Leaving.]
memymo has quit [Quit: My Mac has gone to sleep. ZZZzzz…]
Guest1739 is now known as bramc
moa has joined #bitcoin-wizards
hashtag has quit [Ping timeout: 260 seconds]
GAit has quit [Quit: Leaving.]
<bramc>
I'm tempted to write up a FAQ about Bitcoin for crypto experts
GAit has joined #bitcoin-wizards
<nsh>
goforit
<bramc>
Q. What's all this stuff about macroeconomics and golbuggism? Isn't goldbuggism not taken seriously by real economists? A. Yes, Bitcoin is a commodity not a currency, just ignore that discussion
<bramc>
Q. What's keeping Bitcoin's price up? Couldn't it swing down to $10 tomorrow? A. Yes it could. Caveat Emptor.
<bramc>
Q. Doesn't Bitcoin mining inherently waste lots of resources doing nothing? A. Yes, it does.
<gwillen>
heh
<bramc>
Q. Doesn't Bitcoin have a hard scaling limit at a not very high amount? What happens when it reaches that? A. Yes it does. At that point fees go up. People are working on net settlement mechanisms to dramatically improve scaling.
bendavenport has quit [Quit: bendavenport]
<belcher>
have you often come across crypto experts interested but ignorant of btc ?
bendavenport has joined #bitcoin-wizards
<bramc>
Q. What is Bitcoin good for? A. Bitcoin is good for transferring value reliably and incontrovertibly. To a lesser extent it's good at sheltering assets from being seized, but they could be frozen much more easily than people think. There's also some interesting work on smart transactions going on.
Starduster has quit [Ping timeout: 260 seconds]
<bramc>
belcher, Yes all the time in crypto circles. Often they've looked into bitcoin enough to be aware of its puke-inducing characteristics and are either turned off by them or think they misunderstand or think that bitcoin partisans are a bunch of liars who misrepresent what it is.
<bramc>
These people always find it reassuring that there's a core group of devs who are well aware of all the problems and don't bullshit about them.
<kanzure>
bramc: a cryptography-oriented FAQ about bitcoin might be particularly interesting for recruiting reviewers etc. iirc one in particular called bitcoin a "denial-of-service attack on crypto review time".
<kanzure>
wait that's not the quote... what did djb say?
Erik_dc has quit [Remote host closed the connection]
<bramc>
kanzure, It could certainly be expanded into something which is actually helpful for crypto people trying to learn, the basic idea here is to get across the message 'Yes those are all serious issues and we know about them and don't have great solutions'
skyraider_ has quit [Quit: Connection closed for inactivity]
<kanzure>
ah the quote was "Bitcoin is a particularly effective DoS against security professionals"
<coinoperated>
<bramc>: Q. What's keeping Bitcoin's price up? <--- the more intriguing half of the question remains unanswered
Tiraspol has quit [Remote host closed the connection]
Tiraspol has joined #bitcoin-wizards
Tiraspol has quit [Changing host]
Tiraspol has joined #bitcoin-wizards
mihar has joined #bitcoin-wizards
WyzBTC1 has quit [Ping timeout: 256 seconds]
<zookolaptop>
coinoperated: What's the more intriguing half? "What's keeping Bitcoin's price down?" ?
Tiraspol has quit [Remote host closed the connection]
Tiraspol has joined #bitcoin-wizards
Tiraspol has quit [Changing host]
Tiraspol has joined #bitcoin-wizards
<coinoperated>
zookolaptop: the less intriguing half was truncated
T23WS has joined #bitcoin-wizards
bramc has quit [Quit: This computer has gone to sleep]
atgreen has quit [Ping timeout: 240 seconds]
conner has joined #bitcoin-wizards
bramc has joined #bitcoin-wizards
bramc is now known as Guest63476
<moa>
What is keeping Bitcoin's price down?
Tiraspol has quit [Remote host closed the connection]
Tiraspol has joined #bitcoin-wizards
Tiraspol has quit [Changing host]
Tiraspol has joined #bitcoin-wizards
IAmNotDorian has quit [Ping timeout: 250 seconds]
conner has quit [Ping timeout: 256 seconds]
Tiraspol has quit [Remote host closed the connection]
Tiraspol has joined #bitcoin-wizards
Tiraspol has joined #bitcoin-wizards
b-itcoinssg has quit [Quit: Connection closed for inactivity]
<bsm117532>
moa: -> #bitcoin-pricetalk
bendavenport has quit [Quit: bendavenport]
<moa>
:( #bitcoin-pricetalk is invitation only, wish I never knew it existed now, thnx bsm117532
veleiro has quit [Ping timeout: 252 seconds]
bendavenport has joined #bitcoin-wizards
toffoo has joined #bitcoin-wizards
<gwillen>
moa: try ##bitcoin, I hear that's a place you can go
<gwillen>
although I have not been in it myself
rusty has joined #bitcoin-wizards
hashtag has joined #bitcoin-wizards
ThomasV has quit [Ping timeout: 272 seconds]
soupcan has joined #bitcoin-wizards
zookolaptop has quit [Ping timeout: 245 seconds]
Guest63476 is now known as bramc
ThomasV has joined #bitcoin-wizards
<bramc>
That is so annoying
erasmospunk has joined #bitcoin-wizards
Tiraspol has quit [Remote host closed the connection]
PeterR has joined #bitcoin-wizards
[\\\] has joined #bitcoin-wizards
Tiraspol has joined #bitcoin-wizards
<PeterR>
Statement: "The expectation value of the time between when a miner solves a block and when 50% of the hash rate has received and validated that block increases with the size of the block mined."
<PeterR>
Choose the best answer:
<PeterR>
A. TRUE
<PeterR>
B. FALSE
zookolaptop has joined #bitcoin-wizards
[\\\] is now known as tripleslash
<bsm117532>
PeterR: obviously true.
erasmospunk has quit [Remote host closed the connection]
<PeterR>
What would be the best way to prove that?
<kanzure>
validation cost
<bsm117532>
Prove that if A > B, A takes more time to transmit across the network; and block B takes longer to validate.
<phantomcircuit>
PeterR, your question is under specified, is there a single miner which is > 50% of the hash rate?
<bsm117532>
haahaaa phantomcircuit
<rusty>
bsm117532: um, no, consider pre-publishing solutions. The publishing of the coinbase + nonce is all that is required then, which is not proportional to block size.
<rusty>
This ground has been well covered previously.
<kanzure>
pathological case of exotic block from non-cooperative-pre-publication miner
<kanzure>
*non-pre-publication
<phantomcircuit>
rusty, pre-publishing solutions do not work under adversarial conditions, which we must assume for security considerations
<phantomcircuit>
rusty, the answer to his question is true, unless more than 50% of the hash rate is centralized then false for blocks found by that miner
<phantomcircuit>
which is to say that larger blocks inherently cause centralization pressure
<phantomcircuit>
but im sure he's just trying to manipulate people into answering an under specified question
<kanzure>
i wonder if maaku would agree that bandwidth cost needs to be considered in validation cost metrics
<kanzure>
for validation cost metrics purely for rating/scoring block contents, i think the answer is no
<PeterR>
bsm117532: would you agree that--ceteris paribus--communicating more *information* takes more time, as per Shannon's carrying capacity theorem?
<rusty>
phantomcircuit: depends on what you're trying to measure. The question of "is it possible to mine s.t. costs are not proportional to block size" vs "can we force others to mine s.t. costs are not proportional to block size"
<phantomcircuit>
rusty, "s.t." ?
<rusty>
phantomcircuit: such that
<kanzure>
PeterR: the more centralized the network the less information that needs to be propagated
<bsm117532>
PeterR: yes of course. But the block is artificially time-restricted to be "as fast as possible after creation" (my assumption) so that the weak block ideas spread it out over 10 minutes (rusty's correct comment).
<PeterR>
Agreed.
<phantomcircuit>
rusty, the answer is of course yes, but only if we assume a level of coordination between miners such that they are effectively a single entity
dnaleor has joined #bitcoin-wizards
<moa>
kanzure: excellent point ... taking the converse we could show how decentralised the network is by how much information is being transmitted?
<kanzure>
phantomcircuit: that's a non-statement. your other statements were better. :P
<PeterR>
BSM117532: OK good. Just going over the basics to find the points of contention.
<coinoperated>
phantomcircuit: is there a good reason to assume this is not the case?
<kanzure>
moa: no? anyone can voluntarily saturate their bandwidth. that's not interesting.
<phantomcircuit>
kanzure, if miners are cooperating on transaction selection and pre-propagation then they are effectively a single miner and the system has failed
<moa>
kanzure: meaningful information
<bsm117532>
PeterR: if blocks were so big that it takes 10 minutes or more to validate or transmit, you're back to entropy limits.
binaryFate has joined #bitcoin-wizards
<kanzure>
phantomcircuit: huh? i think you mean "if it's mandatory for participation", right?
<phantomcircuit>
kanzure, it *would* be mandatory to participate with sufficiently large block size
<kanzure>
well okay. right, if centralization pressure (whatever the metric is in aggregate) passes some threshold, yea i could see a run-away effect being more pronounced.
<PeterR>
bsm117532: Dr. Nicolas Houy and Andrew Stone (independently) showed that if a miner *expects* that a block would take > 10 min to propagate/verify, that the Nash equilibrium is to not bother mining on it at all.
<kanzure>
PeterR: btw it is not just 50% of the miners that need to receive the block (or instead of 50%, whatever the largest consensus is necessary to achieve for that 10min period or w/e), it's also the validating nodes on the network that need to receive the block
<PeterR>
In other words, if a miner publishes a block that I would expect to take 12 min to verify, then I should immediately orphan it.
<bsm117532>
PeterR: You mean someone else's block that you just received? That makes sense.
<phantomcircuit>
kanzure, there is no known mechanism for improving block relay which works without coordination between miners in transaction selection
rusty has left #bitcoin-wizards [#bitcoin-wizards]
<PeterR>
kanzure: that point is orthogonal to my question.
<PeterR>
bsm: yes.
<kanzure>
PeterR: perhaps if you only want mining .... but nobody wants only mining. it's useless.
<maaku>
kanzure: I suspect the concerns are separable. By adopting a validation-cost metric with a witness discount, and combining that with probabilistic validation we can remove validation cost as an active concern
<phantomcircuit>
kanzure, and actually thinking about it more as blocks become larger the cost of that coordination would grow non-linearly unless there was a single central entity selecting transactions
<kanzure>
maaku: oh. okay then.
<bsm117532>
PeterR: likewise in the absence of code-imposed artificial restrictions on block size, there's a Nash equilibrium maximum block size that can be computed.
<maaku>
kanzure: then we do something else (e.g. something weak blocks derived) to address bandwidth concerns
<PeterR>
exactly!
<phantomcircuit>
PeterR, are you going to at least re-state the question such that it is fully specified?
<kanzure>
there's no possible way that it's orthogonal to your question unless you don't mean bitcoin mining
<kanzure>
if you don't mean bitcoin mining then you should have said so :-)
<PeterR>
No, because I asked for the "best" answer.
<maaku>
kanzure: i mean both need to be addressed, but not necessarily by the same mechanism
<phantomcircuit>
maaku, weak blocks at best result in a 50% bandwidth reduction (literally best case and unachievable in practice)
<kanzure>
maaku: indeed. limit can be separate from the validation cost metric. seems reasonable to me.
<bsm117532>
PeterR: You're one of those people that writes SAT tests, aren't you...
<phantomcircuit>
PeterR, then your question is intentionally deceptive and you should feel bad
<PeterR>
:P
<kanzure>
i feel bad about almost everyone all the time. so it works out.
Starduster has joined #bitcoin-wizards
<maaku>
phantomcircuit: by weak blocks I'm including a whole category of things like bitcoin-ng, braids, dag trees, etc.
<PeterR>
The tongue out was at BSM, not you phantomcircuit
Guyver2 has quit [Quit: :)]
<maaku>
phantomcircuit: that's less about bandwidth reduction so much as not making transmission of the tx data in the critical path for block relay
<bsm117532>
Speaking of which, I'm gonna switch locales and pound out, now that I'm done with company stuff for the day... *sigh*
<phantomcircuit>
maaku, all of those require the transaction data to be sent at least once
<phantomcircuit>
maaku, ok, still doesn't even begin to address initial block synchronization costs
PeterR has quit [Quit: Page closed]
<kanzure>
PeterR: "use the same pool" is not the same thing as "use slightly more efficient block propagation techniques"
<dgenr8>
PeterR: you might need to say "does not decrease"
<maaku>
phantomcircuit: yes, but right now in an adversarial case the entire block must be sent within say 30s of the block being found
<kanzure>
wtf he left that's not fair
<maaku>
if we can at least spread that out over the full 10min we could reduce bandwidth requirements by an order of magnitude
<phantomcircuit>
kanzure, i told you it was a trap
<phantomcircuit>
he's a genuinely bad person
<maaku>
but existing ways for doing that (e.g. IBLT, relay network) do not have adversarial guarantees. it's possible an advsarially secure alternative exists with weak blocks, dag trees, etc.
<phantomcircuit>
maaku, uhm... an adversarial miner simply doesn't send their weak blocks
<kanzure>
i am confused as to why phantomcircuit has had to explain adversarial mining to rusty and maaku
<kanzure>
is there an actual disagreement here? what is it.
deego has joined #bitcoin-wizards
<maaku>
kanzure: phantomcircuit is confused about what things like bitcoin-ng are promising
<smooth>
bad person is unfair but i do think his behavior has deteriorated over this conflict
<smooth>
not alone in that either
<kanzure>
he didn't say propagation efficiency improvements aren't promising
<kanzure>
he jsut says they have at most a 50% bandwidth improvement
<kanzure>
50% is pretty great
Yoghur114_2 has quit [Remote host closed the connection]
<coinoperated>
the guy gers pilloried in here every time i see his name mentioned. surprised he even pops in
<coinoperated>
gets
<kanzure>
oh i see, maybe some disagreement about initial block synchronization costs
<phantomcircuit>
kanzure, maaku's point is that their goal is to pre-stage the transaction data to avoid a bandwidth spike when a block is found
<maaku>
kanzure: 50% is great. But I'm more interested in the 2000% improvement it would yield if it make block transmission not dependent on tx relay
<kanzure>
have we been actively monitoring initial block synchronization costs
<kanzure>
because that's sort of something we must not let explode
<phantomcircuit>
kanzure, which i do not believe can be accomplished if a miner with significant hashing power is acting adversarially
<kanzure>
where is the 2000% coming from? why talk about the average case?
<coinoperated>
kanzure: has there been a DDoS attack built around spurious initial sync requests yet? seems like such an obvious exploit
<phantomcircuit>
kanzure, he's talking about latency improvements, not total bandwidth reduction
<kanzure>
single packet based on IBLT + weak blocks? that's... look, we're not designing a system where we assume there is agreement.
<phantomcircuit>
kanzure, which is interesting for mining centralization risks; but is completely un-nteresting if you look at the cost to operate a full node
<phantomcircuit>
un-interesting*
<coinoperated>
just keep asking for more sync blocks, never save them to disk. BottomlessBit
<kanzure>
if we want efficient agreement we know how to do that in centralized models
Tiraspol has quit [Remote host closed the connection]
<kanzure>
but yea i agree that latency can be improved
<kanzure>
but also, phantomcircuit points out that efficiency improvements might also be centralizing pressures
Tiraspol has joined #bitcoin-wizards
<kanzure>
bramc: i haven't looked at that one yet.
<maaku>
kanzure: the design question is can we add consensus rules which punish blocks that include transactions not previously seen
<phantomcircuit>
kanzure, this is yet another point where the optimal strategy for miners and the optimal strategy for regular users are radically different
<kanzure>
bramc: thanks for the link
<bramc>
I'm puzzling over what they're claiming their construction does. It seems to be a single thing which is an either/or between proof of space and proof of time
<phantomcircuit>
for regular system users the optimal strategy is almost always to operate in blocks only mode such that they receive only the block data and maintain nothing in their mempool
<maaku>
kanzure: it is surprising that the answer is not a downright no, but more research is needed
<phantomcircuit>
for miners the optimal strategy is to have all of the transactions possible pre-staged in memory to avoid signature validation and bandwidth costs when a new block is found
<kanzure>
punishment wont work for censorship resistance, unless you can convince miners to somehow commit to transaction details they haven't seen
<kanzure>
also, centralizing pressures around punishment could probably harm low-bandwidth miners
<maaku>
kanzure: my assumptions regarding maximum acceptable blocks is predicated on block relay taking no more than 30s. if you can presume all txns are pre-staged, then you can take the full 10minute block window to relay for a 20x improvement
<phantomcircuit>
the only proposal so far that even attempts to reduce the cost of operating a full node is segregated witness and that only works because we dont validate most of the signatures anyways
<phantomcircuit>
maaku, if that's your metric then you are going to end up overstating what the system is capable of by many hundreds of times
memymo has joined #bitcoin-wizards
<kanzure>
yeah i am not sure about that assumption
Tiraspol has quit [Remote host closed the connection]
<maaku>
kanzure: there's basically two approaches. one is mining policy for a majority of miners to share weak blocks and not build on blocks that diverge too far from the "pre-consensus" the weak blocks established
<kanzure>
there is no such thing as pre-consensus :-)
Tiraspol has joined #bitcoin-wizards
<kanzure>
that's like a pre-photon
ThomasV has quit [Ping timeout: 272 seconds]
<maaku>
kanzure: the other is to have a block contain NO transactions, but with a one-use-only signature that they later use to sign the blocks
zookolaptop has quit [Ping timeout: 260 seconds]
<maaku>
*sign the list of transactions
<coinoperated>
kanzure: pre-photons conditions exist though
<kanzure>
pre-photon conditions are subject to the same constraints as regular reality
<kanzure>
nsh where are you we need you
adnn has joined #bitcoin-wizards
<nsh>
(certainly not lurking in the scrollback, muttering to himself)
<phantomcircuit>
maaku, even if the cost to relay blocks could be reduced to effectively zero under adversarial conditions the cost to complete initial block synchronization must not be so large that no new participant can complete it
<maaku>
kanzure: ok. imagine that a >50% mining cabal decided to only build on p2pool blocks, and recent p2pool blocks at that
zookolaptop has joined #bitcoin-wizards
<nsh>
'<PeterR> In other words, if a miner publishes a block that I would expect to take 12 min to verify, then I should immediately orphan it.'
<nsh>
(game-theory does not work that way)
<nsh>
(the utility of mining on a block is dependent on how others are going to treat it, not your personal calculation)
<coinoperated>
nsh: isn't this the same thing as saying that calculating a 12 min block time is impossible
PeterR has joined #bitcoin-wizards
<coinoperated>
or any arbitrary verification time estimate
<PeterR>
nsh: that
<nsh>
you can always calculate something that approximates the propagation time, that's just a terribly poor proxy for the decision of other miners to proceed on the tip
<PeterR>
that wasn't my statement
<nsh>
(or wait for a block below the propagation threshold)
<PeterR>
Apparently there's a Nash equilibrium. Miner's should orphan any block they believe will take longer than 10 min to propagate/verify.
<nsh>
i don't think anyone who is doing mining is even making such assessments, or has the tools so to do, or knows how to make the tools if they wanted them
<PeterR>
*miners
<PeterR>
Nsh: agreed.
<nsh>
well, yes, that's how toy maths works
<nsh>
but the real work depends on what code is running
<nsh>
not what the Nash equilibrium is
Tiraspol has quit [Remote host closed the connection]
<PeterR>
And the code converges to the Nash equilibrium
Tiraspol has joined #bitcoin-wizards
Tiraspol has quit [Changing host]
Tiraspol has joined #bitcoin-wizards
* nsh
rolls eyes compassionately
<PeterR>
As miners learn more
eudoxia has joined #bitcoin-wizards
<nsh>
it would be good to simulate what happens when blocks are no longer propagating to some kind of node saturation before the expectation time for the next block
<nsh>
that's definitely something to investigate, as empirically as possible
<PeterR>
kanzure: sorry for leaving for a moment.
GAit has quit [Read error: Connection reset by peer]
<kanzure>
in the absence of anything but mining, it would make sense to say that 50% (or less) of the hashrate needs to receive a block during propagation, but since the network does intentionally have non-mining fully-validating nodes, they also have to receive their money and blocks
<phantomcircuit>
PeterR, i like how you're listening to the channel after you leave
<coinoperated>
nsh: would it be simpler to just do this with Tx alone, and pretend blocks don't exist, for a naive model.
<nsh>
how do you mean sorry?
<coinoperated>
ultimately even if block prop is "solved" Tx prop is next
<PeterR>
phantomcircuit: why was my question deceptive?
<PeterR>
Statement: "The expectation value of the time between when a miner solves a block and when 50% of the hash rate has received and validated that block increases with the size of the block mined."
<coinoperated>
nsh: model bitcoin network as if it propagated Tx without need for consensus mechanism that filters out double spends. just: broadcast a Tx, model how fast 51% of the network sees it and can trivially ack to 49%, 48%, 47% that has not seen it. how long would this take if everyone plays fair