00:47:04aburan28:aburan28 is now known as ab428
02:38:37dgenr8:sipa: because every full nodes validates the rules, AND the nodes are connected in a cohesive p2p network. let it never be suggested that the p2p network is disposable, not integral, etc.
03:08:54gmaxwell:dgenr8: "the p2p network" is complete disposable and already unnecessary (there are nodes connected in ways other than the p2p network)
03:09:31gmaxwell:there are at least three other actively used alternative transports that I'm currently aware of, and we'll see more in the future.
03:09:54gmaxwell:(relay network, bitcoin-over-freenet, and the DVB radio blockchain stuff)
03:17:02dgenr8:the transport is not what's important, the way nodes find each other and relay inventory is
03:17:47gmaxwell:No thats not important either.
03:17:58phantomcircuit:dgenr8, censorship needs to be difficult, but that doesn't require a p2p network really
03:18:08phantomcircuit:it's just a relatively easy thing to build with bitcoin
03:18:09gmaxwell:It's important that you learn the blockchain. (that you're not partitioned from the honest users) but anything that accomplishes that is acceptable.
03:18:18gmaxwell:E.g. having transport over facebook would be fine.
03:19:20gmaxwell:(p2p of bitcoin's type is actually a very weak tool, easily attacked.. but fortunately the blockchain is strong enough that that weakness hardly matters. A friends network like cjdns links would be better, except the setup costs are unacceptable)
03:19:36phantomcircuit:gmaxwell, i wonder if facebook would let me upload pictures with blockchain data stuffed into EXIF fields
03:20:54mappum:phantomcircuit: or you can just steganographically put it in the image in a way that survives facebook's JPG compression
03:21:57dgenr8:if bitcoin were just the blockchain, it would not survive the 70% miner rule change scenario discussed above
03:23:17dgenr8:this is going to end up with the realization everyone was saying the same thing, sorry
03:24:07phantomcircuit:mappum, that might be hilarious
03:24:48phantomcircuit:i bet i could do giant qr codes
03:24:57phantomcircuit:lol im going to try that
03:25:56mappum:phantomcircuit: My photo albums: Summer Vacation 2012, Hawaii Trip 2013, Bitcoin Blockchain
03:26:02dgenr8:gmaxwell you say some crazy things in your quest to discredit my every utterance ;)
03:26:29phantomcircuit:are they crazy if they're true?
03:26:34phantomcircuit:im thinking no
03:55:26gmaxwell:genr8: if it were "just" the blockchain it wouldn't be a working system at all, there must be some method to communicate somehow.
03:55:29gmaxwell:dgenr8: We make two assumptions up front normally for bitcoin, that the majority of the hashpower is honest and that the honest participants
03:55:32gmaxwell:are not partitioned. The current p2p network is not very partitioning resistant, but it doesn't really need to be because partitioning alone
03:55:35gmaxwell:is mostly just a dos attack.
03:55:37gmaxwell:dgenr8: Under the example you're posing there (majority of miners want more subsidy) the current p2p network would not be ideal because it is
03:55:40gmaxwell:relatively cheap to partition. Other transport mechnisms may be better in a case where partitioning would be more valuable.
03:55:43gmaxwell:Certantly in no case is the current p2p specifically essential to the system.
03:57:01gmaxwell:dgenr8: As for "quest to discredit" insult, keep in mind that you're the one that decided to make a completely out of left field argument here; I didn't seek out anything. And your argument is just attempting to continue a position you've advanced before that I think is pretty easily demonstrated to be incorrect.
03:57:17gmaxwell:(by the fact that people aready use the full bitcoin system without using the p2p network)
04:00:56gmaxwell:(e.g. CJDNS and freenet non-opennet are example networks where partitioning a honest subgraph should be substantially more difficult; hub/spoke topologies are generally strongly partitioning resistant, though have other limitations)
04:01:20gmaxwell:Fortunate Bitcoin doesn't have just one transport so it can benefits from the benefits of multiple without suffering the all the costs of any single one.
04:15:35op_mul:phantomcircuit: I have done that. though for facebook the compression makes it almost useless. try somewhere like imgur or flickr which support PNGs. you can even leverage the pretty good compression in a PNG to get some savings over the original binary.
04:16:01op_mul:phantomcircuit: there's other methods, too. https://soundcloud.com/blockchain/00000000000000000f54a1fe88f6a4a2bf98952e60b57376eb15d80e9c753c04
04:21:19user7779_:op_mul: can you explain what that is exactly
04:21:28user7779_:or how the block info is being rendered via sound?
04:22:02op_mul:http://www.whence.com/minimodem/
04:22:58user7779_:awesome, thanks
04:23:25op_mul:there's no reason you couldn't do bitcoin-over-shopping-center-PA if you felt so inclined. though you might drive people inside a little insane.
04:24:32user7779_:hahah, yes probably.
04:25:45contrapumpkin:contrapumpkin is now known as copumpkun
04:25:50copumpkun:copumpkun is now known as copumpkin
04:36:10luke-jr_:luke-jr_ is now known as Luke-Jr
04:40:39rusty:maaku: do you have an implementation or more details on your maarkle tree structure for prev blocks?
04:43:18op_mul:rusty: nice portmanteau.
04:45:18rusty:op_mul: thanks
05:24:25nuke_:nuke_ is now known as nuke1989
05:40:05maaku:merkle tree structure? sure see this (unfinished) draft bip: https://gist.github.com/maaku/2aed2cb628024800044d
05:40:24maaku:regarding the block header back links, just what was posted here to -wizards and the mailing list
05:40:41maaku:but I intend to use the same structure in that bip
05:40:57maaku:the keys are just meaningless
09:05:14hitchcock.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
09:05:14hitchcock.freenode.net:Users on #bitcoin-wizards: andy-logbot soundx adam3us Guyver2 RoboTedd_ paveljanik kyletorpey Muis CryptOprah BlueMatt mappum bosma_ dgenr8 Baz____ btc__ OneFixt Rynomster copumpkin austinhill thrasher` rusty Adlai hashtag ryanxcharles tlrobinson Dr-G2 wallet42 mortale DougieBot5000 PaulCapestany hashtagg Starduster waxwing shesek afk11 bobke AdrianG wumpus isis_ [\\\] Luke-Jr MRL-Relay Dyaheon- phantomcircuit Greed` MoALTz stonecoldpat Aquent morcos pi07r c0rw1n
09:05:14hitchcock.freenode.net:Users on #bitcoin-wizards: prodatalab samson_ bsm117532 coiner GAit mr_burdell eric atgreen Shiftos dansmith_btc spinza tacotime berndj nuke1989 Graftec NikolaiToryzin mkarrer ompik [d__d] rfreeman_w Emcy toddf jaromil gues Iriez luny heath Guest3112 jgarzik go1111111 Graet_ Transisto bbrittain husebyAFK DoctorBTC BananaLotus Apocalyptic EasyAt tromp espes___ jbenet null_radix hollandais midnightmagic starsoccer gavinandresen postpre danneu catcow TD-Linux ryan-c
09:05:14hitchcock.freenode.net:Users on #bitcoin-wizards: roasbeef amiller smooth optimator_ mmozeiko Alanius epscy kobud JonTitor Hunger- fluffypony kanzure warren Krellan nanotube fenn otoburb LarsLarsen coutts jchp nsh_ asoltys_ sl01_ K1773R nickler_ a5m0 @ChanServ Meeh comboy btcdrak ahmed_ lnovy eordano nsh gribble gwillen Nightwolf Anduck cfields digitalmagus8 gmaxwell pigeons andytoshi lechuga_ artifexd hguux_ michagogo kumavis warptangent Fistful_of_Coins HM2 paperbot maaku Guest38445
09:05:14hitchcock.freenode.net:Users on #bitcoin-wizards: coryfields iddo harrow BrainOverfl0w phedny Keefe helo BigBitz so crescendo petertodd throughnothing Taek poggy tromp_ wizkid057 burcin livegnik sipa harrigan sneak Eliel_ s1w lclc_bnc Logicwax yoleaux azariah kinlo
09:05:14hitchcock.freenode.net:[freenode-info] channel flooding and no channel staff around to help? Please check with freenode support: http://freenode.net/faq.shtml#gettinghelp
10:11:29lclc_bnc:lclc_bnc is now known as lclc
11:54:07Greed`:Greed` is now known as Greed
12:18:36samson2:samson2 is now known as samson_
12:51:34lclc:lclc is now known as lclc_bnc
13:55:44wallet421:wallet421 is now known as wallet42
14:30:31fanquake:fanquake has left #bitcoin-wizards
14:46:55NewLiberty_:NewLiberty_ is now known as NewLiberty
15:39:31amiller:petertodd, it seems like you're trying to do formal reasoning with proof-of-publication and anti-replay-signatures as primitives but there's not a clear enough definition of either yet to really call it a theorem
15:40:44amiller:one thing is that you're willing to say that Proof-of-publication can implement anti-replay-signatures but you're not mentinoing the efficiency loss
15:41:20amiller:you're saying it counts as an "anti-replay-signature" protocol even if you have to rescan through every transaction in the blockchain to check it
15:43:16amiller:the other thing is you aren't saying anything about efficiency
15:43:52amiller:especially whether it's part of "proof of publication" to assume that everyone making proofs and checking proofs is *storing* or can readily access arbitrary pieces of data that have been previously published
15:48:23amiller:i think that for a proof-of-publication protocol that satisfies your definition, the audience has to be determined at the outset and everyone involved has to do retain all data forever
16:00:25petertodd:amiller: re: efficiency, I've pointed out elsewhere how rescan's aren't a given; appropriate indexes lead to obvious O(n log n) non-publication proofs
16:00:41petertodd:amiller: hell, with UTXO commitments you get an easy ~O(log n) way of doing it
16:01:57petertodd:(and I only say ~ because that's trusting miners fairly heavily)
16:02:59amiller:yeah but that's like
16:03:34amiller:well basically it undoes the entire assumptions about embedded consensus, that you can do this more cheaply than accessing the shared index (utxo index) for every message you want signed inthe system
16:03:50petertodd:huh?
16:04:13petertodd:embedded consensus just assumes we have some consensus system and we can do proof-of-publication (or really even just anti-replay) with it
16:04:26petertodd:e.g. colored coins *is* embedded consensus
16:04:39amiller:right, so, you can achieve 'proof of publication' while implying horrible efficiency for the resulting anti-replay-signature
16:04:52petertodd:well, yeah...
16:05:17petertodd:that's the whole point of something like treechains: O(n log n) non-publication proofs
16:06:24petertodd:you're log n is the path from top chain to specific part in the tree, the other n is just # of block headers needed to prove a non-spend
16:06:39petertodd:call it O(m log n) to be more specific
16:08:13amiller:okay well nevermind then i'm back to not understanding the significance of this philosophical difference at all then :)
16:09:18amiller:actually no i still want to dig apart at your definition
16:09:22petertodd:I understood it as an emphasis on trust vs. verification - if you're heavy on anti-replay you're more willing to accept third party trusting solutions
16:09:59amiller:in the anti-replay spend definition, you need to check that there's no previous message m' under the same pubkey p
16:10:01petertodd:e.g. how merge-mining is willing to accept some very dodgy economic assumptions to reduce that O(m log n) non-publication proof cost down to O(log n)
16:10:07amiller:that doesn't typecheck against proof-of-non-publication
16:10:07petertodd:yup
16:10:19amiller:in proof of non-publication you assert for a specific message m that m has not previously been published
16:10:38petertodd:ah, wait, no you see in the anti-replay spend definition you *trust* that *someone else* did that for you - for them the cost is still O(n log n) and probably worse
16:10:40amiller:so your reduction doesn't work because you can have a proof-of-publication system that can't support anti-replay-sig at all
16:11:21petertodd:nah, thing is you can arrange for that message to also have a signature, and prevent others from spamming invalid signatures
16:11:36amiller:within your current definition?
16:11:49petertodd:or equally, with an index mapping pubkeys to messages with valid signatures
16:12:19amiller:so it's not just about publication but also signature validation
16:12:27petertodd:amiller: there *isn't* a specific definition of exactly what we're saying when we say proof-of-publication; statements about it's efficiency are always going to be in the context of some actual implementation
16:13:14amiller:well look you tried in one letter to give it a definition and in another letter to establish some kind of reduction theorem
16:13:19petertodd:for starters, notice how in a treechains system w/o sig validation by anyone but end users it's O(m log n) with a k being the blocksize
16:13:29petertodd:yeah, you're way overthinkign that
16:13:54amiller:and this is supposed to be helpful for disentangling/comparing some different approaches re: tree chains, side chains, embedded consensus, etc
16:14:23petertodd:yeah, but all this stuff is so tied to actual implementations that getting abstract about it is silly
16:14:45petertodd:I mean, hell, I'm talking about taking advantage of UTXO indexes in my last post - that's got fuck all to do with abstract theory
16:15:13petertodd:equally, the other half of all this is economic considerations, even political considerations, which aren't exactly well captured by any of this
16:17:50amiller:yeah but it's so hard to actually evaluate all these actual implementations and concrete proposals that some abstract theory might help, so i guess my point is to encourage you for trying
16:19:24petertodd:all this stuff boils down to some pretty basic questions about who you are trusting and at what cost; I don't see the difficulty in comparing this stuff at all
16:20:43gwollon:gwollon is now known as Guest39855
16:20:48Anduck_:Anduck_ is now known as Guest70465
16:23:24amiller:"I don't see the difficulty in comparing this stuff at all" <-- maybe i'll get there someday
16:25:45lclcd:lclcd is now known as lclc
16:28:53Anduck__:Anduck__ is now known as Anduck
16:30:01wizkidO57:wizkidO57 is now known as wizkid057
16:33:34irc.freenode.net:Disconnected from irc.freenode.net (ERROR :Closing Link: wpsoftware.net (Ping timeout: 245 seconds))
16:44:09sendak.freenode.net:Users on #bitcoin-wizards: @andy-logbot
16:49:47sendak.freenode.net:*** Notice -- TS for #bitcoin-wizards changed from 1419093849 to 1361863487
16:49:48sendak.freenode.net:sendak.freenode.net has changed the topic to: 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
16:50:251JTABGVNA:1JTABGVNA is now known as nsh-
16:50:43maaku:maaku is now known as Guest64655
16:52:29otoburb:otoburb is now known as Guest2104
17:12:23kanzure_:kanzure_ is now known as kanzure
18:05:59luny`:luny` is now known as luny
19:26:54Guest39855:Guest39855 is now known as gwillen
19:50:00OneFixt_:OneFixt_ is now known as OneFixt
20:46:39mr_burdell_:mr_burdell_ is now known as mr_burdell
20:47:08mr_burdell:mr_burdell is now known as Guest29370
21:10:22BlueMatt_:BlueMatt_ is now known as BlueMatt
21:43:38Aquent:do we have any estimate of when sidechains might be impelemented?
21:55:53tacotime:two weeks
21:56:45tacotime:™
21:57:16tacotime:the last read from the core devs was that it's still in its infancy unless you want to make a federated peg sidechain, which is more or less centralized
21:57:41tacotime:the git for that is referenced in the appendix of that paper
22:07:36Aquent:that seems what reddit is looking to do
22:07:50Aquent:centrelized sidechain reddit notes
22:08:22op_mul:I don't think even reddit knows what reddit is looking to do.
22:08:30Aquent:and, u saying, they dont have to wait?
22:26:09gmaxwell:Today's concetrated laughter dose: http://blog.bettercrypto.com/?p=1008
22:28:45nsh:lol
22:38:56user7779078:ol whats that person's deal? are they just spewing nonsense?
22:45:22gmaxwell:user7779078: in this case it's halarious becuase the parameters he's pushing are widely discredited (they originate from NSA provided unjustified magic numbers).
22:46:43user7779078:gmaxwell: ah, thank you.
22:47:01gmaxwell:user7779078: in general I don't know what up with him. He makes a lot of weakly sensible very loud and over stated allegations about bitcoin. When he started up I'd wondered if he was not right in the head, but now I think he's just trying to make a job for himself as a "bitcoin critic". Too bad, because the people trying to make a job for themselves out of being critical usually also make genui
22:47:07gmaxwell:ne critics with helpful criticisms look stupid too.
22:47:44user7779078:interesting, not surprising either. someone was bound to do it