log☇︎
312000+ entries in 2.291s
mircea_popescu: now europe is going off a cliff, and the laggard is in the best position all things considering.
ThickAsThieves: includes a whore, go
mircea_popescu: romania is one of the eu members. it has always been a laggard. this was always decried as a huge disadvantage.
mircea_popescu: the idea is just that having a car that's 1ms faster wins you nasdaq, but on a well designed exchange having a bot that's 1ms faster wins you little if anything at all.
asciilifeform: (whenever I read the word "goxlag" I can't help but imagine a future Gulag for BTC users...)
mircea_popescu: there's no real advantage to being quick in a game of wits.
ThickAsThieves: i guess i'm a hypocrite tho
ThickAsThieves: i just want the market to be a market
asciilifeform: everybody says that bots/HFT are a fact of life, but what if you have to solve a Sokoban (or other AI-complete nonsense) to move a coin...
asciilifeform: Diablo-D3: the wronged party needs to be able to deal damage to a scammer without benefiting himself.
asciilifeform: Diablo-D3: a Shitcoin that can result in moving coins is a terrible idea. Moral hazard for bond-holders.
asciilifeform: Scrat: see "CAPTCHA" as a (somewhat flawed) example.
Diablo-D3: I can sign a sharecoin using the scripting stuff
Diablo-D3: say, shares of a company
Diablo-D3: but the countersignature can ALSO be a normal tx
asciilifeform: jurov: the idea is that you need to pick a work function which remains hard even when a million people (some of whom with arbitrary pull) want it to be easy.
Diablo-D3: send a tx that contains your signature like normal, but contains a list of addresses who can countersign it
jurov: encode a block asenglish text tobe submitted into peer-reviewed scientific journal
asciilifeform: F is a trapdoor function.
asciilifeform: Diablo-D3: the idea is that a user has to sign a message with his key: "This coin can be shat on in the future, by anyone who can provide x such that F(x) = somenumber." He gives x to the bond holder(s).
Diablo-D3: which isnt a big deal
jurov: sorry for not reading it all but have a proposal for slavecoin
Diablo-D3: asciilifeform: we'd need a new tx format
Diablo-D3: where I send a tx to the network that says "these coins are shitcoined for the following users" and then list their public keys
Diablo-D3: asciilifeform: we'd need a limited form of this
asciilifeform: Diablo-D3: the only coins you can shit on are those for which a Shitcoin bond has been issued to you. but it is true that you could do so maliciously. hence a certain amount of dirt would be seen as commonplace.
exfortuna: Thank you, but is there a way to get raw data? I'm doing a regression analysis for an econ class
Diablo-D3: and then people can coutersign your coins if you're a dick?
deadweasel: when frying a skinny fish, do to poke at it too much, this spoils the fish.
Scrat: Diablo-D3: a coin with coloring built in
Diablo-D3: thestringpuller: I request a new analogy
thestringpuller: that's like arguing with a vegan about honey
Diablo-D3: WHO THE FUCK THOUGHT THREADS WAS _EVER_ SUCH A GOOD IDEA
Scrat: but to be fair such a scheme will in no way be a USP for a new coin
Scrat: well i guess it does make a bit of sense, just 2-3% of overhead for a great increase in availability
Diablo-D3: send 195kb for a 192kb block
Scrat: ditch the OS/lang idea imo and make a new coin
Diablo-D3: I send in total 256kb for a 192kb
Diablo-D3: so if I need 3 or 4 of a 128kb block, 3 RS chunks will be 128kb
Diablo-D3: Scrat: a properly done RS setup of n of m, the size of n chunks is going to be at least the size of the original
Diablo-D3: but it has a lot of shit on top to handle error correction
Scrat: well I need to think about it but it looks as if you're gonna need to have a lot of RS redundancy to make that usable
Diablo-D3: and then that finishes, and I send chunk C to peer 3, and while Im doing that, peer 1 sends A to another n (where n includes 2 and 3), peer 2 sends B to n (which includes 1 and 3)
Diablo-D3: I send chunk A to peer 1, that completes, then I send chunk B to peer 2 and while Im sending it to 2, peer 1 is sending A to peer n
pgp: wow a lot of coin just printed...
Scrat: so you compress it and get to 60% of the size, then a healthy application of reed solomon codes will give you another 40%
mircea_popescu: proof of work could be then accomplished by showing a path of solving the map in at most N steps, where n is the difficulty.
mircea_popescu: suppose we find a liniarisation of sokoban maps, such that any arbitrary number comes to a map
asciilifeform: mircea_popescu: there are a great many interesting hard problems, and life is short.
asciilifeform: mircea_popescu: I might, if I lacked anything else to do. Personally I'm not sitting on a stable of slaves waiting to be Sokobanned.
Diablo-D3: mircea_popescu: but a lot of people have already built mining farms in their basements
asciilifeform: what remains is to work it into a workable scheme for a hash chain
asciilifeform: mircea_popescu: here's a starting point: http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.52.41
mircea_popescu: cpus have a large mkt already
Diablo-D3: the cost of an ASIC with 1TB of ram is going to be the same as a cheap CPU with 1TB of ram
asciilifeform: the only long-term solution is a work function that is "AI-complete." Hence SlaveCoin.
Diablo-D3: and asics + ram isnt a winning solution
asciilifeform: the fun part in all of this is that "for every wily arse, somewhere there is a cock with threads." Eventually someone builds an optical computer operating on standing-wave physics, that pwns every new hashing function you come up with the day it comes out.
asciilifeform: Diablo-D3: all you need is a high-speed interconnect that lets you connect arbitrarily more of whatever the necessary resource is (RAM, hashing power, etc.)
mircea_popescu: floating -M may be a great idea.
asciilifeform: right now you can put the BTC world in your pocket for less than the cost of a jumbo jet.
Diablo-D3: mircea_popescu: shush for a sec
mircea_popescu: we had 100k nodes and a bunch of miners.
mircea_popescu: bitcoin was a LOT stronger back when ppl cpu minded it
asciilifeform: Diablo-D3: but raising the cost of attack is a loser's game. eventually someone will pony up the cost, and rule.
mircea_popescu: this is a lost bet.
mircea_popescu: but scrat has a point
asciilifeform: Diablo-D3: a PC with 1TB of RAM will still be beaten by an ASIC connected to the same 1TB of RAM.
Diablo-D3: like, I can buy a machine that has 1TB of memory
asciilifeform: think about the endgame in the ASIC trend. eventually, only powerful industrial nations will matter in the BTC game, because they can afford to build fabs. Just a handful each, even. And we're back to the central-baking clusterfuck we see with fiat.
Diablo-D3: a key derivation function
mircea_popescu: a ok
asciilifeform: mircea_popescu: the slaves need a certain minimum IQ, for Sokoban.
mircea_popescu: asciilifeform there's more people owning slaves than semiconductor fabs. this is a fact.
asciilifeform: all of the Bitcoin-like altchains are doomed, because the first serious guy with a semiconductor fab will own the world. SlaveCoin is the future!
Diablo-D3: you know how blocks are named by a 256 bit value?
mircea_popescu: you'd have to keep a bunch of vats somewhere which'd be the blockchain
asciilifeform: Diablo-D3: at this point, an alt-chain is guilty until proven innocent of being a clever pre-mining scheme.
Diablo-D3: we have what equates to a giant torrent network
Diablo-D3: theres a lot of interesting ideas I want to try out on making chain transmission faster
mircea_popescu: a full re-doing of satoshi's stuff would be great even if only an exercise.
mircea_popescu: this is true, he has a point there.
Diablo-D3: mircea_popescu: there will be a spec, yes.
mircea_popescu: Diablo-D3 if you actually take satoshi's stuff, write A SPEC and then have some implementations
asciilifeform: Diablo-D3: LTC ~= 1.3 USD, as of a few seconds ago. Not exactly gum wrappers yet.
asciilifeform: a scenario where there are multiple blockchains, but automated exchanges connect at least the well-behaved ones, is in some ways equivalent to having one blockchain.
mircea_popescu: the roman empire coinage only had much relevance in those places 10 miles or closer to a castrum
asciilifeform: mircea_popescu: without ionosphere bounce (shortwave, <30MHz) or something like MBC you will have a very short usable distance.
mircea_popescu: anyway. why not just make a bunch of normal ham radio stations ?
mircea_popescu: maybe a kb/sec but most trails don't last a second
asciilifeform: mircea_popescu: a few kb/sec. will suffice in the hypothetical Dark Internet-less Future and its transaction volume
asciilifeform: mircea_popescu: one thought: this could even be done for profit. Say, transmit the normal blockchain with a few hours of lag, but those who paid a transaction fee get broadcast in real time.
pgp: i like it - a bit more intuitive than a candlestick chart...
orkaa: ugh, does this thing have a tube that goes inside your dick
thestringpuller: they made a joke about that in californication
thestringpuller: I expect a blog post please.
gribble: http://www.urbandictionary.com/define.php?term=spoiled%20brat | A kid who wants everything they see... and they cry and scream thir heads off until they get it.
c1n0: i usually consult oracles and ask them to gather every different type of drug out there and make a cocktail and inject it into my eyeballs just before i make a trade so i can read the market better
ThickAsThieves: this volatility is a lil too risky for me, not riding this rollercoaster just yet
c1n0: hehe bitpay now doing a ama