log☇︎
336000+ entries in 0.212s
mircea_popescu: 69 begs to differ.
mircea_popescu: in the same way, we also go to bed with the empress-consort of china every night.
mircea_popescu: "multiverses" and all that.
mircea_popescu: i suppose h. everett is the foundational lemma of sf.
asciilifeform: (for n00bz: pistol is connected to a recognizer box, shoots you in the head in all universes where rng did not shit out the solution)
asciilifeform: (most elementarily - tricks involving time travel)
mircea_popescu: for the obvious reasons.
mircea_popescu: but yes, expect all such items to be buried in well - curated burial pits.
mircea_popescu: i originally gave it as an example to stand pro toto for a larger set. specifically - that hunting the p = np thing is a waste of time, when there are in point of fact items known to be harder than np.
asciilifeform: if this, it rose and fell with the career of s. cook.
mircea_popescu: actually : the cost is shown to be double-exponential. fischer & rabin 1974.
mircea_popescu: no different bloom, this.
asciilifeform: ah, i was thinking of Schönfinkel et al
mircea_popescu: the why and wherefore be left for students from the future.
mircea_popescu: just like it - ALSO disappeared with virtually no trace.
mircea_popescu: as a curio - this was a very promising field just about the time of mit ai being a very promising field
mircea_popescu: however - that decidability has a cost, and the cost is > exponential.
asciilifeform: i wil eat six airplanes before i go through this proof.
mircea_popescu: anyway, get his guy stuff from library, if can't be web'd (i spent a little trying to dig that one up).
asciilifeform: 'Sorry, your request could not be processed because the format of the URL was incorrect. Contact the Help Desk if the problem persists. [SD-001]' << l0l
mircea_popescu: basically showing that a+b < c is true or false for a, b, c in R is a harder-than-NP problem. ☟︎
asciilifeform: expand on this ?
asciilifeform: berman's theorem
mircea_popescu: did i say that somewhere ?
asciilifeform: is NP-complete then P = NP'
mircea_popescu: jurov doesn't have to be invented in crypto to be fundamental for crypto
jurov: not sure it's even a crypto invention. noisy channel theory says basically the same (think channel with 50% bit flip noise)
mircea_popescu looks forward to a time when say berman's is a fundamental theorem in crypto
mircea_popescu: that's actually true huh
asciilifeform: xor lemma is probably the closest thing we have to a fundamental theorem in crypto
mircea_popescu: yes but i'm not in the business of carpentry.
asciilifeform: for whatever timber - arithmetic is the straightest saw we've got.
mircea_popescu: just like that, of the crooked timber and all that.
asciilifeform: how's that
mircea_popescu: eh there is no grasping fully.
asciilifeform: and in the process, grasp it fully...
asciilifeform: incidentally, imho mircea_popescu could write a splendid article about the xor lemma.
asciilifeform: one demonstrable benefit is that it multiplies the cost of brute force.
mircea_popescu: so i have to leave pc overnight / rng on for ten minutes to make a key ? big whoop, i make a key a year if that, and i'm an intensive user.
mircea_popescu: quite. now 1mb IS probably too large to be useful (unless actual mathematical need/benefit for it is shown)
mircea_popescu: getting rid of one of the most idiotic constraints people-who-are-not-me put on cipher design
asciilifeform: ah this - yes
mircea_popescu: merely, that the "key over the wire in the jungle" aspect need not be accounted for.
asciilifeform: this is harmless, but to rely on the enemy having difficulty getting your pubkey is lunacy
mircea_popescu: just like with chicks and their phone numbers.
mircea_popescu: (for the future, the ~ONLY way to get a key would be to ask trusted third party for it. such as hey x you know y ? send me his key!)
mircea_popescu: similarly with the key. so large key can't be sent over the wire. bwahahahaha mkay.
mircea_popescu: if the increase is 100x this impacts nothing
asciilifeform: plus whatever overhead from the individual cipherators.
mircea_popescu: look at the usual case : we exchange 1kb messages.
asciilifeform: you get 2x the ciphertext bitz.
asciilifeform: what this gambit ~does~ cost, is weight
mircea_popescu: i misread the split.
asciilifeform: if i give you one half of a xor pair, it tells you ~nothing~ about the plaintext...
asciilifeform: so hard to grasp ?
asciilifeform: again the motherfucking xor lemma
mircea_popescu: there are some schemes which, if the result is split, lose strength
asciilifeform: he cannot use knownplaintextisms created by one, to help in the other
asciilifeform: ^ in this case, enemy actually must break both ciphers ~independently~
asciilifeform: (incidentally, the problem of an inner cipher introducing known-plaintextisms is solved routinely by splitting the payload into xor-able halves, using rng, and enciphering each ~half~ with different cipher, rather than box-in-box composition) ☟︎
asciilifeform: the paper is pretty tame, actually
mircea_popescu: i for instance had never read it / wouldn't care to now.
mircea_popescu: i dunno that that particular item informs anything.
asciilifeform: but i'm blown away by the maurer & massey thing
mircea_popescu: "Since most conventional block ciphers are in fact product ciphers, they testify to the worth of multiple encryption, even with weak internal ciphering functions."
asciilifeform: scroll to 'deceptive article'
assbot: Ritter's Crypto Glossary and Dictionary of Technical Cryptography ... ( http://bit.ly/1Q9fzCq )
asciilifeform: rate is zero across currencies and issue Bitcoin 2.0 based on incoming bitcoin and signatures made by participating addresses. This means the issuance honors mining until the end of bitcoin and can be more than total bitcoin present. I suspect 1 BTC 2.0 will have the spending power of 1 USD current by issuance. My pedigree.'
asciilifeform: 'I have been involved in Bitcoin since before the "Mt. Gox Fast Pay" thread that produced Bitinstant. My law cases make me the bearer of Bitcoin 2.0 no matter what. Bitcoin 2.0 is simply my daemon forked from a sued Ripple Labs qualified as an invention with features added and the Bitcoin upgrade address 1J7CAm6bavGcG52xgobn8n9ScnAzH8JZUm - I have pledged to drain the exchanges until exchange
asciilifeform: just 1 of these is enough to stay in roquefort and orc-free street
mircea_popescu: fiat money tho, not terribly useful.
mircea_popescu: sed the hirethisguy.
asciilifeform: 'hire this guy'
assbot: Want Your Own Version of Bitcoin? Hire This Guy - Bloomberg Business ... ( http://bit.ly/1WeCNp9 )
BingoBoingo: ;;later tell danielpbarron improving
asciilifeform: i can even see the dirigle algebra involved here !
mircea_popescu: check out how complicated things get.
mircea_popescu: so as you can no doubt follow, asciilifeform from the political considerations discussed it follows that while i don't allow for the possibility of "quantum computer", i don't disallow your implementing a scheme which would be deemed (by idiots) as "quantum resistant".
asciilifeform: eliminate things enemy may rely on.
asciilifeform: this is really a specificity-of-diddling theorem application.
mircea_popescu: (this reduces to what the idiots do with their s boxes incidentally)
mircea_popescu: branch it out like a tree.
mircea_popescu: yes. always large block to small ; always few to many options.
asciilifeform: incidentally, to pick up a thread from last night, one useful lemma re: chaining is,
mircea_popescu: the two are much closelier related than people generally wish to accept.
asciilifeform: srsly, if there is one mega-lesson mircea_popescu taught me, it is that by DEMANDING sanity, you will get it
mircea_popescu: alowing the possibility that this scam is actually worth taking seriously costs the people who follow what i say whatever it may cost them. ☟︎
mircea_popescu: asciilifeform things may cost you nothing, that is your privilege. things always cost me something, because of my peculiar relationship with the fucking future.
mircea_popescu: even going under that : afaik not yet usable in crypto
asciilifeform: back to earlier's thread, i am willing to pascal's wager on shor's algo SO LONG AS it costs me nothing.
asciilifeform: it is provably not doable for arbitrary turing tape.
mircea_popescu: re other thread : the only point of interest in lattice stuff would be homomorphisms. which don't work yet.
mircea_popescu: exact equivalent of the anabolic/catabolic processes in the human body. if there aren't people falling over in the street out of sheer hunger as an unremarkable phenomenon, there's no industrialization afoot. if there aren't any "self-betterment" dudes moving about, there's no dezindustrialization occurring.
mircea_popescu: industrialization is driven on the backs of starving slave/iobagi, and deindustrialization is driven by lazy, entitled idiots.
mircea_popescu: what color lipstick you put on that pig - vanitas vanitatum.
mircea_popescu: neither fervor nor bayonet. "people", if you can call english speakers that, will industrialize as they do : out of pompous laziness. that's how it always goes.
asciilifeform: plus the whole climatology thing was a misguided excercise - the reich for some reason thought people would happily deindustrialize and starve out of religious fervor, but then understood that bayonet is good enough
mircea_popescu: get with the motherfucking program, bunch of snake oil salesmen.
mircea_popescu: "Lattice-based cryptography is another important candidate for post-quantum public-key cryptography, and is the main topic of this blog post. The main practical reason to consider lattice-based encryption rather than code-based encryption is that lattice-based systems advertise much smaller public keys." << WRONG.
mircea_popescu: i expect there's going to be a "climatology winter" too starting next year, as opposed to "we lied for pay and after that the employer had no further use of lying pseudoscientists".