log☇︎
52100+ entries in 0.369s
a111: Logged on 2017-10-07 22:39 phf: http://btcbase.org/log/2017-10-07#1722379 << this is probably true but only because ffa mutates an array of bigits, where's any language level bignum system produces a whole new one for each operation
a111: Logged on 2017-10-07 21:53 apeloyee: the primorial has to be, say, 2^32 times less than the ffa maxint. then you can add randomnumber*primorial, and such a number is equally likely to any prime from some interval
asciilifeform: the ONLY correct method of generating cryptoprimes, is to 1) get N bits from FUCKGOATS 2) determine, in fixed spacetime every single time, whether that string of bits constitutes a usable prime. ☟︎☟︎
a111: Logged on 2017-10-07 21:48 apeloyee: http://btcbase.org/log/2017-10-05#1721485 << alternatively, can *construct* numbers which don't have very small factors. pick a nonzero remainder mod 2, mod 3, ... mod largest-prime-fit-in-your-primorial and find what number of primorial is congruent to it using chinese remainder theorem
asciilifeform: http://btcbase.org/log/2017-10-07#1722402 << this is a fundamentally wrong way to generate cryptographic primes. we had a thread about it, http://btcbase.org/log/2017-08-14#1697562 ☝︎☝︎☟︎
a111: Logged on 2017-10-07 21:28 apeloyee: http://btcbase.org/log/2017-10-05#1721485 << i thought bernstein's "how to find smooth parts of integers" suggests a remainder tree, not gcd?
a111: Logged on 2017-10-07 21:25 apeloyee: the multiply-by-approximate quotient in barrett's also needs only the lower part (plus 2 extra bits to the left), and lower part of product can be computed exactly (since rounding is not a problem)
a111: Logged on 2017-10-07 21:09 apeloyee: asciilifeform: turns out a simple, ffa-suitable O(N^2) algorithm exists for GCD. This is adapted from GMP docs with one extra operation in the loop: http://p.bvulpes.com/pastes/oupUJ/?raw=true . Note: the code as posted is likely wrong, but I'm sure the idea can be made to work.
a111: Logged on 2017-10-07 21:53 apeloyee: the primorial has to be, say, 2^32 times less than the ffa maxint. then you can add randomnumber*primorial, and such a number is equally likely to any prime from some interval
mircea_popescu: http://btcbase.org/log/2017-10-07#1722405 << this may actually be a better check than any miller-rabin, and at any rate a good complement. gcd with primorial. ☝︎☟︎☟︎
BingoBoingo: Trilema re-read of the day http://trilema.com/2014/how-i-was-wrong-cuckolding-or-a-story-about-sigmas/
phf: a whole new bignum that is
phf: http://btcbase.org/log/2017-10-07#1722379 << this is probably true but only because ffa mutates an array of bigits, where's any language level bignum system produces a whole new one for each operation ☝︎☟︎
ben_vulpes: meanwhile, found a 20160728.tar.bz2
apeloyee: the primorial has to be, say, 2^32 times less than the ffa maxint. then you can add randomnumber*primorial, and such a number is equally likely to any prime from some interval ☟︎☟︎
apeloyee: http://btcbase.org/log/2017-10-05#1721485 << alternatively, can *construct* numbers which don't have very small factors. pick a nonzero remainder mod 2, mod 3, ... mod largest-prime-fit-in-your-primorial and find what number of primorial is congruent to it using chinese remainder theorem ☝︎☟︎☟︎☟︎☟︎☟︎
apeloyee: http://btcbase.org/log/2017-10-05#1721485 << i thought bernstein's "how to find smooth parts of integers" suggests a remainder tree, not gcd? ☝︎☟︎☟︎
apeloyee: the multiply-by-approximate quotient in barrett's also needs only the lower part (plus 2 extra bits to the left), and lower part of product can be computed exactly (since rounding is not a problem) ☟︎☟︎
apeloyee: asciilifeform: turns out a simple, ffa-suitable O(N^2) algorithm exists for GCD. This is adapted from GMP docs with one extra operation in the loop: http://p.bvulpes.com/pastes/oupUJ/?raw=true . Note: the code as posted is likely wrong, but I'm sure the idea can be made to work. ☟︎
asciilifeform: so a word mul is actually five MULs
a111: Logged on 2017-10-07 16:26 phf: asciilifeform: wait, that seems like a cheap sleight of hand. obviously increasing number of iterations in an iterative algorithm that you gave is going to increase run time
a111: Logged on 2017-10-07 16:49 mircea_popescu: my guess is that it's as close to closed form solutions as possible, hence all the barrett fucking etc, but then again i'm a weak programmer and a very dubious mathematician.
mircea_popescu: my guess is that it's as close to closed form solutions as possible, hence all the barrett fucking etc, but then again i'm a weak programmer and a very dubious mathematician. ☟︎
mircea_popescu: a
phf: i suspect that ffa's take on expmod is to iterate over every bigit of the exponent, which will have to perform base operations no matter what the numeric size is, but that's a guess.
phf: well, it's conveniently two strategies: closed form solutions and constant iterators. if you don't have a closed form solution, you have to iterate, which you simply do at the upper bound constraint by a data type size. i don't see how theoretically it can be anything else
phf: mircea_popescu: well he either has a constant time algorithm in ffa, in which case if the goal is to compare speed specifically we should be comparing fixtime ffa and fixtime something else. otherwise he has a variable time algorithm running at worst case constant time, in which case the comparison is between base operation speed, which is still going to come out on top
mircea_popescu: phf his point is that if you're going to compare fixtime with something else, better make sure you get a long case in there too.
phf: asciilifeform: wait, that seems like a cheap sleight of hand. obviously increasing number of iterations in an iterative algorithm that you gave is going to increase run time ☟︎
mod6: and same version of py there too. ok just a sec.
mod6: <+asciilifeform> out of curiosity, how long the py item takes on mod6's box ? << was just saving... lemme give it a try here. want me to try it on the i5/8gb box ?
asciilifeform: phun phakt, this calculation is taken from the gpg autopsies last summer, when asciilifeform was chasing imaginary rng boojum after somebody found a real one
asciilifeform: this is even though python uses a c bignumatron internally.
mod6: yeah, something simple like this is a good starting spot.
mod6: mainly, I read through them. because, there's still a lot for me to grok here. and it's easy to fool oneself into groking if you treat it like a blackbox instead of actually reading the code.
asciilifeform: currently i generate them with a pyturd
a111: Logged on 2017-10-02 19:31 asciilifeform: note also that the calling style from early versions will not work, there is no longer a .Z , FZ is not a struct any moar, it is just a word array
asciilifeform: ( it will need a small adjustment in re http://btcbase.org/log/2017-10-02#1719728 but otherwise oughta work ) ☝︎
mod6: i think ima make a quick one for myself just so i can see what youre sayin on stuff like that.
mod6: btw, do you have a simple test harness setup for this just to assert some known output values?
asciilifeform: ... could even live with this, if i had a hard proof that it's never moar than 3.
asciilifeform: '...I think a formal apology should be handed out, and the teacher involved should be reprimanded,” he added. '
asciilifeform: '“the year is 1935 and you have been tasked with creating a mascot to represent the Nazi party at its political rallies.” “Think about all of the information you have learned about Hitler and the Nazi party,” the assignment directed. “You will create a COLORFUL illustration of the mascot. Give the mascot a NAME. You will also write an explanation as to why the mascot was chosen to represent the Nazi party.”'
mod6: 2.107 shows extended euclid that yields greatest common divisor d of two integers a and b, but also integers x and y satisfying ax + by = d
mod6: 2.103: FACT If a and b are positive integers with a > b, then gcd(a, b) = gcd(b, a mod b)
mod6: while b /= 0: r <-- a mod b, a <-- b, b <-- r. return (a).
asciilifeform: we're prolly stuck with a muxed euclid
mod6: yeah, i read the thread a few times.
mod6: her haggard old bag of a mom didn't approve? or did i dream this up?
mod6: especially if it rains a lot too.
mircea_popescu: used to. i suppose this place being ~a village doesn't help.
mircea_popescu: mod6 anyway, i'm not a young man anymore. i maybe do 10km a week these days, if that.
mircea_popescu: asciilifeform consider : i've yet to throw out a pair of shoes because ~broken~.
asciilifeform: recently asciilifeform bought a whole case of identical chinesium shits, because they're ~disposable
mod6: ah, that's a deal!
mod6: certainly a lot better than the $100 shitters.
mod6: i spend $400 on mine. so yah, not cheap... but shit, what a difference.
mircea_popescu: they're nice though, a brazillian take on the longwing brogue
mod6: get yourself a pair of shoes. it'll change your whole outlook on shit
mircea_popescu: http://www.anatomicgel.com.br/br/marca.html/ <<< i suppose it could be theoretically called handmade, pushing a point.
asciilifeform regularly runs into this species, 'i'm into btc!!' '...node?' 'wat's a node'
mircea_popescu: which one ? iirc he had a whole herd.
asciilifeform: as for 'does arson count', if d00d had just a megaphone and persuaded 500 derps to drink bleach, imho also counts.
mircea_popescu: or a block of cinderwood.
mircea_popescu: only been a day, these days medicare can keep even tom petty alive a day.
asciilifeform: these folx trample one another regularly, there is even a commercial 'holiday' ( 'black friday' ) when they have india-like mass tramplings fighting over misc. retail crud
asciilifeform: problem is that 60 is fewer than die of drink in baltimoristan in a week
a111: Logged on 2015-08-19 23:55 mircea_popescu: what, "i'm a boy from tenesee here to die for some fat bitch's right to marry her dog" ?
mircea_popescu: what is that, smile from within a basket, torso aside ?
mircea_popescu: problem is -- nobody in current generation has what to retire on. so im guessing dimon will get a job to pay for their 401ks or w/e they need.
mircea_popescu: buying jpm is a bad strategy about half the time, and bitcoin overperforms dimon by about 1000% EACH SINGLE YEAR
mircea_popescu: i wish to fucking know, at what point has buying bitcoin been a bad strategy ?
mircea_popescu: "Will bitcoin ever be a safe investment or always a gamble? - The boss of JP Morgan was unequivocal about bitcoin at a recent conference in New York: the digital currency was only fit for drug dealers"
asciilifeform: as if d00d with 100k in the bank ( per the most recent link, he wired it to some pinoy chick ) couldn't get a proper kalash wherever
asciilifeform: mircea_popescu: they've switched back to the 2010 tack of 'it's because he too easily found a spring to attach to his stock' etc
mircea_popescu: seems a good indication of merit, that they've put the effort into persuading the original poltroons into hiding it.
asciilifeform: ( that last one doesn't stand on a hash )
mircea_popescu: anyway, if anyone can cough up a manner to evaluate, any particular scheme, or even a comparison of a pair, i'm of course all ears.
a111: Logged on 2017-09-20 19:28 asciilifeform: incidentally iirc we did the proof of 'if there is a good hash, there is a good blockcipher, and vice-versa'
mircea_popescu: basically the scheme is, you rsa a random bitfield, then you expand that into as much otp as you want by doing recursively Fi = hash(bitfield + Fi-1). there's a limit on i, obviously, which can be set to 1. ☟︎☟︎
mircea_popescu: "In case you did, i hope this will be a lesson to you. Stop trading obscure tokens/coins!" << if THIS is "the lesson", then one wonders wtf is supposed to be the "ethereum value proposition" ?
mircea_popescu: i may opt to include a pubkey. but i don't have to.
asciilifeform: yer trying to not be spoofed, rather than hide under a rock.
asciilifeform: all i got is a guarantee that the arithmetic happens as specified, every time, using a number of cycles known in advance.
asciilifeform: sadly, rsa per se dun come with a guarantee.
diana_coman: obv atm nobody can actually give any sort of guarantee anyway so it's more of a model /ideal thing
diana_coman: asciilifeform, eulora needs a client-server communication protocol that at least allows client to decide for self the compromise between cost and some degree of confidentiality+integrity for the info it exchanges
asciilifeform: 2s modexp, btw, is a ~10-20min keygen time.
asciilifeform: ( asciilifeform makes a general-purpose numbertheoretical item, broader than 'here is modexp' )
asciilifeform: incidentally ~95% of the work ffa does in modexp, now, is multiplication. which means that there is further 20-25% speedup waiting to be had when i get bernsteinian optimization for karatsuba ( haven't yet figured it out, he buried it deep in a paper , as if he were an alchemist, quite cryptically ) and another 10-20% optimization if we move to unrolled comba ( see august thread. )
asciilifeform: so right now every 8192b mod 4096b op ( of which 4096b modexp has 4096*2==8192 ) costs 2 8192b*8192b -> 16384 mul's and a few subs. which still beats the living shit out of knuthian div.
asciilifeform: ( not to mention did not yet even try to use apeloyee's secret-normalization method : instead we use a completely unshifted dividend , and cut it in the end)
mircea_popescu: and in other random dribbles : i was going by teh sabana park (huge park here), noticed they had cut down some rainbow eucaliptus trees. so i had car stopped and... got a chunk for the trunk.
BingoBoingo: Allowing a tort against the car with such a massive and sclerotic mitigating circumstance is insanity.
BingoBoingo: If morbidly obese and has coronary while excited, cause of excitement is not material. Could have just as well been seeing a pallet of "Count Chocula" "breakfast cereal" causing terminal excitement.
mircea_popescu: BingoBoingo "coronary was BECAUSE SHE WAS FAT, and NOT because she was an excited fat person someone drove a car into" ?
BingoBoingo: In other ???, Heather Heyer, the fat Charlottesville "victim" apparently died of a heart attack and not Dodge Challenger induced injuries http://www.vdare.com/articles/anarcho-tyranny-update-mounting-proof-that-the-charlottesville-five-are-political-prisoners
mircea_popescu: meanwhile in "for the end user, ux is not a part of your product -- ux IS THE PRODUCT!!" lulz, https://thehackernews.com/2017/10/macos-high-sierra-apfs-password.html
asciilifeform: BingoBoingo: if there existed a sane crowned head we would probably never have this thread to begin with
mircea_popescu: i think they eat up almost a gb of trilema