asciilifeform: apparently mircea_popescu dun buy the 3/4 proof ?
asciilifeform: say you have n for which the entire bottom quarter of the 2048bit witness space is liar. how does this prevent working rng from still finding working witness in the expected # of shots ?
asciilifeform: how does 'contiguous set of liars' play into scenario with working rng ?
asciilifeform: even 'factoring is hard' -- unproven, etc
asciilifeform: the 1 device which dun rely on unprovens is... otp ( insert oblig old thrd here !11 )
asciilifeform: correct, the bound presumes a flat-spectrum rng.
asciilifeform: this type of failure hinges on imperfection of rng, rather than hidden boojum in m-r
asciilifeform: the m-r proof is unambiguous re the number line thing. i.e. the p of n ~actually random~ witnesses all lying, is bounded.
asciilifeform: it's sorta like the proposition of hiding 4 people in phone booth
asciilifeform: to the point that i'm at a loss to construct a crackpot hypothesis for the negative ( what would the loch ness monster here look like ? erry composite n, we know has 3+ / 4 of integers as proper witnesses. so where wouldja hide'em so that working rng doesn't find 1 in 32 shots before asteroid hits machine ? )
asciilifeform: all we have is the http://btcbase.org/log/2019-03-28#1905286 ( from elementary proof ) + the observation that nobody ( or at least not asciilifeform ) has ever found a composite that doesn't properly light up m-r 'composite!' indicator for 3+ / 4 rng stabs.☝︎
asciilifeform: when we 1st had m-r thread, i also considered a hybrid algo, where you take e.g. 32 rng witnesses, and 32 that are kept in bottle and known only to you , for 64-shot test that is slightly moar immune to rng failure. but then thought 'rng is jesus bolt, if fails, yer candidate is also fucked' so couldn't think of why to do such a thing.
asciilifeform: whereas if you actually lift 32+ rng witnesses from a working rng (as in asciilifeform's demo, or diana_coman's proggy, and elsewhere where not koch.. ) actually converges (for so long as you actually have working rng)
asciilifeform: mircea_popescu: possibly i'm thick, but what does b win ? seems like any hardcoded list, if becomes known to enemy, opens you up to theoretical 'bake a n for which the witnesses are liars' item
asciilifeform: afaik that was the last word on the tech.
asciilifeform: there was a sovok train engine with no boiler . for last-mile into factory with combustibles. filled with steam, off-site, went for ~40min after.
asciilifeform: compressed gas aint so great as energy store, you lose just about errything you put in , to pv = nRT
asciilifeform: airtank only good for vertical. battery if you care to also move..
asciilifeform: mircea_popescu: 500 m. for the typical
asciilifeform: ( seems extravagant, until you remember that you can't submerge on diesels )
asciilifeform: was thought, if you have 160 hands on the ship, may as well also battery
asciilifeform: funnily enuff, sov nuke vessels still had'em. backup to diesels which in turn backup to the 2 reactors
asciilifeform pictures castle mircea_popescustein with uboat-style battery room, what with the little cart that one rides lying on back to test electrolyte etc
asciilifeform: i.e. what actual hamster could turn out in his wheel..
asciilifeform: this is 1 of the wins of rk, thing can live on 3W
asciilifeform dun like to have moar microwave background than strictly must
asciilifeform: i have the latter, but it's kept in cold reserve, rather than auto
asciilifeform does not have diesel, thought about it but ~0 point, given that upstream fiber people seem to only have ~2hr of battery, so went 'i'ma also 2h..'
asciilifeform: a healthy biped can put out 100+ watt , sustained, on bike
asciilifeform: so conceivably you dun need a gigantic hamster wheel, could rewire ordinary bike thing
asciilifeform: imho mega-waste, oughta come with plug
asciilifeform: mircea_popescu: funnily enuff, even cheapo current gen. of gym stationary bikes, actually include gens, they simply dump the current in resistor
asciilifeform: wainot, if they already turn wheels at gym..
asciilifeform: mircea_popescu: generating cage ? i.e. with hupet in large wheel turning gen ?
asciilifeform: ( for 'honest' prime gen, where there's actually 2 primes of the traditional bitness, this ~never happens. but for others.. )
asciilifeform: incidentally, and for thread-completeness, gotta add, for certain patterns of failed m-r sequence , you end up with output that gets you a factor. this item is actually on the phuctor conveyor, when i get a free hand to crank it again
asciilifeform: it's what i distinctly recall from that piece, will revisit when i dredge up a copy
asciilifeform: all of this , as i understand, only is interesting if enemy knows (or can set) which will be your witnesses. cuz by basic m-r , 3/4 of possible inputs for witness will still give correct diagnosis of compositude
asciilifeform frustrated, could've sworn he saw it with own eyes at some pt
asciilifeform: iirc it is how the linked item worked
asciilifeform: ( even moar interesting, would be to find litmus for some variant of this being in use, then could point it at phuctor collection & pull trigger )
asciilifeform: what'd be interesting is to find algo that bakes a large composite n , and shits forth arbitrary # of 'plausible-looking' large but false witnesses
