asciilifeform digs for ancient report of zoologies who lost fingers, toes, nose, ears, eyes, cock to chimps in one evening. beasts he worked with for a decade, who decided to have some sudden fun.
asciilifeform did not personally measure chimp strength, cannot guarantee results.
asciilifeform: so, to him we're sorta like plasticine.
asciilifeform: mircea_popescu: nontrivial - chimp has >2x the strength of a man gram for gram (muscle fiber quirk. they get greater synchronicity in fiber motion in place of our precision.)
asciilifeform vaguely recalls the zoology lesson - if condemned to choose what animal to be locked with unarmed, and choices are chimp and, e.g., grizzly, or lion - pick the mega-carnivore. at least quick death.
asciilifeform: diametric: an adequate trng is incredibly easy to build. but don't use adc noise on a general-purpose micro - is heavy on quantization bias, and easy to nudge from outside.☟︎
asciilifeform: diametric: likewise, no provision for powerline discipline.
asciilifeform: diametric: FST-01 << useful for many purposes, but not crypto. on account of being reflashable via usb.☟︎
asciilifeform: but 'printing' is a uniquely inefficient way to shape plastic. it is justified for prototype, usually - would not want to manufacture en masse this way.
asciilifeform: and actually represent, e.g, 1/3 as a 1, a /, and a 3; sqrt(2) as a root and a two of it, etc
asciilifeform: mircea_popescu: actually plenty of folks use 'rational arithmetic' systems that don't touch the ieee/flop cpu circuits
asciilifeform is actually rather fond of analogue computing as a concept, and as it once existed, but has yet to meet a 'meatware' proposal that isn't transparently grantsmanship.
asciilifeform: jurov: turing's apparatus is not even a particularly good abstraction for electric computer. much less 'wet' kind.
asciilifeform: takes about a year of skilled labour.
asciilifeform: now, getting the answer out in a usable form...
asciilifeform: why even bring this up: the 'answer' - physical structure of properly folded protein - is 'computed' instantly.
asciilifeform: now, think about why (2) costs roughly same as (1).
asciilifeform: (2) gives a result more in line with reality, and so tends to be favoured unless there is something exotic about the system (weird ligand, perverse conditions otherwise, whatever)
asciilifeform: to every n00b's surprise - the cost is roughly similar.
asciilifeform: one - approximate solution of shrodinger's equation for the system of protein-cum-solution
asciilifeform: so, for those who don't know, there are two basic ways of 'folding' a protein (that is, in physical reality it folds, but you want to know its structure)
