asciilifeform: ben_vulpes: what's that mean

asciilifeform: http://btcbase.org/log/2017-06-09#1667638 << i contemplated whether to show ben_vulpes ~all~ of the mistakes in his proggy, or to let him find for himself, decided on the latter ( it's a healthy thing ) ☝︎

asciilifeform: ( my interest in publishing anything else in c, ever again, is ~0 )

asciilifeform: http://btcbase.org/log/2017-06-09#1667623 << believe or not -- i dun have one yet. at some point i'ma write an ada proggy ☝︎

asciilifeform: http://btcbase.org/log/2017-06-09#1667608 << ~100% of the cpu-bound operation is already offloaded to optimized multithreaded c. ☝︎

asciilifeform: http://btcbase.org/log/2017-06-09#1667603 << very peculiar question -- fella didn't say what, specifically, he proposes to do, in this involvement ☝︎

asciilifeform: every once in a while i also post, publicly, copy of the db: it's ~12G, gzipped, presently. this oughta be enough for anyone.

asciilifeform: ( pgpdump -i gives you the modulus etc. from any pgp key; then multiply the factors : instant verification. )

asciilifeform: http://btcbase.org/log/2017-06-09#1667584 << it is not, and ain't going to be, i'm not especially interested in helping folx to 'embrace and extinguish' it. as far as i'm concerned, 100% of the output is verifiable with the bignum calculator of your choice (e.g. 'bc') and it is no concern of anybody's how it was obtained. ☝︎☟︎☟︎

asciilifeform: thing is due for major rewrite, literally not 1 existing line of code oughta survive.

asciilifeform: http://btcbase.org/log/2017-06-09#1667600 << i'ma depythonize and desqlize it, but currently lack the time ☝︎

asciilifeform: ( the cruncher, the mass dump submitter, and the wwwtron )

asciilifeform: http://btcbase.org/log/2017-06-09#1667588 << i have an ad-hoc version of something quite similar, in there already. it doesn't help much, because phuctor is actually THREE independent (communicating solely via the db) unix processes ☝︎

asciilifeform: Run Moar Intel.

asciilifeform: microshit fud machine in full swing

asciilifeform: in other lullies, https://archive.is/OExoO >> nsaware using ~second~ intel fritz chip boobytrap, found ☟︎

asciilifeform: !~google flwm

asciilifeform: mod6: before ratpoison, i used 'flwm'

asciilifeform: http://git.savannah.nongnu.org/cgit/ratpoison.git/log << some of the changes are at least somewhat wtf

asciilifeform: and holyshit, there's a new release?!!

asciilifeform: mod6: http://www.nongnu.org/ratpoison

asciilifeform: mod6: ever tried 'ratpoison' ?

asciilifeform: the lead pig in which it is to spend the rest of its life, is - fortunately - also here.

asciilifeform: thing is regulation-sized , but theoretically such an item oughta have been sent by fedex. but, go figure.

asciilifeform: in other noose, the (microscopic) ra-226 ( for MB/s 'gold standard' rng test ) came. asciilifeform was half-convinced that it'd get swallowed by the post

asciilifeform: (check your mod queue)

asciilifeform: posted.

asciilifeform: aaha.

asciilifeform: ... etc.

asciilifeform: (assert (equal expected actual))

asciilifeform: (actual (invert original)))

asciilifeform: (expected #*10101010)

asciilifeform: (let* ((original #*01010101)

asciilifeform: where you

asciilifeform: or it'd appear

asciilifeform: oddly enough you knew this when writing the tests

asciilifeform: again i did not even go over whole thing, this was only the 1st bit to stick out to naked eye

asciilifeform: http://wotpaste.cascadianhacker.com/pastes/3mo8X/?raw=true << spoiler

asciilifeform: (format t "~A~%" foo))

asciilifeform: (format t "~A~%" (bit-not foo))

asciilifeform: (let ((foo #*11101010))

asciilifeform: try it yourself,

asciilifeform: but it ends up merely returning (bit-not element) which does nothing

asciilifeform: (invert s) wants a destructive op

asciilifeform: this might not be the only mistake

asciilifeform: !!

asciilifeform: ben_vulpes: bit-not is not destructive

asciilifeform: oh hey.

asciilifeform: ben_vulpes: post yours, then reconciliate?

asciilifeform: he had a pretty tall pile of sofas.

asciilifeform: they happily return same favour.

asciilifeform: ben_vulpes: post?

asciilifeform: incidentally, http://btcbase.org/log/2017-06-08#1667388 >> recall harrison, inventor of temperature-compensated chronometer, per http://btcbase.org/log/2016-03-10#1429335 ; in other contexts mircea_popescu does not even seem to find this kind of thing objectionable, e.g. http://btcbase.org/log/2015-03-06#1044115 ☝︎☝︎☝︎

asciilifeform: ^ lol, apparently was entirely other thing, long ago

asciilifeform: !#s MPHF

asciilifeform: !#s MPFHF

asciilifeform: dun seem to be in the logs

asciilifeform: ben_vulpes ?

asciilifeform: did he ??

asciilifeform: link to 1 other than mircea_popescu's php ?

asciilifeform: did it ?

asciilifeform: ( and it wants 'mphash' also.. )

asciilifeform: ( unless mircea_popescu already has )

asciilifeform: also somebody gotta come up with the tourney judge proggy itself, neh

asciilifeform: how many entries counts as 'quorum' ?

asciilifeform: lolyes

asciilifeform: (at the most recent rate)

asciilifeform: btw, lulzily, if mircea_popescu makes the prize ~11 btc, it will be moar than the eu thing

asciilifeform: neato

asciilifeform: from whence the 16 ?

asciilifeform: i'll bite :

asciilifeform: mircea_popescu's is actually the slightly more complicated one.

asciilifeform: both is exercise in breaking a gnarly and largely arbitrary prng.

asciilifeform: i dare say the pot is a little bigger, in that casino

asciilifeform: funnily enough, asciilifeform already plays a variant of this game regularly, http://btcbase.org/log/2017-05-23#1660316 ☝︎

asciilifeform: because without it, the info you get re the 'functions of the machine' is ~0

asciilifeform: seems like the 'UMM' turns the thing into an exercise in breaking 'MPHash'

asciilifeform: ( i was half-convinced that mircea_popescu did this exercise as a response to the dieharder thread )

asciilifeform: if you can 'learn' from rng -- it's a poor rng

asciilifeform: Very Realistic!1111

asciilifeform: where you cry and then die

asciilifeform: sorta like, what was it, 'depression game' that some chick wrote

asciilifeform: ( rather, poor gameplay )

asciilifeform: this imho makes for a pretty good teaching tool re reality, but poor game

asciilifeform: possibly

asciilifeform: sounds like... lottery

asciilifeform: ( or shitoshidice or whatever the current vegas is )

asciilifeform: so why would anybody play, instead of vegas

asciilifeform: reading mircea_popescu's recipe, asciilifeform begins to suspect -- but cannot yet prove -- that no better-than-chance strategy exists for this game .

asciilifeform: http://www.mazepath.com/uncleal/average.htm << oblig!1111

asciilifeform: *rather

asciilifeform: well, ratherm int8s

asciilifeform: 2).' << sounds like the params are uint8's, from this description

asciilifeform: 'This ISR describes a total of 256 linear functions, indexed in a one-byte list. Each function is of the format ai X + bi, with i taking values from 0 to 255. For each byte in the ISR : the first bit will set ai to negative if 1, and to positive otherewise ; the second bit will set ai to either 1 or 2 ; the third bid will set bi to negative if 1 and to positive otherwise ; the remainder five bits will set the value of bi (from 1 to 3

asciilifeform: plox to point?

asciilifeform: this is not obvious from the recipe

asciilifeform: ( nobody in fact knows how to generate 'uniform float' from rng )

asciilifeform: floats suck mightily, incidentally, especially for 'model reality' or whichever involving rng -- because your uniform rng now becomes a skewed shitsoup because the number line is now unevenly populated

asciilifeform: ieee754 'double' is still 64b .

asciilifeform: mircea_popescu: how many bits of space do you want ?