log☇︎
20200+ entries in 0.368s
phf: i'm trying to figure it out from first principles :) (i haven't had time to look at the recent, i.e. past month, versions yet)
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
asciilifeform: betcha it won't be 1.2s nomoar
asciilifeform: ( if it isn't obvious from where the error comes : observe the 3 Karatsuba_Term additions. in ordinary K., they walk over the upper half of XYLo ( lower half of result.) but in TopOnly K. we lose XYLo, so that carryolade is lost. )
mod6: (for those who don't have the text handy)
a111: Logged on 2017-10-07 01:36 mircea_popescu: today for eg, felt like walk, but didn't feel like climbing 20% inclines, so had girl take me to park. 20km driven to walk 3 or so. imagine the decay.
asciilifeform: http://btcbase.org/log/2017-10-07#1722233 << what's yer altitude, mircea_popescu ? because i can't picture why else ☝︎
mod6: her haggard old bag of a mom didn't approve? or did i dream this up?
mod6: didn't some lady throw some snizz your way one day while walkin by?
mircea_popescu: today for eg, felt like walk, but didn't feel like climbing 20% inclines, so had girl take me to park. 20km driven to walk 3 or so. imagine the decay. ☟︎
mod6: ah, and walking up mountain sides dodging bikers in bazerker mode doesn't help
mircea_popescu: used to. i suppose this place being ~a village doesn't help.
mod6: well, if you walk for 2/3h per day, maybe don't think of them as "shoes". perhaps, "proper foot equipment" or something.
asciilifeform: can't wait for next forkbase, will be hilarious to watch the 'into btc' folx get anally reeducated
asciilifeform: 'if it doesn't offend sense it dun get called OfficialTruth' or how it went.
mircea_popescu: and in continuing lulz, "Bitcoin price is up but top Wall Streeters aren't on board"
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
mircea_popescu: or don't i understand "how the world works"
asciilifeform: http://btcbase.org/log/2017-10-06#1722033 << lemme know mod6 if there's any part that dun immediately make sense. FZ_Make_Barrettoid looks complicated but really is identical to the optimized knuth divider, with the difference that the dividend consists of all 1s and known ahead of time, so we don't shift it ☝︎
a111: Logged on 2017-09-16 02:57 PeterL: by the way, I stuffed the keccak ada stuff (and, speaking of OAEP, here is one of those too) into https://github.com/PeterMLambert/keccak since I don't have my own server up yet
asciilifeform: ( that last one doesn't stand on a hash )
mircea_popescu: the categorical alternative (literally what gpg does now) fell on the grounds of "at least we don't need aes" ; the obvious "cut R into bits and use each" seems to my eye weaker, tho who even knows.
mod6: <+asciilifeform> mod6: this is the most brutal possible method that can still be called barrett : didn't even make use yet of the fact that we only want the upper halves of the mul results << spent the afternoon looking through your changes posted today vs. http://btcbase.org/log/2017-10-02#1719721 ☝︎
mircea_popescu: i may opt to include a pubkey. but i don't have to.
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: mod6: this is the most brutal possible method that can still be called barrett : didn't even make use yet of the fact that we only want the upper halves of the mul results
BingoBoingo: Not like it hasn't already been shaped by tree's executioner
phf`: mircea_popescu: if you're implying that my plan for colo was to ~hire~ people ~from~ colocation facility to do a sensitive operation like hard drive replacement, then i can't even! :)
mircea_popescu: phf` that wouldn't be a problem in principle. just has to be done.
mircea_popescu: i don't expect overnighting is the point. what's the point is that there's little difference between a colo you email to and run of the mill "dedicated servers" hosting.
phf`: i've no idea where that straw man came from. shipping things from u.s. to russia is tricky, but everything else is inherent in the problem. a ~colo~ in russia only makes sense if i also can somehow facilitate trusted upkeep. whether or not we can overnight a hard drive isn't going to make a difference in that case
mircea_popescu: if i wanted that i could have it for the same coupla hundy or w/e. heck, apparently they'd even do it for free if we don't make fun of them for it.
mircea_popescu: on the other hand, this "your raid needs new hdds hurr" thing is not flying. i don't need to "hire" a "colo" for them to re-do the whole cherryservers bs for me, now do i.
asciilifeform: phf`: we install , for instance, FG. ain't nobody else who will do it.
mircea_popescu: i don't even care, sure, if they have their own state courier they're trying to prop up in this way, more power to them, we'll use it.
phf`: asciilifeform: i can't find any news updates, last i heard from people is that fedhl packages go, but spend weeks to months in customs, for "inspection"
mircea_popescu: i can't imagine the country actually works without some kind of such service though. i mean really, how do yo uget a wrench, walk to factory ?
phf`: are the servers sitting at vendors ready to ship to a destination, also is it going to go proper shipping container or through a consumer service? latter doesn't work reliably in russia and is generally discouraged
mircea_popescu: or rather, that's why normal people don't pick a hill in the midle of the jungle to plop down servers on.
mircea_popescu: i definitely want the dc to provide the uplink with w/e guarantee they can manage. i get it, outside of nato 99.999% isn't so common, but whatever they got.
phf`: the place to be is of course http://www.mmts9.ru they do have 1/2 rack but i haven't tried them, since wasn't sure about our requirements. they can even get you a personal cage. antenna placement on the rooftop so we can put a microwave relay, the works :p
mircea_popescu: anyone inclined to go "oh i'd have done a much better job" after any particular doing is going to find himself buried under a thick wall of "why the fuck didn't you DO IT then!".
phf`: mircea_popescu: right, i figured paying is not going to be a problem, either directly with btc or in the format that you've described. finding a reliable colo, getting a number down. unfortunately i don't have a good feel for russian colo space yet, and people i talked to while happy with their colos didn't necessarily have tmsr demands.
mircea_popescu: phf` so the question is very specifically, what do they take, and how do we sort it out ; it's in no case a "let's filter out whoever doesn't take bitcoins"
mircea_popescu: phf` i don't specifically care what they take ; but whatever it is they take, we'll have to figure out a process.
phf`: there's a bitcoin craze in russia right now (i.e. talk shows on t.v. discussing bitcoin as investment platform. dorky former FIDO node administrators talking about bitcoin mining on a rig in their apartments, etc.) so i wouldn't be surprised if i can find a colo that takes bitcoins. i haven't noticed one yet though
ben_vulpes: http://btcbase.org/log/2017-10-06#1721789 << gears grinding. i don't know anyone in beijing, but i've got a line on some folks in chengdu for translation services and building lists of eg datacenters ☝︎
asciilifeform: the ruler doesn't lie.
phf`: not home, and i also didn't have time to process the information in detail yet, but i piped in since there was conversation. depends on what we want, baseline avg 3000 rub/mo $50 to put a 1u rack, with a 100mbit line, 300W. line is typically not metered, but also not guaranteed. then depends on our needs, an unmetered guaranteed 100 mbit line will cost another ~~$100 mo (to 1gbit around ~$500), plus whatever over cap on power usage.
mircea_popescu: every nonprofit has one, including nonprofits that don't actually exist.
mircea_popescu: http://btcbase.org/log/2017-10-06#1721786 << this, fwiw, isn't even the giveaway it used to be. i personally supervised the release of piles upon piles of posters in us university campuses, for a very modicum costs. people off fiver just print what you give them, and put it up. nothing keeps you from having them print something decent, and with the abundance of "kids who fucked around with adobe illustrator instead of doing t ☝︎
mircea_popescu: you know the above ru may well be a case of "we don't deal with weirdos trying to get 1u for 1mo"
mircea_popescu: asciilifeform i really don't see point of buying less than 1 rack.
mircea_popescu: in any case, these aren't even that huge.
asciilifeform: in ru of course. but i don't have a phone or post box in ru.
a111: Logged on 2017-10-05 18:20 mircea_popescu: ok so results of audit : Framedragger, shinohai, phf : your bot has no help implemented whatsoever, in spite of spec. trinque Framedragger you don't follow the json/sexpr portion, bot simply puts out the same help.
mircea_popescu: asciilifeform the amusing bit to my ear was the implicit "only good citizen is citizen who isn't doing anything"
shinohai: Those 72 virgins ain't gonna fuck themselves.
asciilifeform: 'don't run from sniper, you'll just die tired' (tm)(r)
a111: Logged on 2017-10-06 11:55 Framedragger: ty diana_coman. sad to see the state of tmsr isp affairs. i suppose it was kind of always like this, the state simply got.. actualised, so to speak. still, am planning on giving small talk on ssh scan to hackerspace people, wanted to link to phuctor, now - can't. :(
diana_coman: well, one can't commit what one doesn't have, sure
diana_coman: anyways re state of isp affairs I don't see it getting any better by itself
Framedragger: ty diana_coman. sad to see the state of tmsr isp affairs. i suppose it was kind of always like this, the state simply got.. actualised, so to speak. still, am planning on giving small talk on ssh scan to hackerspace people, wanted to link to phuctor, now - can't. :( ☟︎
a111: Logged on 2017-10-05 18:20 mircea_popescu: ok so results of audit : Framedragger, shinohai, phf : your bot has no help implemented whatsoever, in spite of spec. trinque Framedragger you don't follow the json/sexpr portion, bot simply puts out the same help.
a111: Logged on 2017-02-19 16:12 mircea_popescu: (the source was 80s dispute wrt to general public's apparent bottomless appetite for purchasing luxury goods -- in eastern communism this includes meat, washing machines, electricity, etc -- for which they won't even have the fucking decency to pay in hard currency, but aim to use the worthless reminbo internal scrip)
a111: Logged on 2017-10-05 19:54 apeloyee: grind until you did (2*bitness) steps. since remainder can't be more than a/2.
asciilifeform: apeloyee: maybe i'm thick, but i don't see how to grind 'shift so b has one digit less than a' a fixed number of times for all possible a and b.
apeloyee: grind until you did (2*bitness) steps. since remainder can't be more than a/2. ☟︎
asciilifeform: that ain't constant time !!
apeloyee: if they're equal -that's the end, the gcd is found. don't throw out the unshifted b.
apeloyee: I don't see any GCD in miller-rabin
apeloyee: and in one side only, as obv can't be bigger than (A/N)*4^K
apeloyee: if N is 2^(K-1), then ordinary quotient won't fit in K+1 bits. but pseudo-quotient (one less the actual quotient) still works.
asciilifeform: apeloyee: it won't differ at all, once i have working shifter using your scheme.
apeloyee: I assumed the elementary shift doesn't leak. whatever, just do the sub-word shifts with the same algorithm (with word size 1)
asciilifeform: not only an idea, but afaik the only practical method that isn't W-shifts-by-1-with-mux
asciilifeform: hm i think i finally see what algo apeloyee was trying to implement ( the pseudocode doesn't actually do it )
apeloyee: goddamnit. it's FIXED!!! first sub-word one (which doesn't leak in our model), then by 1 word, then by 2, then 4, 8 and so on
apeloyee: fix word-size and bitness. otherwise, you don't
asciilifeform: apeloyee: it ain't log(bitness) ! not if you don't want to leak any info re the shift amount.
apeloyee: the operations themselves leak anyway. i thought they shouldn't leak operands?
mircea_popescu: why can't i grep "hurr" over a cat #trilema bots ?
ben_vulpes: and what, a machine-understandable description of what each extra-spec command does? i didn't think anyone here operated machines that /thought/.
ben_vulpes: the response from the bots to what they can do. either they're all spec-compliant, respond the same way to the same set of commands, or they don't.
ben_vulpes: should be binary, #t-spec compliant or not.
mircea_popescu: but practice aside, shouldn't i have the ability to query a bank of seated whores, "what can you do ?"
mircea_popescu: ben_vulpes because we just don't care about machine discovery, now or ever ?
mircea_popescu: ok so results of audit : Framedragger, shinohai, phf : your bot has no help implemented whatsoever, in spite of spec. trinque Framedragger you don't follow the json/sexpr portion, bot simply puts out the same help. ☟︎☟︎
mimisbrunnr: Logged on 2017-10-05 16:14 mircea_popescu: phf ben_vulpes Framedragger excepting deedbot from the "don't read out bods" ? feasible ?
apeloyee: if q' <= q <= 4q' , it means you don't even know ONE significant digit of q'
mircea_popescu: either they use your money or else they don't get shit.
mircea_popescu: asciilifeform obviously, you didn'[t expect the actual USG has any resources ?
asciilifeform: nah, simply for same reason street sweepers don't -- if you give'em terminals, ~0 packages will get stamped
cruciform: didn't get hit with import tax, unlike diana_coman
asciilifeform: lemme guess, cruciform , you didn't get the paper manual ?
a111: Logged on 2017-08-13 20:36 mike_c: i know, right. It's frustrating prioritizing time. Didn't have time to hang, definitely not to work, so felt weird dropping in once a month and saying "hey, haven't done anything, don't have time to do anything"
mircea_popescu: phf ben_vulpes Framedragger excepting deedbot from the "don't read out bods" ? feasible ?
asciilifeform: in ordinary barretting, x loses its bottom however-many bits, to make the multiplication smaller, depending on the 'bitlength' of x; and then its gets restored with a leftshift . we can't do this, because ffa.
mircea_popescu: i don't think it can be put into words, how much of the whole "poor people need help" the pay-for-your-tits disabused in my mind.
shinohai: If I made that offer on some South-American camslut site, it wouldn't last a week.