log☇︎
2400+ entries in 0.313s
asciilifeform: apeloyee was right in that c is always 2^oneplusbitnessofx / m .
apeloyee: ok, so quotient of division of 2^4096 by 2^4096-1 is 1.
a111: Logged on 2017-10-05 16:06 asciilifeform: compute 'barrettoid' c for the current m : c := 2^4096 / p . this is done via the knuthian longdividotron, once per modexp.
apeloyee: http://btcbase.org/log/2017-10-05#1721060 << c := 2^4096 / p; i thought you only have integer division? ☝︎
asciilifeform: 2 ^ 4096 is a 4097-bit number
apeloyee: ok, let m =2^4096 - 1. then c=1
asciilifeform: ^ also misprint, good catch
asciilifeform: of x is 0 .. (2^8192) - 1 .
asciilifeform: range of m is 1 .. (2^4096) - 1 in this example.
apeloyee: did you mean 2^8192?
a111: Logged on 2017-10-05 16:06 asciilifeform: compute 'barrettoid' c for the current m : c := 2^4096 / p . this is done via the knuthian longdividotron, once per modexp.
trinque: mircea_popescu: ^
asciilifeform: the (unsurprising) surprise is that ^method is wholly absent from the public lit ☟︎
asciilifeform: compute 'barrettoid' c for the current m : c := 2^4096 / p . this is done via the knuthian longdividotron, once per modexp. ☟︎☟︎
mod6: ^
asciilifeform: ^ for completeness would also say ' a bsd would, in theory, work ' but practice has not confirmed : http://btcbase.org/log/2017-08-06#1694302 ☝︎
asciilifeform: seems ^ plausible.
asciilifeform: hey BingoBoingo ^ steroisism ?
asciilifeform: midnightmagic are you aware of ^ term of art ?
asciilifeform: ^
mircea_popescu: ^ early noether, by any account.
phf: ^
mod6: <+asciilifeform> mod6, diana_coman , whoever else was tuned in for ffa ^^^ latest . << got it, thanks!
asciilifeform: ^^^
asciilifeform: ^
a111: 10028 results for "\"^\"", http://btcbase.org/log-search?q=%22%5E%22
asciilifeform: !#s "^"
asciilifeform: mod6, diana_coman , whoever else was tuned in for ffa ^^^ latest .
asciilifeform: ^ various threads re how and why asciilifeform did this
asciilifeform: ^ asciilifeform's very painstaking 'trbfication' of koch
trinque: ^ for some curious reason this thing's behaving strangely dropped on a bsd
diana_coman: trinque ^
asciilifeform: this ^ is first time i see 'chinese' and 'depressive disorder' in same sentence
asciilifeform: ^ dun weigh much, and even not all rubbish
asciilifeform: ^ prolly most detailed record of usg gasenwagen session to date
asciilifeform: ^ offers a solution to the megapuzzler, but takes 'complexity classes ARE disjoint' as a lemma.
mircea_popescu: suppose there's a function that does hashing over a domain consisting of 2^100 possible distinct values.
mircea_popescu: so notrly. only thing left for him is to joestack before they ^
trinque: ^ brb
trinque: ^ migration in progress.
asciilifeform: ^ somewhere in there
asciilifeform: ^ old tale with similar premise
mircea_popescu: elaineo ^
asciilifeform: ^ incidentally : if you can get hall-effect switches, these will work even better ( the important thing is for whatever switch to be 'bounce-less', contact-bouncing creates predictable, in theory, temporal patterns )
asciilifeform: ^ not massively different from prev snapshot, but here it is.
asciilifeform: ^ latest.
asciilifeform: ^ fans of al schwartz will recognize
trinque: ^ wd
asciilifeform: ^ 'armory' seekrit-splitter leaks keybits
trinque: BingoBoingo: ^
shinohai: ^ that one
a111: Logged on 2017-09-20 23:12 apeloyee: longmod procedure in linked pdf says "2^b < m <2^(b+1)" on page 9
apeloyee: longmod procedure in linked pdf says "2^b < m <2^(b+1)" on page 9 ☟︎
asciilifeform: ^ no barrett yet. and there is still room for polish in barrettless variant, there is still a great deal of avoidable shifting and subtraction of guaranteed-empty words in FZ_Mod ( exercise for alert reader, to see where ! )
trinque: ^
BingoBoingo: ^ EPA approved
phf: ^ https://archive.is/e7TuU/a9e266b5e26e862285dbfc1c2fe15369c3f6d2db.jpg << physiognomy is real(c)
asciilifeform: ^ 0 mentions until nao.
mircea_popescu: yes, but as long as you do rsa, a ^ b mod m = either a ^ b or else, a ^ b - m. that's the complete story, there are no other cases. there's specifically no k * m parameter to be explored there.
mircea_popescu: fine. and the reason your mod does not reduce to "At most one substraction op" is "because i believe the product of two prime numbers 2kb each can in fact be less than 2^4095". that it ?
mircea_popescu: so then! calculating a ^ b mod 2^4096 is 0.26s ; calculating a ^ b mod m is 51.3s. that it ?
asciilifeform: FZ_Exp does not calculate a^b, it calculates a^b modulo ffawidth. this is important difference.
mircea_popescu: soooOOOOooooOOOO.... calculating a ^ b is 0.26s ; calculating a ^ b mod m is 51.3s.
asciilifeform: a ^ b mod 2**4096 , rather than a^b ( which our galaxy hasn't enough atoms to represent for 4096bit a and b )
mircea_popescu: sooo.... calculating a ^ b is 0.26s ; calculating a ^ b mod m is 51.3s.
mircea_popescu: so, calculating a ^ b is 0.26s ; calculating a ^ b mod m is 51.3s.
mircea_popescu: asciilifeform so to try and extract actual reality from this : a ^ b mod m takes 0.26 seconds in koch writing ; same a^b mod m takes 51.3 on your box, notwithstanding a ^ b without any modding takes 1s ?
asciilifeform: ^ and yes there is an 'obvious' 2x speedup possible and not yet done, even in this naive method
a111: Logged on 2017-09-17 11:10 mircea_popescu: (pro tip : since the modulus is 4096 bits, all powers of two up to 2^4095 are unchanged by modding.)
mircea_popescu: (pro tip : since the modulus is 4096 bits, all powers of two up to 2^4095 are unchanged by modding.) ☟︎
BingoBoingo: ^ mats
BingoBoingo: ^ AHA alf
asciilifeform: kanzure ^ for your enjoyment
BingoBoingo: ^
kanzure: 10^14 kg of wafer-grade silicon vs 1 kg dna
kanzure: anyway the concern is that there's not enough pure silicon available to make enough flash memory for future projected demands of like 2^28 bits in ~20 years.
kanzure: sorry, that was the wrong number. 10^14 bits is today's write demand.
kanzure: i was in dc yesterday attending a dna synthesis meeting organied by iarpa. they want 10^14 data write with dna.
asciilifeform: ^ thread somewhere
asciilifeform: BingoBoingo: i must admit that i found it utterly impossible to determine what the piece ^ is actually about
BingoBoingo: ^ ty alf for the pickets
asciilifeform: ^ which does mean that i'ma have to a) audit the binary when ffa built for use in the field b) patch gcc/gnat not to emit DIV ☟︎
shinohai: ^.^
phf: ^ http://rollingstoneaus.com/assets/Uploads/1401x788-GettyImages-74702183.jpg
asciilifeform: as of my last reckoning O(n^3 log n)
mircea_popescu: n^3 ?
a111: Logged on 2017-09-13 20:15 mircea_popescu: http://btcbase.org/log/2017-09-13#1713497 << man your answers are terrible! what happened to "yes, our current item is O n^3 log n ?" or "no, our current item is n log n" ?
mircea_popescu: http://btcbase.org/log/2017-09-13#1713497 << man your answers are terrible! what happened to "yes, our current item is O n^3 log n ?" or "no, our current item is n log n" ? ☝︎☟︎
a111: Logged on 2017-09-13 14:34 asciilifeform: ^ http://ulady.ru << exactly what it looks like
apeloyee: standard barrett requires a*b<n^2, yes. hence I specified the double-precision reciprocal.
asciilifeform: and incidentally barrett , as iirc i mentioned in yesterday's thread, requires that a*b < n^2.
apeloyee: ((maxint+1)^2)/N
apeloyee: what woud be wrong with barrett reduction if you calculate a sufficiently precise reciprocal of modulus N,i.e. (maxint+1)^2/N?
apeloyee: would O(N^2) modular multiplication be too slow?
BingoBoingo: ^
asciilifeform: ^ http://ulady.ru << exactly what it looks like ☟︎
BingoBoingo: ^ In other "Think of the Children!"
mircea_popescu: consider the number 97. is is 1100001. they do mp_mod (2^6, 2^5, 2^0) ; you can do (2^6, 2^5, 0* 2^4, 0* 2^3,0* 2^2,0* 2^1,2^0). the list method will sitll work, but this time in constanttime.
asciilifeform: so as not to feed a massive turd into an O(N^3) division algo
asciilifeform: well we are talking about a O(NlogN) rsa vs a O(N^5) one