114900+ entries in 0.874s
ascii_butugychag: how many us folk realize that usa was mega-power in mid-20th ~because~ it was
a place where folks like ulam wanted to live ?
danielpbarron: i went to college for math, even if i did drop out right before actually getting
a degree
assbot: Logged on 18-11-2015 14:54:26; asciilifeform: shinohai: 'i heard that ivan won
a car in the lotto.' 'yes! but not ivan but piotr, not
a car but an overcoat, not in lotto but at cards, and not won but lost'
danielpbarron: ascii_butugychag> danielpbarron was
a cryptographer ? << notrly
ascii_butugychag: the one hitler has, which emerges at the press of
a button, when suction is required ?
assbot: Logged on 04-02-2016 20:37:03; ascii_butugychag: for
a degreed mathematician, or even talented amateur, the life choices are generally
a) web dev b) usg.
mircea_popescu: as is the ballet teacher holding
a switch or light cane.
mircea_popescu: "Dear Ms. Tarzian: Here's another idea I've had: Weight Beaters. Weight Beaters are
a method of encouraging participants to lose weight.
A participant who does not lose the desired number of pounds in
a month is beaten up. This negative feedback can, of course, be combined with more traditional positive-feedback weight-loss mechanisms."
mircea_popescu: ascii_butugychag speaking of which, this very expensive rare custom job of
a server that was made to be
a phuctor host is going to be paid for the 2nd idle month.
mircea_popescu: sure, but i wish to hear
a bunch of "oh there's nothing wrong with tiny block ciphers - bitcoin needs larger blocks!!11"
ascii_butugychag: yes but what precisely is the point of
a block cipher that is slower ~and~ bulkier than abused-rsa.
thestringpuller: ascii_butugychag: if the looms are destroyed in
a fire, there may be no way to reproduce exactly the same end product.
ascii_butugychag: thestringpuller: this is
a very 'hollywood' conception of 'blueprints'
phf: heard similar stories about sending monkeys to moon from nasa people,
a lot "research" is
a large multi-cultural team trying to reverse engineer 1970s suite glove, etc.
ascii_butugychag: thestringpuller: it is
a dead field for different reason entirely.
thestringpuller: ascii_butugychag: re: nuke designer. I remember reading
a wired article saying that practical nuclear engineering (bombs) is
a dying field since no one can technically test nuclear detonations. Most of the students who were learning from manhattan project-age engineers were learning via word of mouth.
ascii_butugychag: sorta how i'm being paid, through
a pyramid of derpatronic intermediaries, 'not to tmsr' for 8h/d
ascii_butugychag: usg is doing
a decent job of paying most qualified folks ~not~ to crypto.
ascii_butugychag: the champs of recent history re: both-at-once was prolly israel. hence
a dead gerald bull.
ascii_butugychag:
a wild cryptographer is as tolerable to usg as
a wild nuke designer.
mircea_popescu: if the snake was as competent as is required for
a living animal, it would certainly not harry djb with inept shit of that level.
mircea_popescu: it also has
a lot of bitcoin experts, scattered all over reditg
ascii_butugychag: for
a degreed mathematician, or even talented amateur, the life choices are generally
a) web dev b) usg.
☟︎ mircea_popescu: what's more "well specified, mechanical" than
A FUCKING AIRPLANE
mircea_popescu: nope. they can only build items that look enough like the image of
a plane as hashed through their culture.
PeterL: did you set
a time limit for it?
ascii_butugychag: cipher designers have
a severe occupational disease, of confusing own ignorance for actual intractability
mircea_popescu: there's
a sad dearth of research in this field. we don't even klnow which of pi*e and pi+e are transcendental. etc.
mircea_popescu: give people
a perfectly valid reason to only take crypto : "i don't want to be spammed"
mircea_popescu: contrary to piously fraudulent consensus in the field, expensiveness is
a quality of cryptographic items.
ascii_butugychag: this is sort of why i'd like to take the opposite approach, rather than take
a 'this looks confusing!111' item like transcendental digits, take
a proven-nphard problem and 'cryptoify' it somehow
mircea_popescu: the properties which º must obviously exhibit make it only partly similar to plain multiplication - perhaps group theory may produce
a good candidate, i have not currently
a very clear picture of what this item should actually be, but seems like some sort of modulo-multiplicator.
mircea_popescu: to decipher you calculate the maxint+key to maxint+key+block digits of e with
a plouffe-like algo (not here included) which you º' with the ciphered message yielding the original plaintext.
mircea_popescu: you need
a proper "destructive multiplication" item, which i'll discuss later, noted here º. you proceed to calculate the maxint+key to maxint+key+block digits of pi via repeated applications of the original plouffe algo, and you º the result with the plaintext message which yields the ciphered message.
phf: mircea_popescu: i've been using his guide in the early 2000s, but then i think he caved and switched to laptops, because he stopped updating it. he has
a recent build recommendation
https://blog.cr.yp.to/20140602-saber.html from his "saber cluster"
mod6:
A great idea, overall. Thanks for posting that and putting up the reward.
mircea_popescu: from
https://cr.yp.to/bib/online.html : "instead of signing
a copyright transfer agreement. If you ever encounter
a publisher that doesn't accept this, let me know, and I'll be happy to blacklist that publisher here. I'm now blacklisting IEEE and ACM."
mircea_popescu: idealism is generally
a stupidity-amplifier in all seen deployments.
ascii_butugychag: the hilarious part about american politicircus is that none of the groups pandered-to actually ever ~get~ so much as
a bite of the carrot
BingoBoingo: Alright the one thing worth taking away from the "Phenomenology of Spirit" is that USSA is evil in
a more profound way than USSR was most of the time.
PeterL: you don't think sanders has
a chance to beat her?
mircea_popescu: it's not, no. what, you use
a single key for all udp packets ever ?!
mircea_popescu: nobody is spending
a dollar to break
a safe that holds five cents.
assbot: Logged on 04-02-2016 17:05:32; mircea_popescu: give me
a 2kb replacement for bassomatic, in 1991. or
a 64kb, today.
mircea_popescu: ascii_butugychag the example then was that the chain allows you to introduce known-weak points in
a subsequent pass without being able to know it
ascii_butugychag: this is rather like arguing that 'cutting
a man's abdominal cavity open will weaken and kill him.' YES if we haven't invented antiseptics yet
assbot: Logged on 04-02-2016 16:53:38; mircea_popescu:
http://log.bitcoin-assets.com/?date=04-02-2016#1396294 << this is like falling in love, you don't make one just because you need one. chaining things you don't understand for accountant's business sense reasons results in unexpected weaknesses. was
a thread here about just that few months ago
mod6: alright, i've got
a local fix for gernika's bug where it'll press the entire tree out if the given vpatch is not in the flow.
assbot: Logged on 04-02-2016 17:05:32; mircea_popescu: give me
a 2kb replacement for bassomatic, in 1991. or
a 64kb, today.
mircea_popescu: after which i will write
a cipher competition, and include my idea.
assbot: Logged on 04-02-2016 16:57:23; mircea_popescu: ascii_butugychag seriously, you are going too far too fast. we are not ready to do this yet. i do not wish to use, and i see no purpose of having deployed in any capacity,
a block cipher that does not satisfy
a) and b) and idealld d) from
http://log.bitcoin-assets.com//?date=02-02-2016#1394698 mod6: yeah, i agree. im hoping that after this round of fixes in place, that the mission-critical features are solid enough to warrent
a genesis.
mircea_popescu: past
a certain fixedness level, it can just fix itself.
mod6: I've got
a local fix in place for ben_vulpes' error that he brought up:
A file (of any type) exists in 'patches', and it causes V to fail ungracefully.
mod6: but yah, in general, it should have its own tree of vpatches. and will save me
a lot of pain too :]
mircea_popescu: fix MY fucking flat tyre so i go where I wanted to go in the first place., don't tell me idea is better than 2kbit bassomatic, i couldn't give less of
a shit.
mircea_popescu: so sure, they "fixed" his error, but the wrong way. i don't want
a fix for
a flat tyre that consists in loading me up on
a trai nthat goes
a different direction.
mircea_popescu: and re bass-o-matic, note that while ineptly implemented, the origianl p.z. had the exact same intuition. it happens to be the correct posture here, with the expansion that 25 years added - back then 256bytes was
a disk block iirc.
mircea_popescu: it's
a toy, or at best
a ready equivalent for haskell crap in
a different circle jerk.
assbot: Logged on 02-02-2016 23:48:55; mircea_popescu: incidentally, thinking vaguely along the lines of making
a b-
a call for papers for symm cypher, what would we actually want ? i'm thinking
a) block sizes of 1, 4, 16, 64 kbytes. none of that bit-denominated bs, wtf is this, 64 bits. fuck that. b) key size of 64kb fixed. c) bonus points for proved hardness, as-hard-as-x etc d) bonus points for not using just basic arithmetics. fuck this shift-and-xor