138100+ entries in 0.022s

mircea_popescu: speaking of sexual offal, why is it that all the marginal women ~nobody'd give the time of day imagine they "speak for all of us", ie, somehow represent femaleness an' womanhood ?
☟︎ mircea_popescu: just because people turn heads when passing crashes doesn't mean running car into wall is "a sexuality"
mircea_popescu: dude gtfo, trans porn is interesting in the same sense any other gore is interesting.
mircea_popescu: in other news i finally replaced a rather loud psu. life's good!
☟︎ mircea_popescu: asciilifeform oh incidentally, node coloring! what's that called ?
mircea_popescu: jurov you recall that story with the trains and the fly ?
mircea_popescu: take patriarchy white man's "house of thought" give it to some rats ?
mircea_popescu: which is starting to explain how chinese ended up chinese rather than a proper language.
mircea_popescu: and there's absolutely no way that i can guarantee when drawing up the math of the cypher, that "all uses of symbol X will always be ambiguous in the space" or w/e dumb shit.
mircea_popescu: that's the bitch now innit. that's at least what i was longfaced about earlier, "in practical eavesdropping you don't EVER care about message integrity. if your algebraic cyrptosystem is porous to someone saying "bring me all messages that discuss ALEODOR" or whatever, and then he can discern something about them - say that ALL possible forms of one use it in the negative - you're still as fucked as a three year old in a pedo
mircea_popescu: but, that said : the principles to be built upon are exactly the "well, i can solve this any way you want" said the cracker to eve. hence, stuff like polynomial roots are very respectable idea.
mircea_popescu: the current discussion is from a "in house algebraicist" considering his job, "what's the worst that can come up when the suits decide to solve X"
mircea_popescu: so we likely need to re-do the hardness measurements anyway, from a crypto pov ; to use it in a crypto pov.
mircea_popescu: saying "x is np hard" is like saying "the us is rich". everyone you meet there is still fucking poor.
mircea_popescu: though some material recently linked made a very good point : "np complete" and other hardness discussions aren't even meaningful here. the term refers to WORST CASE, not to average case or anything.
mircea_popescu: it's because there's no option : either algebra ; or else "control structures".
mircea_popescu: course i guess that's all happenstance, nsa starts building long divisor processors.
mircea_popescu: yeah. but this is a lot more expensive sort of cycling than you know, hashing.
mircea_popescu: well not reinventing, but obviously will be needed for any practical scheme.
mircea_popescu: weaponized, this would fuck up say 1 in 20 messages it'd be so taut.
mircea_popescu: i've not done the actual math, per N size, but there's entire classes of messages that "could" work. you won't get enough precision until you have MANY messages.
mircea_popescu: suppose you know P and Q. ok. what now ? you can't actually calculate N.
mircea_popescu: mno, because you don't actually have something to solve.
mircea_popescu: still. N could be something like 3kb and the scheme'd be still workable.
mircea_popescu: and these trivially have any roots you wish them to have if you parametrize N
mircea_popescu: now in principle if P, Q are the encrypted chunks, (P/Q)^(1/(N-1)) and (Q/P)^(1/(N-1)) yields the message.
mircea_popescu: right. which is what i'm saying : if you're into ditching snake oil, you'll have to come to equations.
mircea_popescu: that's what aes did to try and escape "oh it's just maff", "fine we'll implement control structures".
mircea_popescu: just saying, it's going to be either equations or else... "boxes" or wtf.
mircea_popescu: the idea being that in principle there should exist a set a b c so as to turn any encrypted text into ~any text you wish it to be
mircea_popescu: select a, b and c and take N=(a + sqrt(b)) / c so that N > 1 ; N < 2. for any message M, deflate, crc and split it up in two fragments a b so that neither is larger than 2/3 of M ; to encrypt calculate a^N*b ; b^N*a and pass those.
mircea_popescu: o btw asciilifeform wanna hear my inept symmetric encryption scheme ?
☟︎ mircea_popescu: "well... all my friends are gone...i'm so lonely... listen, why don't you bring them back to me ?"
mircea_popescu: so you know, one goes back to moscow, the other goes back to wherever the fuck
mircea_popescu: "why you're so dumb i'd be loath to separate you from humanity. here, say where you each wanna go and i'll take you there, as an act of mercy (for you)".
mircea_popescu: and the genie's all "of all the dumb fucks wandering in here ever since the injuns figured out how to half'n'half a buffalo, you're easiest the dumbest fucks"
mircea_popescu: so a coupla hours later they're fucking stranded, and about to die, when this genie shows up
mircea_popescu: except they're all old and inept and ineffectual, they take this "suv" that's really running on a qcw-50 engine and take no water and whatever.
mircea_popescu: and eventually they decide to go for a ride in the desert
mircea_popescu: so they keep you know, fucking around, telling each other stories from back in their respective days
mircea_popescu: there's an ex-kgb ex-plutocrat dude ; there's an ex-yale ex-vc dude ; there's this dude used to run arms for rice in somalia ; and finally there's a rapper dude.
mircea_popescu: so there's a bunch of various people gangsta as fuck at the poker room of this casino of this retirement community that had originally been built in what was a us state but meanwhile was lost in some desert.
mircea_popescu:
http://btcbase.org/log/2016-06-05#1477486 << you just calm the fuck down and load the problem in head. if it's an actual blocking thing you just naturally stop sleeping until it's resolved ; if not, head goes on into a sub thinking about it ; which will eventually return with the results. this is how most trilema articles are written.
☝︎☟︎ mircea_popescu: not that they're not getting enough done in 5 minutes.
mircea_popescu: and the principal problem of exceptional people is that they have trouble keeping balance.
mircea_popescu: let's put it this way : the most stimulants can do for you is stimulate ; but on the long term they still debalance.
mircea_popescu: and then they want to legalize pot. worst fucking idea.
mircea_popescu: and finance sprawling suspended gardens of the future where ecofeminist tribes will raise the blabla.
mircea_popescu: if it weren't for trilema, people'd still derp about how bitcoin is here to save the geeks and the poor.
mircea_popescu: hence all the vomit about imaginary links between "cypherpunks" and bitcoin. and so on.
mircea_popescu: people have this capacity to imagine what's going on in their head is most important, and everything happening outside a sort of minor companion point.
mircea_popescu: anyway, i'm sure there were people in america who believed the spanish showing up to be a good thing.