log☇︎
268500+ entries in 0.189s
asciilifeform: not to mention oceans
asciilifeform: then it was deemed impractical, to have separate fibers run in unmarked pipes, set in cement, around entire continent
mircea_popescu: the lulziest part being that they've been aware of this for years and don't know how to fix it.
mircea_popescu: a good chunk of the leaks are due to retarded client-server architectures deployued for no conceivable reason which they don't know how to secure
asciilifeform: and employs several battalions of www dev types.
asciilifeform: i've been told that nsa castle of doom actually suffers from the same profusion of 'dockers', etc. as typical american software co.
mircea_popescu: "here, download docker, configure orchestrator, push this butonetor and write your number. the restful interface will return a bevy of young ladies with the factors written in mascara on their tits."
asciilifeform: just like we aren't having a thread about testing strength of knapsack keys.
asciilifeform: and we wouldn't be having this thread.
asciilifeform: would mean that popping rsa is in P
mircea_popescu: for some reason i was thinking it's going to return a factor. dun ask me.
mircea_popescu: mno. you primality-test the N.
asciilifeform: that's the other thing - when you're massaging your key, you are also radiating it.
asciilifeform: also, mircea_popescu's airgap unit runs for 4 days in its cage ? i cannot boast of this.
mircea_popescu: so yes, vijay menon claimed in 2013 that his implementation of agrawal-kayal-saxena took 5k seconds for a 25 digit number ; and consequently we could expect as much as 1 minute/bit or cca 4k minutes for a tmrs-rsa key.
asciilifeform: i'd like to know from where he got the figure - no source is cited.
asciilifeform: http://cs.stackexchange.com/a/23360 << take with grain of salt, but possibly related.
mircea_popescu: it was however the basis of later stuff that WAS of use.
asciilifeform: (witness, even the '2 minutes per key' was rather too long to be of use.)
mircea_popescu: i just mean the part where "everyone knows the moduli don't factor" "oh look they do"
asciilifeform: if phuctor took 4 yrs/key, it would not be of much interest.
mircea_popescu: moreover, subtle errors in technology are best made evident with radically different approaches. "have both deterministic and non deterministic tests, notwithstanding IT IS PROVEN the latter converges" is substantially the same thought process as produced phuctor.
mircea_popescu: but i dunno exactly how loing it'd take ; and i suspect not THAT long.
asciilifeform: if crypto-academia were something like a going concern, rather than elaborate circle wank, we might actually have had one.
asciilifeform: and i'd personally love a deterministic test that doesn't take 4 years of cpu for a 2048-bit mod.
asciilifeform: there are other probabilistic tests, which also converge.
mircea_popescu: asciilifeform which is the reasoning here - yes rm can be made to converge. so what's it hurt to have another test ?
asciilifeform: i.e. there is always a - calculable - chance that your result bit uncorrectedly flips right before you read it.
mircea_popescu: by definition, the efficiency of the single given method to do anything is always 100%
asciilifeform: it also helps to remember that there is no such actual thing as a fully deterministic algo for a physical machine.
mircea_popescu: the definition of this "efficient" is suspicious.
asciilifeform: 'The very fact that the check for whether the number 9965468763136528274628451 is prime ( in Table 5.3 ) took as much as around 70 minutes shows how inefficient it is.'
asciilifeform: re, aks, if mircea_popescu can't be arsed to wade through the linked piece,
asciilifeform: 'oh we can't have keygen take a week, and besides, who will sit and wiggle the mouse.'
asciilifeform: since the dark ages, and up.
asciilifeform: mircea_popescu: this item re-appears in every single sad crypto proggy.
mircea_popescu: im ill equipped to cut through the inanity of gnupg.
asciilifeform: why is there 'randomlevel' and 'secret' flag ?
asciilifeform: witness the sheer inherent sad in, e.g., http://btcbase.org/log/2016-08-18#1524029 ☝︎
mircea_popescu: there's layers to these dudes' "improvements"
mircea_popescu: and no, obd does not need 1mb, or beenfit from it. point was - if you're going to use spare entropy, it would be for this rather than too much waltzing or w/e.
asciilifeform: though funnily enough koch didn't pump ~that~ entropy from the whitened pool
asciilifeform: the ~implementation~ of the latter in gpg - is other question entirely.
asciilifeform: (interest in the subj is slim, on account of miller-rabin probabilistic test provably converging - i.e. probability of catastrophic failure can be made smaller than probability of machine flattened by asteroid the same night) ☟︎
asciilifeform: i have nfi, it would be interesting to determine it.
mircea_popescu: something, so i can put this in numbers.
asciilifeform: you mean, what is the constant multiplicand ?
asciilifeform: iirc, best variant that does not introduce an unproven conjecture runs in O(log n) ^ 6.
mircea_popescu: "your key will take a week to check on a contemporary box" "this is fine"
asciilifeform: mircea_popescu: iirc the runtime on the one known deterministic test (saxena's algo) is prohibitive.
mircea_popescu: AND moreover, most of that legendary mb of entropy should go into the primality test, not into the prime gen
asciilifeform: incidentally, the traditional pattern of 'push this button for zero security' is everywhere present.
mircea_popescu: incidentally... the primality tests, being probabilistic, are already bs.
asciilifeform: (rng is also used in primality test, but it is switched into 'fast' gear and doesn't use the pool)
asciilifeform: (or was it derived wholly from pw ? i suppose i oughta read the thing, later.)
mircea_popescu: e of your aqents on the street, you'll be recognizing them on youtube, by how their liver and spleen squirms."
mircea_popescu: heh. "Five minutes later, Burnett called back and said, “I don’t believe you actually represent her.”¹ Burnett stated additionally that a phone call from me might suffice, but that the FBI preferred to meet with me in person. After a pause he said, “But… if we happen to run into her on the street, we’re gonna be asking her some questions without you present.”" << "That's fine. And if she happens to run into som
asciilifeform: aha, the symmetric key for ciphering the privkey for storage.
asciilifeform: incidentally, mr. lolcow just gave away, unwittingly, another gpg laugh - recall what else comes out of rng when you generate key, after the primes ?
mircea_popescu: check out the retrospectively-done-it-all man.
asciilifeform: 'i was there, i did it!' -- al gore, 9/12/01
mircea_popescu: back in the tardstalk glory days of 2012 it was "because competition!!!", but by now that nonsensical coverup has been disposed with.
asciilifeform: https://archive.is/2CprW << i wonder how this lolcow is paid. per-lol ?
mircea_popescu: to shari, brenda, cher and madonna whatever, the notion that anything changed is entirely foreign. they'll quite not understand the source of derision eg above, and imagine, to quote, "Author seems to have an axe to grind."
mircea_popescu: once nsa runs out of budget for the project, they are still... "working", just as they did before.
mircea_popescu: see, that's the thing with plowing flies : for as long as the plow appears to plow, they imagine their clicking through lolpics and whatever-not-irc-chat nonsense is "work"
asciilifeform: what did these folks see themselves as having been ~doing~ all of this time.
asciilifeform: the bizarre autolysis is kinda interesting to me.
mircea_popescu: aaanyway. not all graves are worth the stream of piss, and in the case of tor i'm well spent.
asciilifeform: perhaps as a toilet brush.
mircea_popescu: i can't begin to give a shit about people with no parents.
asciilifeform: mircea_popescu: https://archive.is/ImDmM << bonus. this is... shari.
mircea_popescu: oh, really, what did brenda do ? oh no she didn't! that hussy!
mircea_popescu: ved there was victims, why did no one ask if I had bad experiences with Jake? This would have been easy given that Alison spent an hour alone with me in a cafe at the last TorDev meeting and had several communication channels with me. Also I have been at Isis' place several times in the last months as I am dating her roommate Brennan" <<< MOTHERFUCKER. so what, i'm to watch low rent reruns of beverly hills 90201 now ? da fuck!
mircea_popescu: "The first thing I said when I saw this website, is that Shari failed. She knew people (many datapoints indicate the site creators are Alison and Isis) were about to do this. How could she let this happen? I am strongly wondering if this was really about protecting women. Here's what I noticed: I am a woman, and I estimate that more than half of the Tor core people know I was dating him, or at least close to him. If Tor belie
mod6: yeah, there is that part too.
asciilifeform: (incidentally i have nfi whether gnu patch util will apply a patch that simply 'FALSE'-s a deleted file, without the minusola.)
asciilifeform: mod6: but in the case of vdiff, there is no use for it, it simply makes patch unreadable.
asciilifeform: mod6: in traditional diff, turning a deleted turd into a massive ball of minus is arguably necessary, to prevent patch from applying to arbitary item
mod6: <asciilifeform> imho 'vdiff' ought to zap 'minus' changes in cases where an entire file turns 'FALSE' (absent) - they take up space for no reason, given as we display a hash in the --- entry. what do you think ? << ah... ok. let me think about the impact of such a change here.
mircea_popescu: mod6 don't spend more than 15 minutes on it.
mod6: i'll work on an estimate for the 2015 time spent (for me at least).
asciilifeform: http://wotpaste.cascadianhacker.com/pastes/b385a4e6-836d-4829-b157-d157b09c5a4f/?raw=true << and the drepper grep, is too long to take up log space !
asciilifeform: ^ in case anyone did not know this.
asciilifeform: g10/ChangeLog: the new iobuf_skip_reset. Orginal patch by Florian Weimer.
asciilifeform: doc/ChangeLog: * gpg.sgml: Small typo fixes by Florian Weimer.
asciilifeform: ChangeLog: * README: Some typo fixes from Florian Weimer, and bump version
asciilifeform: ahahahaha, this is gold:
mircea_popescu: "But what happened is so unfair to so many people, in so many ways, that I cannot keep this for me." << this counts as "i raise my arms, to either kill or be killed, but to no longer continue this life"
mircea_popescu: that's the objection. they try to import terminology born of 9gag and at home on 9gag into actual discourse about their actual lives.
mircea_popescu: mno. because after being used for microamps, the item reverts to miliamps when exposed to miliamps.
asciilifeform: the moment derp's universe begins to include fallopian hangings, perspective will readjust.
asciilifeform: ;;later tell mod6 i came to this by attempting vdiff of gnupg 1.4.10 vs their .21 release from wednesday, and ending up with an unreadable 8MB turd, they moved/rm'd a buncha files. ☟︎
asciilifeform: ;;later tell mod6 imho 'vdiff' ought to zap 'minus' changes in cases where an entire file turns 'FALSE' (absent) - they take up space for no reason, given as we display a hash in the --- entry. what do you think ?
mircea_popescu: no sense of proportion, like they got a coupla octopus eyes.
mircea_popescu: schmuck seriously thinks "creating a website" "is beyond terrible". really ? and hanging her by her own fallopian tubes is what then ? beyond-beyond-horrible ?
mircea_popescu: chartered accountants / lion tamers, for chrissakes.
mircea_popescu: ever since http://trilema.com/2014/the-all-american-asshole-in-his-own-words-with-my-own-notes/ i can't shake the feeling all these, age irrespective, gender irrespective, are really the same community-college educated, chronic masturbator.
mircea_popescu: incidentally, what is with the redditarded generation, that whenever they have to do one of these important, man-talks-to-world, "soul searching" pieces they always come off as utter dweebs speaking in a dreary monotone about really trivial, boring minutia ?
asciilifeform: eh, mircea_popescu preemptively turned that knob to '11' with own hands, if it is turned any further it will probably come off.