log☇︎
128600+ entries in 0.079s
diana_coman: http://btcbase.org/log/2018-01-05#1764721 <- so I gather the 80 cols habit won the day for code ☝︎
a111: Logged on 2018-01-05 04:41 mircea_popescu will take this to the comment section now.
mircea_popescu: motherfucking mother of isis, the act of arraying buttons together in a guy is no design!
mircea_popescu: chrome tailpipes on already made car/pc/whatever cxhassis) somehow ended up called "designer".
mircea_popescu: anyway, thinking about this whole "fsf was an attempt to *finally* bring about socialist utopia through a fettering of everyone's access to knowledge while open source was an attempt to nevermindthatjustkillM$already leading to java-for-browser because microsoft invented a c++ market etc" broad but really mostly correct summary, the one most striking aspect is that somehow the job of the modder (ie, guy that adds those snazzy
mircea_popescu: these letters from a time before empire-of-idiots was formalized and understood as such are about as fascinating as a child's experience from before it understood any mechanics at all.
mircea_popescu: ey perceive it) blame all their (ample) shortcomings on the hulk barnacled while claiming all the (scant) benefit as own.
mircea_popescu: other than the lovely "if you launched all pantsuit in outer space, do you expect seti would manage to find any ?" putdown, valuable lesson from naggum : inept bureaucrats / insufferable cucks / other "people themselves" try to barnacle their inept nonsense (in original, scheme) on pre-existing "brand" (as they perceive it ; in the original -- lisp) for the transparently transactional reason that this way they "get to" (as th ☟︎
mircea_popescu: https://www.xach.com/naggum/articles/3129644482406982@naggum.no.html << this exchange with fare is pretty amusing/informative both wrt "who is this francois-rene rideau character ?" and as an early "bitcoin improvements from the sewer" sampler/potputre. ☟︎☟︎☟︎
mircea_popescu: "doom is inevitable, BTW. mankind will die out, planet earth will be vaporized when Sol goes nova, if not sooner, and then Common Lisp will have to acknowledge defeat to the unwavering hostility of the universe. for those of us who plan to become immortal, this is a serious concern." <<< 1999 naggum was apparently planning for immortality ?
mod6: <+mircea_popescu> holy hell i lived to see the end of the 4th day of 2018 logs. << ikr!
mircea_popescu will take this to the comment section now. ☟︎
mircea_popescu: diana_coman technically it's incorrect to say "6 bits will always be 1" ; it's the case that the first and the last bits of N are always 1 ; and a further number of bits are subtly affected (ie, biased) by the p, q masks.
mircea_popescu: "go and read some basic books" -> "read, re-read <em>and understand</em> TAOCP" ? or what, add K&R in there ? steele's 2002 thing ?
mircea_popescu: no known pattern ; but no homogenity either. there's always an even count of integers in a given bitsize ; but not necessarily an even count of primes.
asciilifeform: ( i can't see from where mircea_popescu extracts this )
mircea_popescu: (now go, predict things about how it falls)
mircea_popescu: ie, prime numbers themselves do not follow this 'no bit of input is more influential over the output than any other bit' rule.
mircea_popescu: moreover! while there's just as many integers with the 1337th bit set as there are with it nul, nevertheless it's trivially not the case that there's just as many 2048 bit prime numbers with it set as there are with it null.
asciilifeform: would be neat to generate keys in bounded time. and know primality for a fact. BUT i dun have, yet, such a pill.
asciilifeform: i dun have the martian one, ftr.
asciilifeform: can get to it the current way, or, hypothetically, some martian way.
mircea_popescu: even if i admit this objection, which i dunno i should but let's, for fellowship -- 2037 still not much room to work with
asciilifeform: so that's the actual keyspace available.
mircea_popescu: about 1 in 2k numbers is prime there.
mircea_popescu: anything other than 2045 loses.
mircea_popescu: mp-prime is stronger than alf-prime.
mircea_popescu: consider : mp-prime takes 2045 bits from trng, spits out after some wrangling 2048 bit prime number ; meanwhile alf-prime takes 45 bits from trng, spits out after some wrangling 2048 bit prime.
asciilifeform: mno. you have same 2044 or whatnot. just takes you longer. and with m-r , also chance to step on a mine.
mircea_popescu: if your magic maker uses say 2044 bits of input, therefore you have half as many possible states than i do!
mircea_popescu: but any other item is worse than what we have now.
asciilifeform: the described item is trivially impossible
mircea_popescu: so you don't have much to work with.
mircea_popescu: asciilifeform you have a different problem : even if "no bit of input is more influential than other" i STILL want to put in no less than 2045 bits of input ; AND get out no MORE! than 2048 bit long prime.
asciilifeform: and will add, that it is theoretically possible to satisfy asciilifeform's uniformity criterion ( 'avoids no primes' + 'no bit of input is more influential over the output than any other bit' ) without creating the mythical n-th-prime-maker
mircea_popescu: it'll be funny 5mn years from now, when we're all sitting around with whatever pools & eggnogs of the future and rsa still stands, undaunted, in preference of ~everything else.
asciilifeform: nor do i expect to see a solution.
asciilifeform: one of the biggest unsolveds , since greeks
mircea_popescu: asciilifeform there is no method to produce nth prime.
a111: Logged on 2018-01-05 00:12 asciilifeform: right, 'engineer' is more a psychological term to asciilifeform , rather like rpg character class, than diploma.
a111: Logged on 2017-11-07 16:36 asciilifeform: let's model the ideal prime-shitter. it would be an item that takes integer N , of whatever bitness, and produce the Nth prime ( or eggog if the Nth prime is bigger than the register bitness permitted. )
asciilifeform: ( see the thread )
asciilifeform: specifically that a proof of uniformity for an apeloyee-style primeconstructor mechanism, is in nearer reach than a runnable-on-2048b-ints aks.
mircea_popescu: holy hell i lived to see the end of the 4th day of 2018 logs.
asciilifeform: re prime gen -- i suspect that the problem can be finessed ,
asciilifeform: the sheer wtf, srsly, consider the folx who imagine they are somehow cryptoizing on a machine with no trng
a111: Logged on 2018-01-04 23:39 asciilifeform: generally my page will in the end have ~equal area of pencilmark and printerola
mircea_popescu: http://btcbase.org/log/2018-01-04#1764363 << possibly should auction the piles. ☝︎
asciilifeform: 'what if running on toaster, with no rng'
asciilifeform: i mean ffs, koch dun even leave a knob to get ~key~ entropy trngistically.
asciilifeform: 'can't pick witnesses trngistically because reasons'
mircea_popescu: what nsa gains from being able to rely on 2 rather than 44444444444444444 is anyone's fucking guess
mircea_popescu: dja understand motherfucking koch fixed one of the witnesses in mr ?
mircea_popescu: i'm dealing with "four rounds of m-r and a 5th fixed with 2, so as to fix the cone of blindness" bs and you want me to care about the 2047th bit ?
asciilifeform: whereas my algo preserves the entropy of the penultimate bit.
asciilifeform: that wouldn't win anything tho
mircea_popescu: i don't want the 1/4 bit of parity
mircea_popescu: so ? test for parity! throw out!
asciilifeform: asciilifeform's algo was 'if top bit of p*q not set, get new p,q'
a111: Logged on 2018-01-04 23:01 asciilifeform: why not stick to the ancestral 80col, folx??
mircea_popescu: http://btcbase.org/log/2018-01-04#1764342 << cuz i've been mostly the one guy reading her code to date i guess ? ☝︎
mircea_popescu: im not throwing them away 93% of the time
mircea_popescu: yes well, we're looking at ~10 second m-r run per item, so like half a day to produce a pair of primes.
mircea_popescu: faster this way.
a111: Logged on 2017-11-14 01:09 asciilifeform: the way i'd implement the whole shebang, is simply to reject both primes if the highest bit of pq is not 1 .
asciilifeform: so does 11... x 10... tho
asciilifeform: so why not, then, 3 ?
mircea_popescu: the ONLY way to make sure the top bit of a product is set, is to have BOTH top bits in BOTH factors set.
a111: Logged on 2018-01-04 22:30 asciilifeform: 'top 2 bits and bottom bit are ALWAYS 1!' << asciilifeform still doesn't get why to weld the next-to-highest
mircea_popescu: http://btcbase.org/log/2018-01-04#1764319 << twas exhaustively discussed last time! http://btcbase.org/log/2017-11-14#1737379 ☝︎☝︎
a111: Logged on 2018-01-04 21:16 phf: a forced popup that warns you that you're sending to http from https, but there's no about:config to disable the popup. 9 years of "ffs put a boolean in configs, you fucks"
mircea_popescu: http://btcbase.org/log/2012-10-15#-230312 << linked mostly for the bitcoin prices. ☝︎
mircea_popescu: pretty sure we lulzed at this exact item, years ago.
mircea_popescu: (and there was some idiot sheila from down below years ago ; i recall lulzing in the logs as to how this wonder will drive rapist to a) probe and b) beat into a squirming, faceless mess any bitch dumb enough to go around with it)
asciilifeform: mircea_popescu: subj is a type of boobytrap, vs freud's
a111: Logged on 2018-01-04 20:19 ben_vulpes: btw how come nobody in the "stop rape culture!!11" world has done an indiestarter to build the vagina dentata from neal stephenson's snow crash?
mircea_popescu: http://btcbase.org/log/2018-01-04#1764256 << let's leave it at that ☝︎
mircea_popescu: i'm too lazy to dig up where mp says he uses old amds, but w/e.
mircea_popescu: note to journahos of the future : include a plain statement of the above fact in your opening salvo or else.
mircea_popescu: who's ahead and who;s behind in the great technological race of 2010s ?
mircea_popescu: now let's see here : "Meltdown breaks all security assumptions given by the CPU’s memory isolation capabilities." vs "So given that there is in fact no secure memory implementation no matter how much it would be useful if there was one, EuCrypt takes instead the honest and practical approach of making it clear that it uses plain memory and nothing else."
a111: Logged on 2017-09-02 19:42 asciilifeform: biggest ice part is the ICE40HX8K
asciilifeform: http://btcbase.org/log/2017-09-02#1710096 << see also thread, re concretes ☝︎
mircea_popescu: asciilifeform that -- yes, can have, but cripled like so
a111: Logged on 2018-01-04 20:10 BingoBoingo: In other news from the price is right department: "The personal information of more than a billion Indians stored in the world's largest biometric database can be bought online for less than $8, according to an investigation by an Indian newspaper."
mircea_popescu: http://btcbase.org/log/2018-01-04#1764246 << this is a very apt commentary on the value of 1bn orcs piled up together. ☝︎
a111: Logged on 2018-01-04 20:10 asciilifeform: ( likbez : all you need for the mythical holy grail, 'fast iron rsa', is a very large-bitnessed adder-cum-barrelshifter and a few storage registers that can be programmatically shuffled between. )
ben_vulpes: i'm still waiting on my three secret beakers
mircea_popescu: http://btcbase.org/log/2018-01-04#1764241 << see ? three!!1 three secret retorts!! ☝︎
mircea_popescu: so to have closure, i suppose http://btcbase.org/log/2018-01-04#1763975 should read "80 cols plox what is this" ☝︎☟︎☟︎
phf: mircea_popescu: 80col is such an old holiwar you're risking myself or ascii moving in a backyard and building defensive possitions, like those boys they bring from iraq do. "son, you can come out now, we showed those no break somsabeaches"
asciilifeform recently reread 'tigerfibel', really a megaclassic of technical writing for all time ☟︎
ben_vulpes: other lulzies in the same vein http://btc.yt/lxr/satoshi/source/src/bitcoinrpc.cpp?v=makefiles#2418 and http://btc.yt/lxr/satoshi/source/src/bitcoinrpc.cpp?v=makefiles#0309
asciilifeform: mircea_popescu: recall tigerfibel ?
asciilifeform: ben_vulpes: make, why not. then it gets to be 2line
mircea_popescu: oh btw -- variable names like that fucking suck.
ben_vulpes: http://btc.yt/lxr/satoshi/source/src/bitcoinrpc.cpp?v=makefiles#1848 << and fuck you if you want to make a command with a name more than 1char longer than "walletpassphrasechange"
asciilifeform: where autoreflow does therighthing ~100%
asciilifeform: mircea_popescu: imho 1 lf per para is The Right Thing ~in humanolade strictly~
asciilifeform: phf: i force-format my ada. this only affects indents tho