asciilifeform: eats packet (on either end), looks at a few fields, either accepts (then rewrites a few fields and shits out of other arse) or rejects (drops on floor)
asciilifeform: router is a pretty simple item, incidentally, and a good first candidate for proper total deunixation
asciilifeform checks router in room and finds ~year of uptime, and would be longer if i hadn't cleaned it last winter
asciilifeform: i certainly can't bring myself to use the konsoomer crapola, when making a replacement that lasts ~forever and requires 0 nursing, is hardly even day's work
asciilifeform: when have complete proof, will say it in complete sentence!111!
asciilifeform: the one non-negotiable demand of usg.academia, is uselessness.
asciilifeform: asciilifeform for instance notices that his phone is not ringing with offers of tenure.
asciilifeform: apparently you dun get usg.tenure for working on proofs concerning items that map directly to iron
asciilifeform will come back to this item at some point , when it works its way out of his very long digestive tract
asciilifeform: again i dun have a verdict re the verdict of $subj, all i got in my notes is 'd00d appears to have stated the problem correctly'
asciilifeform: author ~does~ have some strange cockroaches in his head: cites shamir's 'proof that factoring can be O(log N)' but omits to mention that it requires a machine that works in arbitrarily-sized integers in constant time...
asciilifeform: hey if yer gonna take 'presently impossibly difficult' to mean same thing as 'uncomputable', then sha256 is goodenuff
asciilifeform: dunno that it is a conjecture when the working set consists strictly of items you can have on a physical comp
asciilifeform: the thing's been in my queue for a while, marked with 'determine what all of the assumptions were'
asciilifeform: quite possibly mircea_popescu read it at some point. but did not yet have the correct itch, to appreciate.
asciilifeform: sane discussion of the question, followed by an answer you can go with ( because if it turns out that oneway functions do NOT exist, all crypto other than otp is worthless )
asciilifeform: ( the flimflam artists who call themselves 'cryptographers' studiously avoid the subject, and will recoil in horror -- guaranteed, try it yerself -- when confronted. )
asciilifeform: afaik that piece remains 'the last word' on the subj, from the maths folx.
asciilifeform: note that, e.g., game of go, is exptime-complete but turns out that good algo exists. complexity class as we have it is a broken concept, which was asciilifeform's argument to start
asciilifeform: ( iirc last time we had the thread, mircea_popescu dug up a few even-moar-painful complexity classes, to place the dragon into. and asciilifeform had to point out that THEIR disjointness from P is equally unproven )
asciilifeform: or rather, requires a proof that P!=NP...
asciilifeform: as for the subj of thread, it would also seem to asciilifeform that it in fact reduces to the P =?= NP megapuzzler.
asciilifeform: the thing is, 'knows answer ahead of time' is not an all-or-nothing. in any non-otp ( i.e. 1:1 mapping of plaintxt to ciphertxt ) there is nonzero bittage of info in ciphertext, of plaintext
asciilifeform: but as i understand we asked for a squarer even circle : that 'CAN make grounded promise that effort will require AT LEAST x'
asciilifeform: incidentally i just realized that von neumann had this thread. and modelled the item in shannon's terms : he asked that the ciphertext contain 0 bits of info re the plaintext. and proved that this is true if and only if you're using... otp
asciilifeform: i've been asking for it since we 1st had the thread, lol
asciilifeform: i get this. and would like one of these. but strongly suspect that it is a provably square circle.
asciilifeform: i get this, this isn't the problem. problem is the dismissal of the knowing-the-answer degenerate case
asciilifeform: now you stand and fall strictly by rsa. P(rsabreak) is <= P(rsabreak OR whateverfucktardationyouusedforablockcipherbreak) , in all cases.
asciilifeform: now you stand and fall strictly by rsa. P(rsabreak) is <= P(rsabreak AND whateverfucktardationyouusedforablockcipherbreak) , in all cases.
asciilifeform: which is why i favour using rsa in place of blockcipher-hash-prng, painfully. the actual averagecase hardness of rsa is unknown and will probably remain unknown. but at least when you use ~solely~ rsa, you avoid introducing ANOTHER unknown.
