#asciilifeform | 2019-12-03

snsabot: asciilifeform: time since my last reconnect : 30d 2h 43m

ossabot: asciilifeform: time since my last reconnect : 24d 5h 3m

ericbot: asciilifeform: time since my last reconnect : 19d 5h 10m

feedbot: http://qntra.net/2019/12/chicago-police-chief-fired/ << Qntra -- Chicago Police Chief Fired

asciilifeform: http://logs.nosuchlabs.com/log/asciilifeform/2019-12-02#1002992 << nm, found >> http://fixpoint.welshcomputing.com/2019/gales-bitcoin-wallet-status-preliminary-work-plan-and-code-dump/?b=Partia&e=dump#select

snsabot: Logged on 2019-12-02 21:46:36 asciilifeform: hey jfw : is your scheme wallet prototype published somewhere ?

jfw: asciilifeform: possibly textbook case for 'peh' from what I gather :)

jfw: but first I need to make it work at all before revisiting numerics.

asciilifeform: jfw: possibly. but i must admit that not yet tried to ecc under peh, there may be required number-theoretical knobs that are missing

asciilifeform: in re bitcoinism specifically -- the archaic hashes required are defo not in ('off the shelf') peh, likewise.

jfw: Nothing but grade-school arithmetic and modular inversion (which uses same) required for ECDSA. The 256-bit modular reductions are the slowest part for me afaik; hashing in Scheme slow too but not the bottleneck.

asciilifeform: theoretically oughta be easily 'pehable' then

asciilifeform focused exclusively on rsa, to date, so not had chance to attempt

asciilifeform: jfw: re hashes, commented on your www .

asciilifeform: jfw: re ecc -- part of why asciilifeform avoided to implement ecc (even tho ultimately it must be done, for a solid btctron) is that safety-critical mechanisms imho oughta be implemented by people who ~understand~ how they work. i understand how rsa worx. but ecc, presently, not.

jfw: ty for that comment asciilifeform, makes good futher context. Isn't RSA in that same snake-oil category though?

jfw: one can understand what it does but factoring not proven to be "hard"

asciilifeform: jfw: not quite same. rsa rests on specific conjecture.

asciilifeform: indeed , factoring not-proven to be hard.

asciilifeform: ( and , worse, rsa (unlike e.g. rabin's method) not afaik proven to rest entirely on the difficulty of factoring. )

asciilifeform: however it is a number-theoretical method, and to date survived 4 decades of number-theoretical approach. unlike the transpository shell games which , historically, fell like rotten trees at 1st serious push.

asciilifeform: merely attempting to do a serious analysis of hash or symmcipher is painful enuff that there's scarcely any public material worth the mention re subj.

asciilifeform: this for instance is the best crit i've found re keccak. and even it, half bilgewater.

jfw: do you reckon anyone understands ECC in the sense you have in mind, beyond "code does what the paper says it should and we can't see how to break it yet" ?

asciilifeform: re symmciphers, ~100% of academitard analysis specifically sits on 'reduced round' (i.e. crippled) variants rather than the real thing

asciilifeform: jfw: it isn't clear to me that anyone understands it, in that (yes, you correctly summarized) sense.

asciilifeform: rather, beyond 'this is equiv. to what paper specified' sense.

asciilifeform: jfw: ecc also has an 'interesting' political history (usg pushed it aggressively for ~20y, and w/ entirely fraudulent claims of '256b ecc is equiv to 4096b rsa' etc)

jfw: alright. Sometimes the reduced rounds findings add up to real breakage, like with sha1, right? (I'm not well versed in cryptanalysis whether bilge or pure)

asciilifeform: for all i (or anyone else) know -- bitcoin per se orig was a constructed bomb, by fella who knows the boojum for ecc.

asciilifeform: jfw: sometimes indeed, but very rarely author of such material substantiates the claim of 'adds up to real break'

jfw: 'we will all go together when we go' -Lehrer

asciilifeform: '...will all bake together when we bake..' indeed