cgra: whaack: got it, thanks!
jonsykkel: http://www.loper-os.org/?p=2071#comment-18560 << had the same question, since X'Length = Y'Length was alredy done, the last condition is checking XY_Lo'Length against a number added to itself then divided by 2
jonsykkel: + are there solutions for the ch5 homevork problems anywher? prety sure my "optimizations" are wrong. curious about intended ones
asciilifeform: jonsykkel: the solution to that one is the subj of ch. 7.
asciilifeform: jonsykkel: thinking about it, i had misread the orig. comment by PeterL, didn't realize that he was speaking strictly of the last line ( XY_Lo'Length = ((X'Length + Y'Length) / 2)); ) when suggested to replace it. you're both right, the division there is superfluous.
asciilifeform: jonsykkel: the egyptian mul is of course not in current ffa, but the precondition is, i'ma fix it in the next ch. (when i get a chance... the coad's been written for more than a year, but not the text..)
dulapbot: Logged on 2021-08-30 16:23:38 asciilifeform: signpost: 'time problem' in fact in past yr got ~substantially worse~ for asciilifeform
asciilifeform: !!key jonsykkel
deedbot: Not registered.
asciilifeform: jonsykkel: now that deedbot is back, register a pubkey with it. that way i can rate you; you can offer folx vpatches when yer ready; participate in other civilized activities later
asciilifeform: !!help
asciilifeform: jonsykkel see ^ instructions.
asciilifeform: !!key cgra
asciilifeform: !!rated cgra
deedbot: asciilifeform rated cgra 3 at 2020/11/27 11:35:21 << careful, attentive student of FFA & TRB. inquire in #asciilifeform.
jonsykkel: http://logs.nosuchlabs.com/log/asciilifeform/2021-09-09#1057075 << ah, was looking for it in ch6. shud have realized from the chapter name
dulapbot: Logged on 2021-09-09 10:21:28 asciilifeform: jonsykkel: the solution to that one is the subj of ch. 7.
jonsykkel: "coad's been written for more than a year, but not the text.." << classic
jonsykkel: ill do key stuff, just i gotta make new ones first
asciilifeform: jonsykkel: writing a proggy is much easier than to prove that it worx, and then to write a readable exposition of the proof. ( which is why most folx stop after 'write proggy'... )
cgra: asciilifeform: how complete was the FFA picture in your head, before published ch1?
jonsykkel: asciilifeform: no doubt
cgra: asciilifeform: or even, how complete per se?
dpb: jonsykkel, check out http://atruechurch.info/ and don't go to hell like the rest of the world!
jonsykkel: looks great, will take a look
asciilifeform: cgra: had (mostly correct) coad & proofs through ch12 -- some of it pasted in #t log -- before ch1 'properly' published
asciilifeform: cgra: how complete per se, i.e. right nao? -- ch21c ( the coad , for extended-euclidean ) is done ; and also extensive revision of the peh interpreter (in particular, properly constant-time i/o of FZ)
asciilifeform: cgra: still remaining after this, will be hashing; and 'breath of life' (iirc discussed in old #t, the feature for placing tape execution params in the head of the tape per se)
asciilifeform: cgra: after this, examples of usage (generation of rsa keys; cramer-shoup; various practical examples)
asciilifeform: cgra: also would like to have a replacement for 'litmus' (gpg compatibility program, for use with e.g. vtron) that doesn't rely on bash kludge.
asciilifeform: possibly other missing things (e.g. at one time asciilifeform promised to include a proof that M-R works; but then put this off, and referred folx to schoolbooks)
asciilifeform: asciilifeform has two uses for ffa -- general-purpose sign/verify suitable for use w/ generalized vtron ; and for the ecdsa numerics in ada trb replacement.
dulapbot: Logged on 2020-03-19 12:32:20 asciilifeform: why the hell is a typical public www bound to 1 box? specifically, why could not exist as an auto-mirrorable directory of signed material ?
dulapbot: Logged on 2021-08-29 21:07:32 asciilifeform: jonsykkel: see linked thrd for basic outline of algo. idea is, if you have a www ('vvv' ??) of signed material, will then have as many mirrors as you have readers. and they'll all have up-to-date, authentic content of yours, trivially verifiably. nomoar ddos, expensive hostings, warez takedowns, etc
dulapbot: Logged on 2020-10-01 15:00:29 asciilifeform: gregorynyssa: there's not a thing that'd rise to the level of 'plan'. asciilifeform in particular wrote 3 things: 'nqb', a largely-complete coder/decoder for the formats used in trb; ffa, with which possible to perform the cryptonumerics; and 'cryostat', to implement a o(1) db .
asciilifeform: other people -- may have yet other uses.
cgra: asciilifeform: the last time i spent time following your ffa curriculum, i felt a need to go a full cycle before returning to ch1, for a review, and until then gaining a proper grasp. now figured must've been similar for a writer, too... fits-in-head after all
asciilifeform: cgra: indeed was; in fact i deliberately started w/ simple (but slow) 'egyptian' algos
asciilifeform: and then went back, replaced one by one with 'adult'. and reproduced this path for the reader.
asciilifeform: ( why did this? asciilifeform wanted to see whether and if so how much optimization is ~actually~ required for reasonable performance on pc on real-life key sizes )
asciilifeform: sooo, for instance, found that the most common heathen 'optimization' -- permitting variably-sized bignums and early termination in iterations -- aint needed at all !
asciilifeform: ( on account of branch prediction mechanics in modern irons, ffa in fact faster than koch's bignum, for the latter's 'worst-cases' )
signpost: jonsykkel: gimme a shout if registering gives you any issues.
asciilifeform: wb signpost !
asciilifeform: signpost: didja get a chance to read asciilifeform's backbreaker ? ( current draft here, for ref. )
dulapbot: Logged on 2021-09-06 00:50:19 signpost: hail satan. can't wait to read this over coffee tomorrow asciilifeform!
asciilifeform: ( and further discussion re same )
asciilifeform must bbl
jonsykkel: will do signpost, ty
jonsykkel: speaking of keys, whats the #a consensus on least retarded gnupg version?
jonsykkel: saw some older ones had this bug affecting key generation
asciilifeform: jonsykkel: you're probably thinking of this one
asciilifeform: jonsykkel: if you're making a 4096b or bigger, and using 1.4.x ( e.g. asciilifeform -- 1.4.10) -- will probably want the patch.
asciilifeform: (koch's key gen is still imho rather shoddy even with fix for this bug tho.)
asciilifeform: for this & related reasons, asciilifeform intends to retire his gpg key (i.e. will promise not to emit new sigs w/ it) after completion of ffa/peh and signing his brand-new peh key w/ the old gpg pub.
asciilifeform genuinely bbl
jonsykkel: thats the one
jonsykkel: i see, guess ill make sure itis patched at least
signpost: asciilifeform: haven't fully digested yet; will try to respond sometime saturday
signpost: definitely on my stack tho
asciilifeform: signpost: a++.
asciilifeform: signpost: there's no reason to hurry.
asciilifeform: meanwhile, in sads, asciilifeform dusts off priceless 'trinitron', and picture wobbles. no known cure.
dulapbot: (trilema) 2015-06-14 assbot: Brand-New Trinitron (circa 2000) - Unboxed in 2015. - Album on Imgur ... ( http://bit.ly/1dFoDwJ )
asciilifeform: ( incl. in debug menu, so it aint the comp. nor the mains power, it was plugged into a lab-grade sinusoidal doubleconverter )
asciilifeform would happily pay, say, $10k, for a ~new~ one, but they physically dun exist
asciilifeform: this item, 'gdm-5402', has barely 10 hours on it!
asciilifeform was not able to find or even conceive of a physical explanation for such wobble, given that electrical/magnetic interference rules out, and unit essentially virginal.
asciilifeform: *ruled
verisimilitude: http://logs.nosuchlabs.com/log/asciilifeform/2021-09-08#1056996 I noticed, but forgot to mention it.
dulapbot: Logged on 2021-09-08 13:28:30 asciilifeform: re 0xFE -- lulzily no one noticed that 4.1 and 4.2 numbering is entirely fucked.
scoopbot: New post on A Syndication of Verisimilitudes: Integrity with One-Time Pads
vex: condolences re: sony alf
vex: 10k is almost worth boots on ground in tokyo
asciilifeform: vex: i dun think you get my meaning -- there aint any in tokyo just as in zimbabwe. last 1 was made in '00ish.
asciilifeform: 20y seems to be at the outer limit of their ~storage in sealed crate~ lifespan.
vex: some greybeard in one of the millions of laneways prolly has one in box collecting dust.
vex: finding- different story
asciilifeform: godfuckingdamnit
asciilifeform: THIS ONE WAS SEALED motherfucker
vex: oh
asciilifeform: and worked 6y ago
asciilifeform: shat self ~in climate-controlled storage~
dulapbot: Logged on 2021-09-09 23:13:56 verisimilitude: http://logs.nosuchlabs.com/log/asciilifeform/2021-09-08#1056996 I noticed, but forgot to mention it.
asciilifeform: http://logs.nosuchlabs.com/log/asciilifeform/2021-09-09#1057145 << verisimilitude have you never heard of error correcting codes, or simply for some unclear purpose pretending not to have ?
dulapbot: Logged on 2021-09-09 23:26:53 scoopbot: New post on A Syndication of Verisimilitudes: Integrity with One-Time Pads
vex: di you try a fistful of g-forces?
asciilifeform: did, after all else. 0diff.
vex: expected
vex: didya ask sony? they might like to see a wobbly one