Show Idle (>14 d.) Chans
← 2019-01-04 | 2019-01-06 → ↓
mircea_popescu: well, meanwhile it's the 5th day of this fine new year, and it occurs to me it might be a good idea to discuss some things, lest april coming around this year like any other year be perceived as some kind of subjective surprise, rather than the objective necessity it ever is.
mircea_popescu: foremost, the lordship is not a state, but an activity.
mircea_popescu: so : as far as i know, bingoBoingo is working on qntra and on pizarro. he's doing a very fine job with the former ; i'm nonplussed with recently discovering just how broken the latter's mp-wp offering actually was ; moreover it seems to me from a distance pizarro's still financially and customer-wise entirely dependent, ie as close to failure as you can possibly get without spelling it out.
diana_coman: fwiw I can't say I saw a surprise so far in any of the april announcements; at most more of a difference of degree at times (i.e. the expected direction/action but to a larger/smaller degree)
mircea_popescu: yet -- it is not my business ; so i'm just going to count it as such, "working on pizarro" without further inquiry (and with the firm expectation that someone IS doing all that).
mircea_popescu: diana_coman well hopefully thusly we smooth that out.
diana_coman: I'm all for the talk, certainly
mircea_popescu: diana_coman is working for s.mg ; we've recently had this exact talk and revised our plans. originally the idea was to have moved over to cuntoo, and do support work for community-driven effort at a new client. the latter completely collapsed over the shocking weakness of such community ; the former's at best delayed.
mircea_popescu: so, she'll be doing client work with her own two hands, something i had every intention to avoid ; and we'll be looking at integrating cuntoo on the server side later on. there's some db work in the hopper also, but that's even further on the maturity vine.
mircea_popescu: hanbot is working on the mp-wp tree, and manages as she long has a rather largeish wetworks i'm not going to get into the details of.
mircea_popescu: other than maintaining the deedbot infrastructure, trinque is working on cuntoo, which is a rather large piece and it taking a [difficult to predict] while is not by itself the end of the world ; but i'd like to see some roadmapping, tentative and subject to change as it may be, lest the effort degenerates.
mircea_popescu: finally, asciilifeform is working on rsa-based ssl-ism replacement (notwithstanding he ~seems to be~ working on any and all wank on the "side" during spare time he doesn't have and all that), which we want so we can finally move bitcoin off sheer cretinity and into cuntoo (and which is principally why we want sane db also, but as i said -- yet immature).
mircea_popescu: other than this, mod6 is taking time off, since mid-november. nothing wrong with this, but i'd like to see some conclusions at some point.
mircea_popescu: danielpbarron seems lost in a bucolic paradise slash midwestern parochial nightmare of his own choosing. nothing wrong with this either, and unless i hear any better that will be the conclusion.
mircea_popescu: i have no idea what ben_vulpes is doing these days, if anything (other than maintaining his logotron, whatever that takes). i would like to hear.
mircea_popescu: lobbes recently unveiled actionbot, which works fine, and is evidently putting all time he can into paying off technological debt he's responsible for if not necessarily guilty of. nothing wrong with this, and it can stand as such.
mircea_popescu: spyked is evidently trying, hence feedbot, but evidently having trouble reconciling saeculum, which i'm going to let stand as such on the grounds that he's new -- even though experience shows that as a dubious idea [for all the eg one could possibly need witness how asciilifeform 's still in the swamp, so many years later].
mircea_popescu: let's just say it's my considered oppinion "difficulty in reconciling" such is almost never actually due to said saeculum.
mircea_popescu: ave1 is, i suspect, silently working on gnating things -- which is fine and valuable except for the silently part. there's this tendency of lone wolf scientist to not properly report failures, out of an imaginary saving of time and resources this permits. it must be said that NOTHING could be further from the truth, nothing at all -- there's more to be gained from a properly reported failure to find than out of ten shiny succ
mircea_popescu: but really, it serves no one to let me suspect. i should know.
mircea_popescu: which leaves phf ; who's doing an excellent job maintaining very elegant and well done extant infrastructure, and a very terrifyingly poor job at communicating himself.
mircea_popescu: i do not know, as i sit here, what the conclusion of the http://btcbase.org/log/2018-12-19#1881988 discussion was, three weeks later. the man sits among a republic built on a certain methodology, aims to quietly use the exact ~other~ methodology, makes no prior mention of this, candidly references the latter as if it worked notwithstanding how howlingly it failed to work to date, on and on in this vein and at the end of it al
a111: Logged on 2018-12-19 20:09 phf: you're constantly in logs, confused as to how anyone can get anything from anyone, yet now you're questioning my methods.
mircea_popescu: l i have not the faintest what even came of the long avoided but eventually unavoidable discussion.
mircea_popescu: so, phf : how about you start clearly communicating yourself, beginning with a complete, correct and true to life adnotation of said discussion in your own hand, because this "ima go meditate on things until everyone involved forgot what i was meditating on" isn't a workable approach to intellectual life.
mircea_popescu: this about concludes the state of affairs ; i'd like nothing more than corrections, disputations &c by they involved.
asciilifeform: http://btcbase.org/log/2019-01-05#1884612 << fwiw i've ~exhausted the short-term milk of the particular side cow ( there's a disk snapshot & a recipe to post, but after that will be stalled for aeons ) ;
a111: Logged on 2019-01-05 14:22 mircea_popescu: finally, asciilifeform is working on rsa-based ssl-ism replacement (notwithstanding he ~seems to be~ working on any and all wank on the "side" during spare time he doesn't have and all that), which we want so we can finally move bitcoin off sheer cretinity and into cuntoo (and which is principally why we want sane db also, but as i said -- yet immature).
asciilifeform: currently focused on ch15.
asciilifeform: orig release sched is blown, asciilifeform's current desire is to get again to the point where can patch its tyre and reinflate and have rel sched again
asciilifeform: ( in orig timeline also did not include problem of constant-time keccak, which i presently do not have, and neither anyone else, but is necessary to fill mircea_popescu's spec for the final product )
asciilifeform: http://btcbase.org/log/2019-01-05#1884623 << at the risk of pouring petrol into that particular fire, fella also promised a http://btcbase.org/log/2018-10-06#1859075 item iirc
a111: Logged on 2019-01-05 14:37 mircea_popescu: i do not know, as i sit here, what the conclusion of the http://btcbase.org/log/2018-12-19#1881988 discussion was, three weeks later. the man sits among a republic built on a certain methodology, aims to quietly use the exact ~other~ methodology, makes no prior mention of this, candidly references the latter as if it worked notwithstanding how howlingly it failed to work to date, on and on in this vein and at the end of it al
a111: Logged on 2018-10-06 23:51 phf: asciilifeform: i don't have binary diffing even in prototype form, if you could adaize your needleman-wunsch i could add it to vtools, the way i did with diana_coman's keccak
asciilifeform: err, bin ~delete~ knob
mircea_popescu: he promised to ask you for an item ? or is this ada n-w published somewhere meanwhile ?
asciilifeform: i have wrong link, 1s
mircea_popescu: maybe it;s not as wrong as all that. did you ever make that thing ?
asciilifeform: bin differ ? yes. but it was never made to current-day tmsr quality , i shelved it when mircea_popescu said it wasn't particularly useful
asciilifeform: ( and i currently agree, really one ought not to have bins in vtrees )
asciilifeform: i posted a complete needleman in cl ( also a draft, rather than troo genesis, nobody stood up and said 'i want this' so shelved ) 2y ago, also.
mircea_popescu: the ~algo~ tho. not specifically for v trees, but it strikes me there doesn't exist currently a bin differ ~at all~.
asciilifeform: mircea_popescu: http://btcbase.org/log/2017-12-29#1760563 ( napkin-level but working item i posted on prev occasions )
a111: Logged on 2017-12-29 05:34 asciilifeform: >>>>> https://archive.is/jMMqT
asciilifeform: and yes it worx. you give it a similarity matrix (i.e. 'what differences are important, in the order of their importance' in matrix form) and it produces an alignment.
mircea_popescu: mkay. in the immortal words of that fellow, http://btcbase.org/log/2018-12-12#1880371
a111: Logged on 2018-12-12 19:41 BingoBoingo: ^ Any awk ninjas want to try this on a few forums?
asciilifeform: mircea_popescu: if you have a concrete application in mind, i can spare an hour or 2 to bake
mircea_popescu: alright ; we leave it like this for now.
asciilifeform: algo's pretty simple btw, 1970s thing ( tho iirc it aint in knuth aop for some reason )
asciilifeform: grr, was looking for mircea_popescu's also for vtron file movements, still not found in log
asciilifeform: the 1 phf went to implement and not yet came back with.
asciilifeform: ( phf if yer stuck, pleez say on what ? )
a111: Logged on 2018-04-03 00:02 mircea_popescu: asciilifeform the logged discussion on the topic was, "if hashes match but paths do not, the file was moved ; if hashes match and paths match, the file is untouched ; if hashes do not match but paths match the file was modified ; if hashes do not match and paths do not match the file was created/deleted"
asciilifeform: there we go.
asciilifeform: ty mircea_popescu
a111: Logged on 2019-01-05 14:17 mircea_popescu: so, she'll be doing client work with her own two hands, something i had every intention to avoid ; and we'll be looking at integrating cuntoo on the server side later on. there's some db work in the hopper also, but that's even further on the maturity vine.
a111: Logged on 2018-11-16 23:13 asciilifeform: it is on hold pending resolution of http://btcbase.org/log/2018-10-26#1866266 ( and is taking back seat to ffa currently )
asciilifeform: ( possibly diana_coman will answer this when wakes up )
a111: Logged on 2018-10-26 02:14 asciilifeform: meanwhile, in gnat bugs : apparently ( and this is documented or mentioned nowhere ) : it is impossible to have a Ada.Finalization.Limited_Controlled type ANYWHERE inside a static library, unless it is generic all the way down (i.e. if the lib package is generic, any sub-packages must also be instantiated as generics )
asciilifeform is refreshing chalkboard, would like to get a sense of the set of items he personally owes
mircea_popescu: asciilifeform not ~just~ on that. also on http://btcbase.org/log/2018-12-04#1878240 and on a larger pile of [not entirely specified, and i suspect not entirely known] edges.
a111: Logged on 2018-12-04 15:14 mircea_popescu: it's not exactly clear to me yet what the situation is. it's altogether possible postgres may be rescuable through a process similar to how "peculiar linux candidate packaging sterilized into cuntoo".
asciilifeform: right; i'm mapping out which 'known and unknown edges' are mine
asciilifeform: if diana_coman is 'stuck on X, cannot move' and it is asciilifeform's X, i'd like to know about it asap.
mircea_popescu: certainly.
asciilifeform: since we're refreshing chalkboards in war room, i'ma take the chance to summarize current level of ffaism. currently exponentiator is mature ( aside from the consideration where http://btcbase.org/log/2018-12-28#1883687 , i.e. it's mature when ~other~ thinkingpeople have fit it into head and concurred ) -- i.e. runs fast enuff for many battlefield applications
a111: Logged on 2018-12-28 20:35 asciilifeform: diana_coman: sadly i do not know how to 'guarantee perfection'. all i know how to do is to bake maximally 'fits in head' and bank that the folx here will find mistake if it turns out that i made one.
asciilifeform: what remains is 1) prime-baking 2) rsa (and similar cryptosystems, tho c-s dun need it) padtron -- requires constant-spacetime keccak 3) optional asmistic speedups.
mircea_popescu: wait, when did we move off oaep for rsa padding ?!
asciilifeform: we haven't
asciilifeform: that's item 2.
mircea_popescu: why does it need keccak ?
asciilifeform: cuz that's how mircea_popescu specified the padding
asciilifeform: diana_coman implemented prototype, using koch
mircea_popescu: ima go have to re-read this now
asciilifeform: rsa-oaep requires hashtronics for padding.
asciilifeform: correct
mircea_popescu: bitcoin, of course, needs this at no juncture.
asciilifeform: otherwise all of the nonleakage guarantees bought at the cost they were bought at, vanish.
asciilifeform: bitcoin dun use rsa at all, at least in classical variant of bitcoin
asciilifeform: it needs an entirely other item ( which can be sewn from ffa parts, but has not been of yet )
asciilifeform: i'ma prolly have to do ~this~ with own hands, too -- no one yet stood up and said 'i will'
mircea_popescu: point being : de-sslification of bitcoin is not stalled on this. it is stalled on prior de-sslification of cuntoo, which as such doesn't yet exist, which is not really stalled on this either.
mircea_popescu: but rather, from what i understand, is taking finishing touches re reproducible builds, consisting principally of hunting for "Where does it piss in date or w/e ruining my sigs"
asciilifeform: mircea_popescu: correct. the item that needs padtron, is mircea_popescu's specced 'fuckng replace gpg already' ; and possibly also koch-free euloratrons.
mircea_popescu: in the sense eucrypt uses mpi you mean ?
asciilifeform: also recall the (surprising to asciilifeform , but apparently nobody else) discovery that ffatron as-is-stands is ~2.5x faster than koch.
mircea_popescu: right. a mpi-eucrypt vs ffa-eucrypt head-on will be interesting to see.
asciilifeform: ( otoh euloratron does not spend much cpu in rsa, as currently sewn )
mircea_popescu: asciilifeform it doesn't ; nor will it, because what truly brings serpent in is the ~space~ not the time problem. ie, because of padding, straight rsa doubles message bulk, which is a major problem for online game.
asciilifeform: this is entirely tru
mircea_popescu: but the fact that we have the eucrypt item is very useful inter alia especially for such simulated-userland-tests for core libs.
asciilifeform: well until last wk it was the only rsatron we had that ran in something like realtime
mircea_popescu: esp because correctly written, with tests etc. so can meaningfully do ffa-eucrypt vs mpi-eucrypt as a benchmark.
asciilifeform: ( diana_coman's , that is )
asciilifeform: and as of right nao it's the only ~complete~ rsatron we have, i.e. that knows how to bake privkeys
asciilifeform: asciilifeform's current thrust is to catch up to ~that~
asciilifeform: ( when i have ~complete~ sys, ~then~ mircea_popescu can go 'hmm, which one to sew ~here~' etc )
mircea_popescu: the more toys to play with, the more playing to be had
asciilifeform: tbh i'm not sure what kochtronic rsa will be good for once i have the keygenning ( it apparently dun win on speed anywhere, even tho it gets to skip 0s in modexp.. ) but this time not yet come.
asciilifeform: the 1 application where ffa defo dunwork, and koch -- does, is phuctor.
asciilifeform: ( i dun have e.g. strassen's multiplication algo, and dun have any plans to implement, it only wins for multi-megabyte ints )
mircea_popescu: irony of ironies.
diana_coman: asciilifeform, what it's good for is at least ...for contrast!
asciilifeform: hey , partizan is perfectly happy to use schmeisser taken off dead german.
asciilifeform: ohai diana_coman
diana_coman: and yes, I'm eating up ffa with an eye on "maybe I can finally get rid of MPI!!"
diana_coman waves and catches up on logs
asciilifeform: diana_coman: i described in this log what currently stands between 'throw out mpi' . lemme know if needs moardetail.
asciilifeform: ( tldr -- asciilifeform needs : wrapup of gcd; then miller-rabin; then keccak. )
asciilifeform: mircea_popescu: possibly i ought add : ~mpi~ dunhave strassen. ~gmp~ (the older, 'uncut' gnu thingie) has strassen.
asciilifeform: mpi had only (ugly as fuck) karatsubatron.
asciilifeform: i've been referring to mpi and gmp interchangeably as 'koch rsa', but this is unscientific, i must remind that they are diff items.
mircea_popescu: what we call mpi is closer to gmp than what the retard crowd does anyway.
asciilifeform: mircea_popescu: mpi is subset of gmp that koch cut ( and ate $mil of microshit payola to do it, somehow ) , aha.
diana_coman: asciilifeform, thing is: from eucrypt and eulora pov, mpi is used for "big num arithmetics" only so I CAN in fact switch to ffa even without ct-time miller-rabin esp if ffa turns out to be...faster than mpi
diana_coman: basically there is no reason NOT TO
asciilifeform: ( phuctor, ftr, uses a (patched, to enable bigger ints) old gmp. with asmisms enabled. )
asciilifeform: diana_coman: it still needs m-r
asciilifeform: ( as well as still needs diana_coman to eat & digest the thing per se )
asciilifeform: diana_coman: my current understanding is that in fact it's ~2.5x faster per bit of exponent length.
asciilifeform: ( i.e. one would have to put in mircea_popescu's specced exponent bitness where 'Bitness' is in http://www.loper-os.org/pub/ffa/hypertext/ch14/fz_modex__adb.htm#85_14 , to get the speedup )
asciilifeform: ( as well as adjusting 'Wi' , etc. but you get the idea )
diana_coman: asciilifeform, re m-r: I implemented it using mpi as per http://ossasepia.com/2017/12/28/eucrypt-chapter-3-miller-rabin-implementation/ ; ofc I'd rather use ffa ct-time implementation but it's not a sticking point per se i.e. I can switch my implementation from relying on mpi to relying on ffa, no?
asciilifeform: diana_coman: correct, you in fact could
asciilifeform: ( if asciilifeform is flattened by an anvil tonight, i'd expect that diana_coman and mircea_popescu will do something of the kind. )
asciilifeform: otherwise can wait for asciilifeform's constant-time m-r ( or not, depending on what's in eulora war room chalkboard, i cannot presume to know what the priority is )
diana_coman: well, don't get flattened please, there's already waay more work than active hands as it is
asciilifeform: i try to avoid anvils, so far succeeded..
mircea_popescu: diana_coman nothing wrong with ~testing~ it at some point, anyway.
diana_coman: mircea_popescu, exactly
mircea_popescu: put some meat on the bones of his "check it out, speed!" thing, as a courtesy if nothing else.
asciilifeform: mircea_popescu: thing was built to be laughably easy to glue to other programs. (if it ain't -- i'd like to know asap why )
mircea_popescu: and yes, in the process find whatever else.
diana_coman: asciilifeform, sometimes I wonder what exactly do you think you need/don't have to move to Romania or wherever else you consider it to be "paradise, can now do just ffa/trb/..."
asciilifeform: diana_coman: moola
asciilifeform: i'd expect that it would cost 1 or 2 diana_coman-days to glue ch14 ffa to euloratron, to see how performs using diana_coman's existing m-r etc.
diana_coman: well yes, but specifically: how much would be "enough"?
asciilifeform: diana_coman: pretty sure i discussed this depressing subj in l0gz prior, would rather not clutter war room log with replay; but it'd have to be enuff to buy new passport, at the very least, and then would need to earn bread somehow ( and as sit-in-torture-room, rather than driving cab , i dun think i'll be of much use to tmsr if living in cab )
mircea_popescu: diana_coman amusingly, spending 50k or w/e it costs to get an old farm (including barn) in whatever, fucking alba county, go walking on the hillsides with the cowsies, would take him to his paradise, yet there he toils.
asciilifeform: mircea_popescu: fields plow themselves in ro, or how does that work , lol
mircea_popescu: how does what work ?
asciilifeform: barn-to-money thermodynamic machine
mircea_popescu: farming is not an economic activity in europe ; nor has it for half a century or more.
asciilifeform: i didn't think so. hence astonishment.
mircea_popescu: i don't know what you're astonished at ?
asciilifeform: what would '50k barn' do for asciilifeform?
mircea_popescu: it'd pay your rent, and most of the cost of your http://btcbase.org/log/2014-10-15#876029 budget.
a111: Logged on 2014-10-15 19:58 mircea_popescu: asciilifeform to follow that model. qntra isn't a seller of 0-dioxin, 1814-equivalent tomatoes, but a greengrocer where they don't hire retards and don't mix shit in the fruit sala.d
diana_coman: mircea_popescu, that's precisely why I asked for concrete sum because the way I see it...he has it
mircea_popescu: specifically, a 5 to 10k% increase in deliverables over what you currently achieve.
asciilifeform: diana_coman: a ro passport sells for 800k usd, last i inquired
diana_coman: why do you need the passport *first*? (i.e. not just get one after whatever hoops are in ro, living there for x years etc)
mircea_popescu: most romanians don't have a passport either.
asciilifeform: so as to have gedanken-farm in own name, say ? or to be able to occasionally cross borders without using u-boat , say ?
mircea_popescu: so if anything, not having one makes you MORE romanian, not less.
asciilifeform: ( otherwise gotta add cost of u-boat.. )
mircea_popescu: asciilifeform you can cross border on your current one also lol wtf.
asciilifeform: not once usg cancels it for unpaid tribute, neh
mircea_popescu: you know for a fact hanbot lived there for what, better part of a decade. wtf 800k passport.
asciilifeform: hanbot had a patron with castle, neh.
mircea_popescu: well if usg cancels it you're then a refugee. they love those in yurp lol
mircea_popescu: apply for a stateless person thing./
asciilifeform: theoretically great, in practice they seem to send those back in chains to washingtonschwitz if they're subjects of trumplincton rather than africa
mircea_popescu: your problems are in your own mind.
mircea_popescu: wth, most everyone travels, you go visit the surrounding places now and again, takes care of the whole problem for you.
asciilifeform: sorta why i dun go around expecting other folx to solve'em for me. i solve with own hands ( or not, if not live longenuff )
mircea_popescu: you don't really want to winter in romania anyway.
asciilifeform: so happens that i like winter on that parallel
asciilifeform: it's my home parallel after all.
mircea_popescu: then you don't wanna summer there.
asciilifeform: do they have mosquitoes yet ?
mircea_popescu: depends on exactly where, but notrly afaik.
asciilifeform: i promise to come back to this thrd, if asked. but would like to not lose the subthread earlier : diana_coman mircea_popescu : is short-term plan to test ffaistic diana_coman-r-m ? and if so, what glue is needed for this from asciilifeform , i'ma bake.
mircea_popescu: i imagine she's going to bake a test as time permits. it's not a top priority item but then again she moves fast.
asciilifeform: i dun expect it will need many litres of glue.
mircea_popescu: might want to read through it / see it finished first / we. i'm not squeezing it.
asciilifeform: what i prolly oughta roll into the conveyor, is a variant of ffacalc that's libraryized (i.e. callable from other program, with string argument containing pcode, and fills a provided buffer with the output)
diana_coman: asciilifeform, first I do need to finish getting the ffa in, so that will still take quite a while; other than that, it's more a matter of "as time permits" and as mircea_popescu says it's not top priority; that being said yes, I'd like to do it and see some timings and comparison for myself
asciilifeform: diana_coman: btw it is perfectly ok also to simply invoke the knobs exported in ffa.ads directly, but then you gotta take care of 1) endianism of the words being put in and gotten out , to match yours 2) testing for div0 , as done in http://www.loper-os.org/pub/ffa/hypertext/ch14/ffa_calc__adb.htm
asciilifeform: ( if you use pcode, these get done inside ffacalc, and ditto the stack memory mechanics, as you prolly knew already )
diana_coman: will keep in mind, ref the log when I get there; and definitely come and shout at asciilifeform when/if I get stuck on something related to this
asciilifeform: user of pcode never has to manually consider memory, so long as he knows how much stack to instantiate it with ( e.g. for modular exponentiation, you need 3 FZ worth of stack ) , and it properly eggogs if you mismeasure.
asciilifeform: so, to illustrate:
asciilifeform: !!up pehbot
deedbot: pehbot voiced for 30 minutes.
pehbot: asciilifeform: FFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFF
asciilifeform: !A FFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFF0
pehbot: asciilifeform: EGGOG: Pos: 0: Stack Underflow!
asciilifeform: !A .FFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFF0
pehbot: asciilifeform: EGGOG: Pos: 65: Constant Exceeds Bitness!
asciilifeform: ^ there we go
asciilifeform: or, e.g.,
asciilifeform: !A .1.1.1MX (this is ok )
pehbot: asciilifeform:
asciilifeform: !A (but this aint ) .1.1MX#
pehbot: asciilifeform: EGGOG: Pos: 22: Stack Underflow!
asciilifeform: and so forth.
asciilifeform: ( this being for the given bitness, which in pehbot is hard-welded to 256 )
asciilifeform: and likewise, you can put as many fz on stack as the stack height given in invocation of ffacalc, but not any moar.
asciilifeform: ( i'ma stop here, folx who ate ch4 know all of this kindergartenisms )
asciilifeform: ffa is designed to be used via pcode (aka 'peh') but i'm not about to tell folx that they absolutely must. given the stated reqs (you gotta test for div0ism, we dun do it internally given as it's thermonuke performance) it can be safely used directly.
asciilifeform: ( for the l0gz, refresher: pcode is meant to give a mechanically simple system where yer privkey is a pcode string, and so is yer pubkey, and so are ciphertexts, and whole mechanism is set in motion simply by feeding pcode to the processor )
diana_coman: well, I hardly see how you can *stop* people from using it directly or why exactly; and the endian + div0 don't sound like a huge layer anyway
asciilifeform: rright, but i dun necessarily mean to even discourage people from using directly.
asciilifeform: so long as they know which end of the rifle bullet goes out from.
diana_coman: well, if they don't, they'll find out quickly :|
diana_coman goes back to chopping client "code"
asciilifeform: ( using it directly, is sorta analogous to https://www.youtube.com/watch?v=A4QagenzsK0 . can be done... )
asciilifeform: http://btcbase.org/log/2019-01-05#1884619 << from ave1 , i hope to see a 'port' of tmsr-gnat that can be hard-welded into cuntoo as primary gcc ( to remove the hack where it builds gcc5, then down to 4.9, and neither of'em being a gnat )
a111: Logged on 2019-01-05 14:32 mircea_popescu: ave1 is, i suspect, silently working on gnating things -- which is fine and valuable except for the silently part. there's this tendency of lone wolf scientist to not properly report failures, out of an imaginary saving of time and resources this permits. it must be said that NOTHING could be further from the truth, nothing at all -- there's more to be gained from a properly reported failure to find than out of ten shiny succ
asciilifeform: also at some point it'd be great to have a mips gnat, so i can plant ffa on pocket-sized irons. but that's for muchlaters.
asciilifeform: ( already hats off to ave1 , who did year+ of gnat cleanup that asciilifeform was solidly convinced he'd have to do with own hands; and the fixed inlining gave us a ~2x ffa speedup 'for phree' ! )
asciilifeform: http://btcbase.org/log/2019-01-05#1884626 << i suspect phf is hunting elk in kamchatka, or similar , atm ( i.e. still waiting for http://btcbase.org/patches?patchset=ffa refresh , so i dun think he's been at console much recently )
a111: Logged on 2019-01-05 14:38 mircea_popescu: so, phf : how about you start clearly communicating yourself, beginning with a complete, correct and true to life adnotation of said discussion in your own hand, because this "ima go meditate on things until everyone involved forgot what i was meditating on" isn't a workable approach to intellectual life.
asciilifeform: hopefully he comes back soonish ( i'ma not even pester him re bolix, bolix is asciilifeform's personal war, learned not to expect help from anyone, and currently in refrigerator, i dun expect to spend much time on it while ffa undone )
asciilifeform: i admit that i'm at least a little curious how phf finally managed escape velocity from the bigzone, but if he doesn't feel like spilling re subj, also won't cry.
asciilifeform: http://btcbase.org/log/2019-01-05#1884603 << BingoBoingo i'd ~really~ like to hear what is current plan for gettin' heathen custom, so as to finally get the hell out of the red. asciilifeform dun have a massive treasure chest that can run pizarro 'on battery' 4evah (hopefully not surprising, this)
a111: Logged on 2019-01-05 14:13 mircea_popescu: so : as far as i know, bingoBoingo is working on qntra and on pizarro. he's doing a very fine job with the former ; i'm nonplussed with recently discovering just how broken the latter's mp-wp offering actually was ; moreover it seems to me from a distance pizarro's still financially and customer-wise entirely dependent, ie as close to failure as you can possibly get without spelling it out.
asciilifeform: and without pizarro, it will be very very cold and dark and we'll be drifting in the unforgiving vacuum of interstellar space.
asciilifeform: mircea_popescu: any idea what happened to esthlos ?
asciilifeform: !#seen esthlos
a111: 2018-10-23 <esthlos> http://btcbase.org/log/2018-10-19#1864316 << apologies alf, I'm running behind! trying to gather time to get caught up in the next week or two
asciilifeform: to round out the 'loose ends' thread -- asciilifeform also has a ~90% built node-walker and www front end for same. but it is in refrigerator, no one is direly starving for the lack of the thing, i expect i'll come back and finish it off strictly after ffa is fielded .
asciilifeform: there is also a serpent-on-ice40 thing, with similar level of unfinishitude; and a ice40-powered 'FG2', ditto.
asciilifeform: ( there is also a 'giant ice40' that amberglint dug up recently, that gotta be tested, but i dun even physically have 1 yet, and deliberately not bought so as not to distract from moar urgent matters )
asciilifeform: to entirely round off the chalkboard, mircea_popescu may be interested to know that no one has called the 1-800 since we last spoke of it. the current summed cost stands at 12 orcbux, and is set to increment by 22 orcbux/mo ( we've spent the vendor's 'test drive' bait). i'ma cover the lunch money cost of this item until given to know that it aint wanted.
asciilifeform: this concludes the 'state of asciilifeform' broadcast for nao -- plox to lemme know if i missed sumthing.
asciilifeform will bbl, maintaining meat systems
BingoBoingo: <asciilifeform> and without pizarro, it will be very very cold and dark and we'll be drifting in the unforgiving vacuum of interstellar space. << Today crunching monthly numbers. Followed on the agenda by reviewing 2018/putting forth tentative plan based on those lessons 2018
lobbes: http://btcbase.org/log/2019-01-05#1884616 << imo, this is a perfect summary of my current state. I walked through the tmsr doors in ~2014 at roughly epsilon and 'learned as I went'. As a result, many of my projects here were built on unsteady scaffolding, and I have been slowly going back and pouring in proper foundations where needed
a111: Logged on 2019-01-05 14:28 mircea_popescu: lobbes recently unveiled actionbot, which works fine, and is evidently putting all time he can into paying off technological debt he's responsible for if not necessarily guilty of. nothing wrong with this, and it can stand as such.
lobbes: possibly the bar to lordship will raise above me while I rebuild, but regardless I'ma keep rebuilding as it seems the only sane move for me. Ultimately, I just want to continue to be +ev for the republic and no way to do that without paying my technological debts
lobbes: !Xsell 140mn 48 500 wired filthy fiats (WU)
auctionbot: Sell order # 1033 created by lobbes: 500 wired filthy fiats (WU) Opening: 140mn ecu Ending: 2019-01-07 08:15:26.113976 UTC (47 hours)
mircea_popescu: asciilifeform no idea re esthlos . maths dude who did summaries, then fell behind, then caught on and wanted to put more effort into it, then fell behind again. maybe he re-emerges.
mircea_popescu: asciilifeform yeah, give the phone thing a year or two, whynot.
mircea_popescu: lobbes on the positive side, at least now you know that should that occur, it won't occur in 2019.
lobbes: this is encouraging. And if it occurs in 2020, then by that time I'll hopefully have built the solid foundation upon which to launch back into lordship in 2021.
asciilifeform: mircea_popescu: in moar pedestrian matters : daykin's algo , turns out, aint good for much -- at least, not in anyffin like orig form. consider the case where P is large ( but not at all close to complement(P) ) and Q is 1.
asciilifeform: you get a potentially 'geological' number of steps that increment by 1.
asciilifeform: ( e.g. 64 bits worth of 'increment this 4096-bit fz by 1' is roughly 400 years... )
mircea_popescu: works well for the example he gave -- numbers with same digit count.
asciilifeform: aha, for napkin numbers.
asciilifeform: i suspect that it's possible to cure it by adding 2^k-multiples of the small Q instead of Q itself, tho
asciilifeform: ( and then shifting out the 2^k )
asciilifeform: rather like what we did in barrett.
mircea_popescu: possibly
asciilifeform: it is interesting to note, i did an exhaustive dig re gcd algos; and found that there are half a dozen sub-quadratic ones, but none of those can be made constant-time.
asciilifeform: (i.e. working subquadratically in worst-case)
asciilifeform: so anyffin we do for gcd is gonna be quadratic, q is strictly re the constant factor.
asciilifeform: this isnt catastrophic (or surprising, apeloyee warned about it yr+ ago) , and the only place where need gcd is the pre-millerrabin primorial 'divisible by small primes?' litmus. but would still be good to cut the constant down.
asciilifeform: and even a dog-slow gcd is still faster than knuth-division by each of million smallprimes.
mircea_popescu: one simple solution would be to just keep digit-appropriate primorial.
mircea_popescu: test as many small primes as their product is as many digits as your proposed large prime and be done with it, daykin will work ok for same bitness
mircea_popescu: ~that~ is one of the exceedingly rare justifications for magic number. "what is this 2048 bit strange ?" "the product of the first as-many-primes-as-their-product-fits-in-2048-bits"
feedbot: http://bimbo.club/2019/01/philosophical-transactions-for-the-months-of-september-and-october-1715-part-ii/ << Bimbo.Club -- Philosophical Transactions. For the months of September and October, 1715 - Part II.
asciilifeform: mircea_popescu: magic # primorials are unavoidable. but i dun immediately see how to make it go with daykin, there aint a bailey-borwein-plouffe-style algo for gcd
asciilifeform: ( i suspect btw that if there were , you could nail rsa, thinkaboutit )
mircea_popescu: consider the simpler case of 16 bit rsa. you thus make two 8 bit primes. you daykin each of these with 210, which happens to be the 8 bit primorial, aka 11010010.
mircea_popescu: had you instead used 32 bit rsa, you'd have had two 16 bit primes you'd have daykin'd with 2×3×4×7×11×13 aka 0x5DD8
asciilifeform: soo hm, what's the idea, 8192 stored primorials ?
mircea_popescu: just a 8192 bit number, equal to their product.
mircea_popescu: you simply gcd each candidate prime with the same "product of primes in order up to bitness"
mircea_popescu: it's a one-shot thing, and it eliminates however many dozen small primes.
mircea_popescu: (i suppose if indeed you want to test MORE small primes than fit in one 8kb, you'll have a number of such composite numbers to test about. however many it takes. and yes, you can clever the knobs so they're not in strict order so that the composites are each exactly 8192 bits)
asciilifeform: 'product but up to bitness' dun do the same job
asciilifeform: ( storing e.g. 8192 primorials, ~would~ work, tho ugly, but ~would~ leak the bitness when fired )
asciilifeform: primorial that fits in, e.g., 2047 bits, is not equal to '1st 2047 bits of primorial that fits in 2048'
asciilifeform: ( would be nifty if it were, tho... )
asciilifeform: on top of this, if you actually carry out a diff stream of instructions for 2047 and 2048, you leak the bitness of the integer under test.
asciilifeform: which is a nogo.
asciilifeform: gcd(x, primorial) gotta use same instructions when x = 0, 1, ... as when it equals 2^ffawidth - 1 .
mircea_popescu: dude, why is every little thing such a fucking uphill struggle with you. suppose you wish to see if x is coprime with the number 2. you run gcd (x, 2). suppose then you wish to also see if x is coprime with the number 3. you run gcd(x, 3). all this is EXACTLY EQUIVALENT to running gcd (x, 6) : if this returns 2, it was not coprime with 2, and if it returns 3, it was not coprime with 3.
mircea_popescu: since your best gcd algo seems to be one that expects x and 6 be same bitness, there's nothing wrong with making a buncha prefab such products-of-primes.
mircea_popescu: then test against them.
asciilifeform: this much is correct, and why i have gcd to begin with. right nao i have a modified stein that goes in constanttime.
asciilifeform: orig q was whether it is possible to improve the constant factor of stein.
mircea_popescu: the whole discussion was re daykin, specifically that for our particular usecase, it's not the end of the world that it wants "napkin numbers" : we enjopy the luxury whereby we can construct them to measure.
asciilifeform: it is used to take gcd(x, primorial(currentwidth)) , and if it dun equal 1 or x , then x is 'cheaply known composite' .
asciilifeform: mircea_popescu: if you gotta do multiple calls to gcd to take gcd(x, primorial) you lose the win from using gcd.
asciilifeform: whole point is that it tests for divisibility by million small primes in 1 pull of trigger.
mircea_popescu: possibly. that's not clear, nor was it ever discussed before now. it MAY BE that a dozen calls of gdc-daykin(x, daykin-primorial) are in fact cheaper than 1 call to gdc-stein(x, primorial(currentwidth)).
mircea_popescu: it all depends on the quadratic and parameters etc.
asciilifeform: and yes it is possible to daykin with a hardcoded list of primorials, 1 for each possib bitness. the issue aint even that you gotta keep around e.g. 8192 primorials; ( you do, they can't be sliced ) , but that it leaks the bitness of X .
mircea_popescu: the bitness of x is already leaked : 2048.
asciilifeform: the factual bitness
asciilifeform: which can be e.g. 1, or 17, etc
asciilifeform: ( index of highest 1 bit )
mircea_popescu: all our factors are 2048 bits per spec.
asciilifeform: bitness in the FZ_Measure sense of the word.
asciilifeform: not of the primorial ( 'Y' if you will )
mircea_popescu: this wasn't a rsa genprime application ?
asciilifeform: of the unknown integer being tested.
asciilifeform: it is, and x is a random string from rng
mircea_popescu: the "unknown integer" being tested IS ALWAYS 2048 BITS.
asciilifeform: why, cuz you pegged the high bit ?
mircea_popescu: cuz im not going to have non-2048 factors in my 4086 bit rsa key, wtf.
asciilifeform: remember, gcd will be an exposed operator, 'G' . can be used for ~anything~ .
asciilifeform: simply so happens that it is also needed for rsa primegen.
asciilifeform: this is the montgomery thread replayed
asciilifeform: no canhaz montgomery, cuz modexp gotta work for ALL ints.
asciilifeform: same here.
mircea_popescu: i thought this entire discussion was a) specifiucally as to daykin (not to stein) and b) specifically as to primegen for rsa secret key baking, (not "in general math functions).
asciilifeform: tho i finally see what mircea_popescu was getting at earlier re daykin.
asciilifeform: yes , you can do this is if all you want gcd for is primegen.
asciilifeform: so happens that it is needed as a general-purpose knob tho.
mircea_popescu: otherwise, i guess daykin gcd can exist as a class, native or extended, w/e.
mircea_popescu: nobody is going to hate your ffa if it includes montgomery, with the proper warning.
mircea_popescu: (whole thing already comes with a "nozero" rule anyways)
asciilifeform: anyway i suspect that i'm overmassaging this particular piece, most of cost of primegen will be in m-r regardless of how i bake gcd.
asciilifeform: mircea_popescu: nonzero rule is elementarily cuz you cant div0.
asciilifeform: btw iirc apeloyee had a comment re 'why do you want to gcd(x, primorial), why dontcha generate a random x and multiply it by primorial + 1 '
asciilifeform: i can't recall why i barfed tho, nao i gotta dig out the notes...
asciilifeform: oh hm i recall nao. ( it was because operator 'P' wants to be a general-purpose primality test, valid for any input whatsoever that fits in the ffawidth, rather than simply 'generate prime' )
asciilifeform: ( why this is, is because for certain types of pubkeycrypto, you want to test adjacent nums for primality. rsa in particular. )
mircea_popescu: i don't get it how you expect to multiply some value by a (product of primes +1) and not get an even number.
asciilifeform: grr nao i gotta find the orig statement, which wasn't obv. broken
asciilifeform: ( mircea_popescu is elementarily correct here )
asciilifeform: aah it was x * (primorial) + 1 obv.
asciilifeform: not x * (primorial + 1 ) lol
mircea_popescu: and this is... an even number.
asciilifeform: as in euclid's proof-there-aint-a-last-prime.
mircea_popescu: !Qcalc 7 * 11 +1
lobbesbot: mircea_popescu: 78
asciilifeform: 7 or 11 is a primorial ?!
mircea_popescu: !Qcalc 3 * (5*7*11*13*17)+1
lobbesbot: mircea_popescu: 255256
asciilifeform: ok this dunwork, i'ma have to pull out the orig item
mircea_popescu: well this was entertaining, bbl.
a111: Logged on 2017-10-07 21:48 apeloyee: http://btcbase.org/log/2017-10-05#1721485 << alternatively, can *construct* numbers which don't have very small factors. pick a nonzero remainder mod 2, mod 3, ... mod largest-prime-fit-in-your-primorial and find what number of primorial is congruent to it using chinese remainder theorem
asciilifeform: not particularly relevant to the problem of general-purpose isPrime() tho, so i'ma put it back on shelf for nao.
asciilifeform: http://btcbase.org/log/2019-01-06#1884935 << i actually considered to have 'if low bit is 0 - i.e. N is even -- then montgomery, otherwise barrett' but what this does is break constanttimeism of modexp -- nao you broadcast the parity of N for whole planet, cuz entirely diff execution profiles for the 2 algos. and montgomery is at the very most a 10% revvup over barrett.
a111: Logged on 2019-01-06 00:08 mircea_popescu: nobody is going to hate your ffa if it includes montgomery, with the proper warning.
asciilifeform: ( and if you montgomery, then you gotta either test whether gcd(N, modulus) == 1 , or ~assume~ , the latter is a mine that user will step on. unlike div0ism , it is not an inexpensive test . )
mircea_popescu: i dunno why you barfed ; but i barfed because it's fucking stupid, you lose a lot of variety in your primes for no gains worth the mention.
mircea_popescu: and i meant include barret ~as an optional~, like a callable function.
mircea_popescu: since you're doing this "general purpose", there's no crime if user can call montgomery.
asciilifeform: mircea_popescu: barrett is the 1 that actually works for all integers without restricted domain.
asciilifeform: so , my spec was 'all operators do The Right Thing arithmetically, and program stops if you demand div0' , like yer cpu does.
asciilifeform: rather than 'watch out, be sure that you constructed right inputs, or you might get soup'
asciilifeform: mircea_popescu: the 'lose variety of primes' was my objection to using wholly-constructed primes. rather than in prelude to m-r .
asciilifeform: ( apeloyee originally proposed it as substitute for primality testing per se; where it elementarily constricts the domain of possible primes, and is unacceptable per my lights )
mircea_popescu: anyway, nothing wrong with that ffa design choice, if you like it ; if you don't anymore, also not the end of world.
asciilifeform: upstack, i'ma polish off stein, and see if the lily in fact needs gilding ~after~
asciilifeform: rather than to hold up orchestra for its gilding.
asciilifeform: ( as for the other thing -- much of asciilifeform's oddball 'must work for all integers!' thrust, is on acct of his interest in cryptosystems other than classic rsa, e.g. c-s and variations on theme )
asciilifeform: c-s actually has 1 interesting win over good old rsa -- it dun need a hash padtron
asciilifeform: but this is a diff thread, possibly.
asciilifeform: ( and at any rate we gotta have trad rsa working 1st, before any such side dishes can be considered )
asciilifeform: i'm carrying out mircea_popescu's orig spec, where 'i want a peh key with my rsa modulus that i carved on the mountain' or how it went.
mircea_popescu: just note that eucrypt having rsa does in no manner hurt your serpent-only-phonecrypto putative app ; just like it having serpent dun hurt a "this is my pgp implementation" usecase, and so on.
asciilifeform: right. whole affair is 'what's the most general arithmetron that is also a useful rsatron', from my pov.
mircea_popescu: libraries, definitionally, contain functionality some end users won't use.
asciilifeform: correct. ( i'ma include jacobi operator, after thing is flying in the field, even tho none of the currently popular schemes use it, for instance. )
asciilifeform: ideally , one'll be able to implement any crypto scheme without having to rip into the ada; simply by writing pcode.
asciilifeform: at ~that~ juncture it'll be possible to ~hardwarize~ the thing, into the ultimate (per horse's mouth) usg nightmare.
asciilifeform: ( i.e usefully general-purpose cryptotron. )
asciilifeform: i.e. that 8192-bit cpu mircea_popescu gedankenexperimented.
asciilifeform: it'll be quite a riot, i expect, when the problem for enemy of determining what to do with ~an unknown pubkey~ is suddenly np-hard.
asciilifeform: ( suddenly, 'which of these bigints do we need to try and factor, and which are noise' becomes an open human-powered q ! for'em. )
asciilifeform: it took asciilifeform 2y ( 3, if you count the mpi dead ends ) to build ~that~ deathray. and it's just about ready to fire..
asciilifeform: i can picture, for instance, that some folx will have a pubkey where 'well, first you gotta decrypt via these 2 rsa keys, and depending on the low 4 bits of the plaintext, the rest is via 1 of these 4 c-s' or the like.
asciilifeform: the problem of even beginning to break, becomes , from empty space, exponential.
asciilifeform: ( of course it also becomes possible to blow out yer own brains by misconstructing a pubkey, but this is a problem that existed before me and will exist after )
asciilifeform: and this is before you even consider that the plaintext itself can be pcode...
mircea_popescu: finally some actual movement in teh damned horses.
asciilifeform: imho yes. actually using 'we have a comp' for sumthing tangible.
asciilifeform: ( in this case, to make enemy have to sweat to simply determine ~what cryptosystem~ particular key is for.. )
asciilifeform: let'em 'ai' or whatever. haruspicy. necromancy.
mircea_popescu: "them" keks.
mircea_popescu: what fucking them. the ~pretense~ is entirely based on cheapness. no cheapness, no pretense.
asciilifeform: for all values of 'thems', present, future.
mircea_popescu: same exact reason preteen boys don't discuss their "sexual conquests" within hearshot of adult woman.
asciilifeform: stupidity, as mircea_popescu pointed out in the old essay, 'has infinite health'
mircea_popescu: i don't recall any villein serfs going around derping about http://btcbase.org/log/2019-01-03#1884205 / http://trilema.com/2014/agency-and-other-notes/#selection-31.0-31.271 five centuries ago.
a111: Logged on 2019-01-03 19:24 stratum: Right now, for the billions, I think it is probably better than nothing, just like easily popped household locks.
asciilifeform: 5 no , 2 yes
asciilifeform: ( even 5 maybe -- hussites ? )
asciilifeform: or hm, who were the derps on mountain fort with 'equalite liberte'...
asciilifeform: cathars ?
asciilifeform: that sorta thing is always there, somewhere, like polio.
mircea_popescu: closer to southern france resistence to growing crown than anything in the derp vein, but why not./
asciilifeform: there's 'resistance' but then there's 'кто был ничем тот станет всем'(tm)(r), former dun necessarily roll in the latter
asciilifeform: i suppose if archaeologizing re the latter, oughta start at spartacus..
Mocky: I just hit double digit grand kids
Mocky: and number 11 is in the oven
asciilifeform: congrats Mocky
asciilifeform: coincidentally, i nailed garbage panda #11 today.
asciilifeform: ( not nearly as sisyphian work as may seem, 1 good hunt removes infestation for 2-3 month )
Mocky: to bring it full circle, I saw one of my grandsons yesterday watching a cartoon about how helpful raccoons are for eating trash so less waste goes to the land fill, and how clean and friendly they are
Mocky: was hoping to see a russian d00d shoot at them from a window, but no, instead the raccoons went down to the creek after dinner to wash their paws
BingoBoingo: If only there were a way to get the locals to put the reject trash back in the dumpster when they finish mining
a111: Logged on 2019-01-06 03:35 Mocky: to bring it full circle, I saw one of my grandsons yesterday watching a cartoon about how helpful raccoons are for eating trash so less waste goes to the land fill, and how clean and friendly they are
a111: Logged on 2015-02-11 06:58 BingoBoingo: A viewer complained that an episode of the animated children’s series Peppa Pig was inappropriate for an Australian audience because it said that spiders were not to be feared. >> http://about.abc.net.au/complaints/peppa-pig-received/
asciilifeform: 'tis said that in gringolandia there's a steady stream of rabies cases from some derp who hand-feeds 'cute furry' shitpanda
asciilifeform: ( fox is rare here, so the primary reservoir of the virus is instead this )
asciilifeform: nicoleci aint wrong, btw, they'll walk right up to you.
← 2019-01-04 | 2019-01-06 → ↑
Random(trilema) | Download hourly DB snapshot | Get Source Code