log☇︎
103100+ entries in 0.828s
trinque: davout: hm seems I might've been wrong
trinque: I'm not; just saw it while rewriting deedbot
trinque: I was horrified to discover yesterday that gpg does not exit nonzero when a sig doesn't verify
davout: http://log.bitcoin-assets.com/?date=21-03-2016#1437476 <<< why's that retarded? kindof the first thing i thought about when reading your piece ☝︎
assbot: Logged on 21-03-2016 20:48:01; mod6: I like the idea though. We just don't have the tools yet.
mod6: oh, i see.
asciilifeform: though i considered, brifly, hijacking the high bits of the timestamp 32b field for it
mod6: I like the idea though. We just don't have the tools yet. ☟︎
asciilifeform: if i'm gonna depend on an antique, i'd rather take one that was made in qty 500,000,000 in 11 countries. than 10,000 in usa-only.
mod6: phf: well, other than saying "interesting" and "cool", i couldn't figure out how to do that either. unless we create a vtron signature style in 'g' or is it 'p'? i cant recall.. cause how will someones vtron be able to pull the bit string out to know how to categorize someones seal?
phf: asciilifeform: but what i'd like to know is why none of you lot commented. << i'm not groking how that's supposed to work with gpg signatures
asciilifeform: i once decided to determine the simplest box that could be assembled from readily-available (preferebly, sov-block) components, that could GENERATE and OPERATE a reasonable onetimepad.
asciilifeform: this, i must add, is not an official s.nsa project of any kind. but it is there on my desk with a dozen other items of 'parachute' variety
mod6: so i/o drivers need to be written by hand alf?
mod6: pete_dushenski: i didn't see anything crazy in there. runnign with -connect or -addnode?
asciilifeform: they also contained various custom i/o, that is gnarly, of heavy incidental complexity, and - importantly - cannot be had in new production
asciilifeform: mod6: i started with this.
mod6: i should look at some 80's hardware and see if I can spin up a rig.
mod6: ah,. yeah vax. i was kidding anyway.
pete_dushenski: asciilifeform: the weirder thing is that the stuck node is on the physical box i have, not either of the two vpses i have. and all are running 99997k or 99996k.
mod6: i think i need to start doing all of my computing with paper and pencil, in a room with no doors, no windows or lights.
mod6: i want like some 1971 vac shit.
asciilifeform: mod6: i am comfortable in the supposition that nsa cannot yet travel backwards in time.
mod6: asciilifeform: hm. im misremembering something. eitherway, outside of well, anywhere really, i can't think of a single place in the known universe where the nsa can't diddle your shit.
mod6: now i gotta dig something up on trilema. thought there was an article?
mod6: ook anyway, one time pad - i can create with dice no?
asciilifeform: mod6: and i probably ought to point out that 'quality random' was baked long ago, ben_vulpes actually took a sample unit home after c3...
mod6: But my good man, a serious cornerstone / building block to sanity and in the direction of "workable" computer will come when your work is complete and I can get quality random numbers.
asciilifeform: am i the only one here right now who isn't sitting in a dark room with a bottle of vodka and loaded nagant ?
mod6: I think if people can make punishment first, they will. cause no == yes && yes == anal.
mod6: <+asciilifeform> http://log.bitcoin-assets.com/?date=21-03-2016#1437337 << the way i see it, the 'punishment' aspect of vtronics is secondary. the main thing is that ACTUAL PEOPLE, i.e. folks willing to ~take responsibility for their actions~, could find and recognize each other. << i think so too. but the road to hell is paved with good intentions. ☝︎
asciilifeform: i actually considered this, in the beginning.
mod6: and you sign your vtron keys with your one regular key? i dunno.
mod6: maybe having different keys (key certifing "i read this" [eyes] / "I wrote this" [hands] / ... etc.) is an analog for the polyphase system that replaced it.
asciilifeform: http://log.bitcoin-assets.com/?date=21-03-2016#1437337 << the way i see it, the 'punishment' aspect of vtronics is secondary. the main thing is that ACTUAL PEOPLE, i.e. folks willing to ~take responsibility for their actions~, could find and recognize each other. ☝︎
asciilifeform: now i could be wrong, entirely, and this-is-as-solved-as-it-gets
asciilifeform: i keep getting the feeling that we only half-solved vtronics
asciilifeform: http://log.bitcoin-assets.com/?date=21-03-2016#1437334 << the distance from workable theory to cold iron here is quite short, and eminently bridgeable. if we had the usable theory. which is a problem i was taking a stab at. ☝︎
mod6: but until we all have shit that doesn't suck... we're kinda just waiting until it doens't suck. in the context of blame and "i don't have to /hope/ this thing works, I know mod6 said it does, ergo, it works."
mod6: i may want one!
asciilifeform: mod6: last week i finally built... a computer.
mod6: until we have computers, and i mean "have computers" in the sense that alf would be happy to say he has a "comptuer", whatever that means.
mod6: mircea_popescu: well, indeed. i mean, i can "swear and affirm" that my code won't fail you, but really, I can't be certain one way or another. too many variables.
asciilifeform: mircea_popescu: eh i didn't wake you up, did i
mod6: but "i swear this werks!!", then someone runs your shit and it ~should~ work, but doesn't because of unforseen hardware, or configuration or cosmic rays.
mod6: <+asciilifeform> well, the implicit 'i swear with my life to this' - leaves me reluctant to sign anything i did not write 100% of << and who's to say, even if you wrote everything at one layer, that something doesn't fuck you in the butt from a layer below? we don't have computers. never did.
asciilifeform: nubbins`: i invite you to comment on my post, possibly suggest a cleaner solution to the problem ?
mircea_popescu: i imagine that's a rabbit with a whiplash sore penis huh
mircea_popescu: ben_vulpes> right now, i expect to sign something half-baked that then crashes on some dirigible and gets me excommunicated << this is a very cogent point ; and i feel rather the same way.
jurov: btw, i have misread first mention of it as "vectored thrust" and imagined "wow S.NSA is finally onto ICBMs. Or space program?"
ben_vulpes: i did.
ben_vulpes: "this may contain pernicious hdd-wrecking nasties i'm unaware of, but is the closest thing to a bitcoinator i know of"?
asciilifeform: i signed it, but it had a distinct feel of falling on a necessary grenade
ben_vulpes: "this bitcoinates - i do swear"?
ben_vulpes: i was even uncomfortable signing the genesis vpatch. what does that even mean -- "this is actually some variant of satoshi's hairball"?
asciilifeform: i rejected this idea immediately.
ben_vulpes: right now, i expect to sign something half-baked that then crashes on some dirigible and gets me excommunicated
ben_vulpes: asciilifeform: i've been waiting for the "what does a signature convey" conversation to sign anything.
asciilifeform: and a pretty big chunk of my efforts in reading ~questionable crud~, which i do a great deal of, is wasted
asciilifeform: well, the implicit 'i swear with my life to this' - leaves me reluctant to sign anything i did not write 100% of
trinque: I can see the pragmatism of it
asciilifeform: i still don't see why we can't have machine-distinguishability of 'i swear this is as i found it 5 years ago' vs 'i read and understood this' vs 'i wrote this and swear with my life that it will not fail'
trinque: sure, for me it's "I cannot yet swear an oath to this" and could instead say "I have been running deedbot atop this successfully for x months"
asciilifeform: i listed a big one for me.
asciilifeform: and i'm not trying to be an arse to folks re: signing, but would like to know some of the ~reasons~ things aren't getting signed
asciilifeform: mod6: definitely there are some i did not sign, i listed one earlier
jurov: maybe i should have been reading and signing patches instead of doing the mempool abortion
jurov: i consider whole trb thing as something that demands my full attention, and there's always something more urgent to do...
mod6: <+asciilifeform> this patch has 'all signers: mircea_popescu trinque asciilifeform mod6' << i'll note that this signatures you're referencing are strictly for genesis.vpatch
asciilifeform: nobody, i should add, really oughta be signing unless he is ~sure~. but sure OF WHAT???
asciilifeform: so what should i conclude, that no one bothered to read ? to sign ? or that folks had reservations? which is it.
asciilifeform: because i DO NOT ENDORSE IT WHOLESALE
asciilifeform: nor do i have the stomach to sign v-genesis of ancient stuff, i struggled with the decision of whether to sign tinyscheme
asciilifeform: i, for instance, can't bring myself to sign anything from polarbeard.
asciilifeform: and the all-or-nothing factor, i think, is part of it.
asciilifeform: but imho i outlined a major reason.
asciilifeform: i can only speak for myself re: why
asciilifeform: trinque: you are reminding me of the thread where i refused to solve the problem v solves initially
trinque: I see the benefit of "this seal is just a messenger"
asciilifeform: i.e. of a vtron being confused by an enemy digging up ancient signed material.
asciilifeform: my argument is that there is a handful of EXTREMELY common situations re: signing that DON'T resolve to 'i approve of this'
trinque: "This is what it claims to be, and that is all I say about it." -- mp
trinque: I don't think all the reasons for "why do I hate" can be represented usefully as a machine-grokkable header
trinque: I have an easier time seeing the benefit of a single bit of tarnish + description why
asciilifeform: right now i am stuck simply ~refusing to sign~ as a means of expressing disapproval
asciilifeform: trinque: consider a scenario where i review, e.g., glibc
trinque: the argument has yet to be made that I'd rather have your anti-signature override my other wot member's signature, rather than your persuasion elsewhere having caused me to delete his signature
trinque: right, I am contradicting you :)
jurov: no, i mean free text + some indexable bits
trinque: a descriptive comment can explain why I signed just fine
trinque: I prefer the latter
jurov: i can see the signature rather like S-expression with ID of the signed artifact, approve/disapprove/etc. bits which could be parsed and would have room for free text with any other nuances the signer wanted to add
asciilifeform: i even thought about subservient keys of some kind, to express disapproval / 'i-found-it-as-it-no-endorsement', etc.
BingoBoingo: <asciilifeform> but what i'd like to know is why none of you lot commented. << compute time on my wetware has been over subscribed lately
asciilifeform: well not really, i'm not satisfied with the result.
asciilifeform: but literally the minimum i could think of.
asciilifeform: the dimensions i ended up with are not accidental
asciilifeform: (i found it necessary to distinguish signing 'i wrote this' from 'i modified' from 'i found it as is')
asciilifeform: jurov: i initially thought about doing just 1 bit, 'sig' and 'antisig'
jurov: i mean, i'd prolly need one signature with "pros" and anther one with "cons"