log☇︎
207700+ entries in 0.139s
davout: if "cutting one's own throat" == "not being able to reverify" i'll agree
mircea_popescu: that the absence of such miracle dieters will scarcely be felt is true ; but not sure if that's relevant to you.
mircea_popescu: so... no. cutting own throat, all the way.
davout: i think pruning would compare better to cutting everyone's throat "just a little" bit than "cutting one's own throat"
mircea_popescu: it's safe to say you probably missed some PARTICULAR arguments ; but it's probably safe to say you did get the fundamental argument, which is : you do, you die.
mircea_popescu: sort-of like asking for particular arguments against cutting your own throat.
davout: so yeah, i'm satisfied with the answers in the sense that i did not miss any particular argument against pruning
mircea_popescu: asciilifeform it is, at that.
davout: trinque: basically your point, which seems absolutely valid, is that past a certain depth you simply can't do a reorg, right?
davout: mircea_popescu: not really because i'm not particularly interested in discussing pruning from a position where i'm somehow supposed to defend it
mircea_popescu: davout you happy with teh responses ?
asciilifeform: davout: this is pretty fundamental material, imho, it is 'what separates us from the monkeys' (tm)
davout: asciilifeform: the "full chain" thing is simply in the context of asking a clarification to the question
mircea_popescu: it's apparently the most deeply cherished delusion, of the young-man-and-his-biddle, that "alternatives!!1"
a111: Logged on 2017-03-10 14:50 Framedragger: i like my rc airplanes. "the will of history necessitates you to X" has a marx'ified hegelian vibe :p
mircea_popescu: the "alternatives" narrative. a deep matter, going all the way to http://btcbase.org/log/2017-03-10#1624181 ☝︎
asciilifeform: silently replace the protocol with promise.
asciilifeform: sorta the whole thrust of gavinism, neh
mircea_popescu: asciilifeform what'd the usg love more than for it to fade! to become "technologies"! to etc.
mircea_popescu: bitcoin really doesn't have that many nodes active, nor has, for a while now.
asciilifeform: which i thought was pretty clear
asciilifeform: this was the subject of mircea_popescu's keccak article
asciilifeform: and the concept of 'the' blockchain -- fades.
asciilifeform: if you have enough of these, bitcoin per se becomes a very questionable proposition
asciilifeform: davout: understand, anyone who runs a node that cannot or will not produce 'historical' record ~from genesis up to currentheight~ is an attacker. doing his own small bit to nail bitcoin .
mircea_popescu: currently, blockchain is a ~whole~ story which has to check out. this can be verified. a partial story can never check out, and consequently can never be verified.
asciilifeform: where was it in the meantime ?
asciilifeform: davout: from where does this 'full chain' come from ? martians deliver it to your house ?
davout: trinque: in that particular case what criterion would you refuse them upon if you had the full chain anyway?
trinque: so you either ignore and maybe you're forked off to nowhere land, or you accept, and blindly
trinque: you have nothing to base it upon
trinque: can't even say they ~ever~ were a part of bitcoin
trinque: say someone comes at you with blocks further back than your prune, and you are compelled to accept them by hash rate
trinque: davout: it goes the other way too
davout: which is basically what everything that has been said regarding pruning boiled down to
trinque: whole thing doesn't even have to be SSD; I routinely shuffle off chunks of bdb to platters
mircea_popescu: asciilifeform i see no need for the subertfuge. if tomorrow thing is pruned to include "last 1k blocks" i'll just mine a 1k long chain in which everyone donates their coin to me, and bomb all the miners who refuse to mine on it. pie.
davout: i'm merely looking for things i'd have missed apart from "being unable to serve peers historical data is a dealbreaker"
trinque: thing wants something like 50Gb per year; couple of terabytes and you're good for a long time.
asciilifeform: it's precisely how all successful usgizations went -- slow, frogboiling replacement of a protocolic guarantee with a promisetronic one that is Just As Good (until it isn't, at which point you're long fucked)
trinque: any "bitcoin on muh smart watchz" notions... things not capable of bitcoinating can connect to something that is, and ask
asciilifeform: proponents of 'pruning' want to replace 'the weight of 400k blocks' with a promisetronic 'checkpoint list'. which, i imagine the game plan is, to eventually include deviations from the usual proof of work, per gavin's unabashed declaration where 'WE say what the blocks were'
mircea_popescu: which is EXACTLY how it works : blockchain replacement now costs the whole weight of 400k blocks.
mircea_popescu: to which the only sane lucentio answer is "fu, i own them, and if you don't like it ask me."
mircea_popescu: and this situation you wish to replace with "because we say so, and if you don';t like it ask another one of us".
mircea_popescu: no but consider. we currently have the situation where lucentio of pisa comes to padua, and inquires, "who owns these houses ?" and gets answer "x, y, z, k" and then inquires "oh yeah ? how do i know that ?" and gets answer, "because they bought from endless list tracing back to the dude that built them".
asciilifeform: there is no way to compress infinite string into finite space.
asciilifeform: and 'i want to not have printolade exist' and 'i want not to have to buy new hdd every year' is such a pair.
asciilifeform: davout: i suppose that works
asciilifeform: means wanting two individually-feasible but contradictory items
davout: "eat your cake, have it too"?
asciilifeform: trinque, no, 'to eat the fish AND to sit down on the cock'
trinque: surely that wasn't 'every clever kid wants a fish to eat and a cock to sit upon' ?
davout: google translate doesn't exactly help: "et le poisson à manger, et assis sur la bite"
mircea_popescu: what data is there OTHER than "the historical data" ?
asciilifeform: mircea_popescu: the thought process is that every clever kid wants to 'и рыбку съесть, и на хуй сесть' (tm)
davout: asciilifeform: yeah, sure you're not serving peers the historical data, that's the single thing that seems very wrong with pruning, at least to me
mircea_popescu: i'm not even sure what the thought process is here, beyond "i don't want to store books because momstate makes libraries and oh, where did anything but cosmo go ?!"
davout: trinque: "if the whole net is trimming" <<< that seems like a valid objection to me
asciilifeform: that mooches off the network, while there still are actual nodes.
asciilifeform: davout: if you aren't storing the historical data, you are not a peer. you are a pseudonode.
davout: doesn't really matter where you get it from, as long as it verifies, you're not asking your connected peers for a wot identity, yet you take their data and verify it independently
davout: mircea_popescu: nobody's asking you to trust someone on his word, to each his own chain copy
trinque: if the whole net is trimming how am I with a new node going to verify once
davout: i contend that verifying "once" is not that different from "verifying at will"
mircea_popescu: davout because who the fuck are you and why do i have to take your word and if your house burns down what do you do.
asciilifeform: davout: what does 'verify' mean if you can't trace the coins from birth to current day ?
mircea_popescu: it'd necessarily be a scheme with expiring coins, seems to me.
davout: mircea_popescu: how is it nonsense if i have personnally verified the chain?
asciilifeform: 'where did this money come from, how much exists in total' 'dunno, in the Great Pruning of 2050 the scrolls of the ancients were lost'
asciilifeform: mircea_popescu: i'd say that the idea is impossible in the general case.
asciilifeform: there is nor cannot be such a thing where a fortranesque '1 := 2' produces a consistent state.
mircea_popescu: as to the direct question, a pruned chain may make sense, if someone came up with a way to do it sanely. this seems impossible on bitcoin-as-it-is, and perhaps unlikely in the general.
asciilifeform: mircea_popescu: even in your demurraging scheme , you would not thereby be able to safely recycle tx id.
asciilifeform: the solution whereby tx id is unique for all eternity, is not stupid. it is the Right Thing.
mircea_popescu: the defense to something stupid isn't picking something else that's stupid to be yelled at top of lungs.
mircea_popescu: i'm not proposing the solution is acceptable. i am showing you that your OVERSTRONG STATEMENT is nonsense.
asciilifeform: mircea_popescu: demurraging coin fails the berlin wall test.
a111: Logged on 2017-02-27 16:56 mircea_popescu: but the correct trb-i might just as well end up this situation where block reward is 1mn bitcoin, and it dies within 1mn blocks. so all mining does is produce ~ a lease ~ on a chunk of bitcoin. and the value of old bitcoin is monotonically decreasing over their lifetime.
davout: your copy of the blockchain being personal i don't see any other problem than peer-service, if, and only if, you've validated it as validly producing the UTXO set you consider valid
asciilifeform: if you cannot show a coinbase origin for EVERY coin that ever existed, you have printolade.
mircea_popescu: davout in principle nothing ; in practice if you can't justify whence the coins came then what do you have
davout: i see not being able to serve historical data to peers as the major one from where i stand
mircea_popescu: mpoe-pr didn't have "dev" talk on her list and who else read that shithole lol.
davout: as a naive question, what exactly is the problem with pruning?
asciilifeform: 'The list of all used transactions isn't readily available, and once pruning shows up, it might not even exist at all. So, it only makes sense to compare the new coinbase to the list of transaction hashes that are unspent at the time ... ' << holy fuck the 1) idiocy 2) nobody challenged it, afaik
mod6: <+davout> it did not occur to these monkeys that somehow it might also make sense to ~not~ include a subject line reading "O HAI YOU HAZ X DEPOSIT KTHXBYE" << lol, huuurrrr
ben_vulpes: don't get me wrong, responding in context is entirely adequate most of the time
ben_vulpes: danielpbarron: yeah, but falls apart if i want to hold onto a signed thing.
mircea_popescu: alf would like this story.
davout: it did not occur to these monkeys that somehow it might also make sense to ~not~ include a subject line reading "O HAI YOU HAZ X DEPOSIT KTHXBYE"
ben_vulpes on tenterhooks for punchline
davout: in other GPG-related lulz, when I decided to give the kraken idiots a chance, i ticked the "Encrypt mail sent to me with GPG" and gave my key
a111: Logged on 2017-03-16 18:41 ben_vulpes: enTIREly unrelated, does anyone know how to get gpg to decrypt a message that is also signed, but to produce the signature in addition to saying that the signature is good?
Framedragger: mircea_popescu: ah ok, i guess the issue is that the sig is part of binary blob, need to convert it, etc, hrmh
danielpbarron: i really like the implied sig of the encrypt part of received message back to sender. that way neither party can hold signed material against the other in public
Framedragger: ben_vulpes: sorry for being obtuse, but if by 'show signature' you mean print signature in ascii-armored way, why can't you `echo 'foo' | gpg --clearsign > a.txt`, then `cat a.txt | gpg --encrypt --recipient recipient-username > b.bin`, then `gpg --decrypt b.bin`? (this assumes gpg is interactive and will ask for password, so best to break it into multiple commands)
danielpbarron: secret message in keys on the ground
trinque: he proposes stuffing the sausage in a sock
davout: ben_vulpes: only a terrorist would want to see the signature