log☇︎
637200+ entries in 0.381s
mircea_popescu: asciilifeform you know you need to fix the links.
asciilifeform: mircea_popescu: i did wonder this when printed my copy
mircea_popescu: asciilifeform why is the pgp corp named twice in the rfc ?
decimation: re: infantilism, charity << http://hopelesslysane.blogspot.com/2014/08/gratitude.html " I watched one brave/stupid older woman approach a very large woman with six kids hanging off her cart ($420+ of free stuff), and tell her "I know gratitude is beyond you, the least you could do is be polite." The oldest of the boys, about 12ish, menaced her, got in her face and said, "Fuck you, bitch! You owe us!", while momma smirked in approval.
mircea_popescu: MolokoDesk ok, so one final time, to check this off.
mircea_popescu: what one woman takes as violence another takes as courtship, a point the state is desperate to hide from the more unfortunate of youze.
decimation: MolokoDesk: I disagree, "monopoly on violence" is a refrain of modern apologists for the state
assbot: You are right, but if Bitcoin ever goes mainstream that kind of regulation that ... | Hacker News
mircea_popescu: all the politics of infantilism, where one makes demands of god, are suddenly exposed to reality, because in either of these degenerate systems of governmance (really, sides of same coin) one actualy may entertain the delusion of it.
MolokoDesk: one of the many defiitions of "government" is "that faction that has a monopoly on the use of force in a geographic area"
decimation: for example, "I demand that the the group give me a "gun free" life" becomes "give us your gun or go to the gulag" for the neighbor
decimation: mircea_popescu: thinking about your definitions of socialism v. fascism, it occurs to me that one quickly leads to another
mircea_popescu: welcome to philosophically sound software design (tm). i hope to see a lot more of it in the future.
MolokoDesk: ok. if this is easier than what I've done so be it.
MolokoDesk: and it's provable that the contract hasn't been altered since signed and timelogged to the blockchain.
mircea_popescu: the verification of identity relies on acts by they who know who you are.
mircea_popescu: MolokoDesk the signature of someone you don't know is worthless to you.
MolokoDesk: ok. is the rationale here that if counterparties signed a document, they usually know who they are and since the contract is not enforceable by an external party, only by their cooperation, it doesn't matter who actually signed it if they all know they did.
MolokoDesk: heh. I haven't tested this with documents signed by anyone else yet. sec.
MolokoDesk: I did, which is why I resorted to using wot's key registry.
MolokoDesk: I didn't see any obvious way to do that.
mircea_popescu: MolokoDeck it can extract a keyid from the signature block.
MolokoDesk: ok. can GPG extract the public key from a cryptosignature without knowing which public key was used to make the cryptosignature?
mircea_popescu: if a matter of repudiation arises, then that is dealt with by testing the sig.
mircea_popescu: just like irl, the registrar of deeds does not verify your signature, merely looks that this shitwas signed
asciilifeform: MolokoDeck: please, absolutely must, take the time to understand how pgp actually works.
ben_vulpes: in that there is a signature and a keyid.
mircea_popescu: ben_vulpes all it determines is that the document formally looks like one signed by that guy.
MolokoDesk: sure, that's what I wanted to do initially, get the public key from the signature.
ben_vulpes: but in the case of an unknown key, pgp cannot determine if a signature is valid.
MolokoDesk: including the public keys in the document then signing it with those public keys would lock this in.
mircea_popescu: simply do this : separate the pastebin into individual signed bits, put each through gpg, take the apparent, unverified signer id, put it through grible to verify wot id, put it through gribble again to verify assbot linkage and you're done.
mircea_popescu: this is not something that may or can be centralised.
MolokoDesk: that's why I'm getting the public key out of the WOT registry at: http://pool.sks-keyservers.net:11371/pks/lookup?op=get&search=0xAF65AE980C825691
mircea_popescu: it compares the pubkey it has stored (and which you hopefully signed) to the shit in the hash of the signed document to establish it was not merely signed but actually signed by x.
ben_vulpes: in the former all it does is...
ben_vulpes: in the latter case it compares the known pubkey to the sig
mircea_popescu: so having a "signature apparently by 8A736F0E2FB7B452, could not verify" is one thing. "good signature from user MP 8A736F0E2FB7B452" is another thing.
ben_vulpes: and gpg verifies that the keyid is valid for the sig?
ben_vulpes: hash == keyid, right? where does gpg get the hash? from the signature itself?
mircea_popescu: ben_vulpes the keyid above.
mircea_popescu: nsa reads like a third of their "secure" comms on this basis.
ben_vulpes: what is "x" in this case - hash or name + email fields?
mircea_popescu: a point so fucking readily lost on derpjournos you wouldn't begin to believe.
mircea_popescu: ben_vulpes gpg saying "signed by x" means nothing if you don't have x's pubkey to check.
ben_vulpes: or that the signature is valid for hash "x" but that there may be collisions?
ben_vulpes: so signed by "x" does not imply that the signature is valid?
mircea_popescu: no ppl i r not insane tyvm stop pming me. i am aware that through the process as described signatures never get in fact verified and one could create a colision and sign for someone else. this is not a bug, it's a fucking feature. you ARE supposed to check YOURSELF the fucking sigs if you intend to rely on the signed documents. it's the only way to implement this correctly.
MolokoDeck: I think that use of "now" requires a comma in the first sentence, since it's used as a colloquialism or interjection.
RagnarDanneskjol: he likes to keep things quiet until they're ready for primetime
ben_vulpes: (mircea_popescu: it's like "i find this amusing, even though nobody else probably does, or someone's trolling me, but it's not quite lolworthy")
ben_vulpes: i have two whole beers to get through before i knock off for the long weekend i want to read more unit tests
ben_vulpes: MolokoDeck: care to share ?
mircea_popescu: MolokoDeck it's included in the signature block of the signed thing.
MolokoDeck: I just sent you a link to the unit tests.
MolokoDeck: rather than keep signing up for more block.io addresses.
MolokoDeck: it's easy to spiff a static address with a few satoshi.
MolokoDeck: at least to do what I'm already doing with it. If this is easier than that then it's as good as done.
MolokoDeck: you need the hash of the public key ... to get that it seems you need the public key. the PGP cryptosignatures on a document are kind of opaque if you don't have the public key already, apparently it uses the public key to determine whether the signatures match the unaltered document and are from the person who owns the private key matching the public key.
mircea_popescu: what is this, kansas ?
mircea_popescu: BingoBoingo "Now this wasn't particularly notable at all. Now a lot of mining companies are crawling" << you are not allowed to use now as the first word in two consecutive sentences
mircea_popescu: when you feed a string to gpg you either get a "nonsense" complaint or a "signed by X" response, with a warning that "we can't know who x is "
MolokoDeck: ok. so it's presumed the signatures are valid?
MolokoDeck: some sort of "give the bot another wallet" IRC commandline function would work for that.
mircea_popescu: RagnarDanneskjol nah i want it to always send from the same address.
RagnarDanneskjol wonders if single use wallets are better for this
MolokoDeck: searching and indexing that blog sounds like a website backend function, not part of the bot.
mircea_popescu: o, that what you meant, "verifies" via google ?
MolokoDeck: it also logs the documents to some kind of repostory or blog. Knowing the preferred format and web address of that ahead of time may be good, but that's like a quick mod to make it do whatever wherever.
mircea_popescu: it does not verify the signatures. it merely extracts w/e signature gpg sees in the document.
MolokoDeck: bot sits in here. People point it at a cryptocontract with multiple signatories. it verifies the signatures. if they're good it creates an unspendable bitcoin transaction with the address encoding the SHA256 hash of the contract. When that transaction clears and is on the blockchain, it notifies the IRC channel and gives the URL of a logged copy of the document, it's associated hash, and a pointer to the blockchain transac
RagnarDanneskjol: wanna run it down one more time for clarity molo? what it does exactly
MolokoDeck: if security concerns are de minimus I'm ready to finish it as planned.
MolokoDeck: so the wallet just has to have a few centiBTC in it at a time.
mircea_popescu: ok, so say what the thing is again, let's see if i say yes this time.
MolokoDeck: yeah, it's like 5 cents a document, about the cost of a xerox copy.
mircea_popescu: if someone breaks in and steals the bitcent, hey, more power to them.
MolokoDeck: or if the bot is started with a shell command the parameters end up in the process list.
mircea_popescu: MolokoDeck it's never going to own more than a bitcent or w/e
mircea_popescu: ben_vulpes it goes back all the fucking way to 1930.
MolokoDeck: probably none, other than that the software has to store the passphrase somewhere. like... on a server.
ben_vulpes: john m kerry thanks mircea_popescu !
mircea_popescu: MolokoDeck so what's the concern ?
MolokoDeck: what is of concern is that to make a valid transaction to the bitcoin blockchain one has to manage a web wallet. The first cut is going to use block.io since they have a simple API that includes the testnet (which I'm using in such a rudimentary way it doesn't matter whether it's obsolete features exist or not. the subset of features is common to the latest bitcoind)
gribble: Decembrie 2013 pe Trilema - Un blog de Mircea Popescu.: <http://trilema.com/2013/12>; ZoneHedge Wow. Such Durden. Much info. About useless ... - Trilema: <http://trilema.com/2013/zonehedge-wow-such-durden-much-info-about-useless/>; Understanding Argentina's Coming Default, For Real This ... - Trilema: <http://trilema.com/2014/understanding-argentinas-coming-default-for-real- (1 more message)
mircea_popescu: ;;google trilema story endless korea
ben_vulpes: regarding borrowing gold and refusing to pay it back?
ben_vulpes: what's the "no congress shall be held to the agreements of a previous congress" citation?
MolokoDeck: ok. not having to sweat keyring security seems a good idea since everyone trusts gribble.
mircea_popescu: all these fucking successful derps for crying out loud.
mircea_popescu: so i figure why no give clueless noobs a chance. send an order " http://mediaparty.info/2014/ << find the afterparty". half hour later, "i can't find anything. nobody is sayinga word on sm, there's ONE picture of a guy and some wine on twitter without enough background to find where it is or anything".
MolokoDeck: the way i've done it makes it somewhat stand-alone. if the public key registry is replaced it would be general cryptocontracts.
MolokoDeck: ok. so you want to use gribble as an open agent.
MolokoDeck: I presume gribble uses the wot api.
MolokoDeck: instead I'm using the wot API and the associated public key server, adding the public key to the bot's keyring each time a contract is validated, then verifying the signatures.
mircea_popescu: this way you don't have to keep updated keyrings locally or verify signatures in any wya ☟︎☟︎
mircea_popescu: MolokoDeck so basically, bot reads each document, extracts declared sig, puts it to gribble
MolokoDeck: actually, if one invites Tao_Jones in here and voices the bot some of these functions already work.
MolokoDeck: not sure I want the test server's URL going into a channel log though.
MolokoDeck: I can show you the unit tests doing that.