33 entries in 0.737s
mircea_popescu: there's exactly nothing similar between rsa packet and serpent packet. for the same money could ask to have busses and flour delivered in single container.
diana_coman: alternatively the hello message stays single-packet and uses a keccak hash of the public key (n,e,comment) as "account ID" so 3.1.5; then key is sent via Data packages and basically I need to define another type for RSA public key; server can ask/expect the RSA key *every time* to preserve same answer behaviour or otherwise only if it doesn't know the key
a111: Logged on 2015-01-07 01:22 asciilifeform: with udp, you can make the 'friend or foe?' decision upon receipt of a single (!) packet.
a111: 28 results for "single packet", http://btcbase.org/log-search?q=single%20packet
asciilifeform: !#s single packet
ben_vulpes: "the greatest website to ever fit in a single TCP packet"
sina: "To craft a valid packet, a sender must collect a single auth string from the receiving node's lighthouse (via whatever means, can be a shortwave tuner), craft auth with it as described by Mircea Popescu earlier, encipher to receiver's RSA pubkey, and send." ?
asciilifeform: http://btcbase.org/log/2016-12-31#1594370 << no, it is a peculiar bug, that i will fix as soon as current packet terminates. though strangely enough every single instance of it to date resulted in a validly phuctured (factors smaller than self) modulus eventually ☝︎
asciilifeform: one is to banish mosquitoes from the atmosphere. for which -- single-packet 'man or mosquito' litmus is required.
asciilifeform: a working gossipd must combine two seemingly-contradictory features: 1) friend-or-foe identification on single packet - no such thing as ddos or replay remains possible -- 2) enemy on the wire can learn nothing about relation graph.
asciilifeform: one other point re http://btcbase.org/log/2016-08-26#1529877 is that divorcing from 'a node is an ip' entirely, makes this or any variant of this attack considerably more painful. with single-packet quanta, and raptor code (i.e. it does not matter whether mircea_popescu gets all of my packets in order, but only that he gets 10% of the packets i send in next second, in SOME order) his 'node' could consist of 15 machines in 6 continent ☝︎
asciilifeform: phf: the handful of interesting aspects (single-packet friend-or-foe, no tcp) were outlined here.
asciilifeform: the cost of validation in a single-packet-authenticating protocol where you crunch the numbers at line speed is effectively 0.
asciilifeform: who can deny you one single packet
ascii_butugychag: http://log.bitcoin-assets.com/?date=04-02-2016#1396374 << if i can't encrypt a 512byte udp (or file off the protocol number as per gossipd thread) single-packet-hello turd, it's useless ☝︎
asciilifeform: but yes, it fundamentally it violates the principle i stated here as 'nobody gets anything just for showing up, or the flies eat the meat' (see the bastard blocks thread, and the single-packet authentication as elixir against ddos threads)
ascii_field: single-packet messages, with no ack.
asciilifeform: go, 'feel safe' because you went along with usg's transparent attempt to ban the only presently working single-packet transmission medium.
decimation: so you are against single packet routing in general?
asciilifeform: over what single-packet proto ?
asciilifeform: what i'd really like to see is what mircea_popescu proposes i use for single-packet auth that 1) isn't udp 2) existing net will reliably route 3) existing os will agree to emit without any new code running as root
mircea_popescu: you can have single packet auth outside of udp.
asciilifeform: mircea_popescu: single-packet auth is the only gossipd worth using.
asciilifeform: but the aspect i'd like to emphasize - one which mircea_popescu thought to be an optional frill and did not include in his spec - is the single-packet authentication.
asciilifeform: decimation: you can't decrypt a single packet until a cell peers you
asciilifeform: you will -never- get single-packet friend-or-foe with tcp or tcp-like apparatus.
assbot: Logged on 07-01-2015 01:22:05; asciilifeform: with udp, you can make the 'friend or foe?' decision upon receipt of a single (!) packet.
assbot: Logged on 07-01-2015 01:22:05; asciilifeform: with udp, you can make the 'friend or foe?' decision upon receipt of a single (!) packet.
asciilifeform: with udp, you can make the 'friend or foe?' decision upon receipt of a single (!) packet. ☟︎☟︎☟︎
asciilifeform: decimation: you're entirely correct that this will eventually be a thing, though. hence the need for stateless, single-packet auth hellos.
BingoBoingo: <asciilifeform> some jobs simply do not parallelize. << Single network interface with packet filter
RagnarsBitch: It has to do with packet loss and pools not being designed for just a single miner. I know I've had these same issues anytime just one miner is active on a pool , then resolves as soon as a couple more come online
decimation: I like the idea of sending a single udp packet as a request and receiving a stream of udp packets in reply