832 entries in 0.575s
phf: jurov: i've been playing with the
gossipd PoC, and it's entirely possible to get a gnupg packet from the wire, defensively, byte at a time, and know at any point how much more you need to get.
mircea_popescu: but be all this as it may, for as long as
gossipd.pop | print yields ((1,2),(2,3)) for me while
gossipd.pop | mathgraph yields whatever ascii wants it to yield, we're happy.
mircea_popescu: hey check that out, zimmerman makes nsa-phone and david chaum makes nsa-
gossipd.
☟︎ ben_vulpes: yes well clearly when we actually have
gossipd auth i can wash my hands of auth too.
trinque looks forward to the day he can fiddle around with the concept of hypertext on a
gossipd trinque: ascii_field: we discussed this as well; even assuming a trusted
gossipd net, there'd still be connections out to the unwashed nether world
mircea_popescu: and i expect
gossipd to be undistinguishable for the phillipino/redditarded eye from current b-a log.
BingoBoingo: ascii_field: mebbe deedbot is afraid of the
gossipd similarity to usenet? Fears #b-a.tasteless?
mircea_popescu: looky. a
gossipd based on the concept that you keep firewalled clients at home and they connect to a random pile of websites, which simply respond with pgp -w text
mircea_popescu:
gossipd must work on a pull model. exactly like old ml.
mircea_popescu: there is absolutely no reason for
gossipd to work on a push model of data transfer.
mircea_popescu: also, thinking about it, it occurs to me that the model we've been using to think about
gossipd is fundamentally broken.
assbot: Logged on 28-11-2015 20:42:43; phf: ben_vulpes: sure, so you're going to cap the size of ciphertext of what you expect to be nonce, which is fine, but after that you rely on state outside of
gossipd (tcp packet from ip such and such) to drop all subsequent
ben_vulpes: jurov: phf's point though is that then you're relying on state /outside/ of
gossipd to run
gossipd. if i understand correctly.
phf: ben_vulpes: sure, so you're going to cap the size of ciphertext of what you expect to be nonce, which is fine, but after that you rely on state outside of
gossipd (tcp packet from ip such and such) to drop all subsequent
☟︎ mircea_popescu: ascii_field your
gossipd node is stuck doing a version of this.
mircea_popescu: but so far we don't even know if we actually want rsa (this for lack of
gossipd) nor have we studied shoup etc.
ascii_field: mircea_popescu: the 'encrypted connection' item necessarily reduces to
gossipd trinque: adlai: assuming a
gossipd wotnet, and there being some way to retrieve things provided by your wot in a cryptographically sound manner, the logs would always be enough
mircea_popescu: so
gossipd is practically going to be lisp hand-compiled ot c ?
phf: i have some code for
gossipd, i'm going to by the spec that's published on trilema, rather then follow up conversation about udp and first packet validation. i wrote enough in c to be able to prime gpg machinery and send packets over the wire (which basically coveres things that i wasn't sure how to do before i started), then i switched to common lisp to get a prototype up. what came out of that so far is updated bindings for lisp
trinque: yep, I think asciilifeform has it, that it'd be something to build around
gossipd v_diddy: *another* killer application of
gossipd.
phf: g is presumably ascii's
gossipd version
mircea_popescu: phf notice that the
gossipd thing was specced, the pgp one has not.
phf: so
gossipd is "in progress", but fella's not been heard from in a very long time. gpg has not been started, but ascii presumably has a lot of ideas about it. then there's pie in the sky projects, like fabricating own cpus or rsa over udp routers.
mircea_popescu: there's also a bunch of large-ish stand-alone items, such as the
gossipd (which might still be done by artifexd, haven;t heard from him for a while) and a proper redesign of pgp
trinque: this is where
gossipd deniable statement mode would be great
ascii_field: BingoBoingo: correctly-implemented
gossipd is intrinsically ddosproof.
ascii_field: BingoBoingo:
gossipd is a computationally very lightweight thing, and also does not suffer from network effect
BingoBoingo: dependence of
gossipd (whenever it emerges) may face geater danger
ascii_field: i had the thought once that
gossipd would make this item quite trivial to implement
trinque: I'd build a dedicated box solely for using
gossipd, browsing the WoTnet, running bitcoind, and very little else
ascii_field: some time during the original
gossipd thread
mircea_popescu: asciilifeform there can trivially be
gossipd-over-blockchain.
mircea_popescu: what, jam tomorrow, maybe one day ima get
gossipd over udp, perhaps, who knows ? meanwhile half the remaining usg online is BASED, and requires udp to survive at all ?
mircea_popescu: ima see
gossipd alright. atm, the infrastructure is not the problem.
decimation: udp can't be easily distinguished, and this is a
gossipd-flavor
trinque: shinohai: I look forward to the day publishing is done over
gossipd.
mircea_popescu: and jurov's working on an irc integration thingee, which i intend to upgrade to
gossipd once that's done, and generally...
mircea_popescu: but yes,
gossipd is the huge sort of project that looks like it'd benefit from a few failure reports from failed attempts before we seriously have a shot.
trinque: I really don't think not having a b-a satellite network should preclude getting started on
gossipd phf: build FIDO on top of
gossipd ;)
mircea_popescu: i guess we're all just holding our breaths for the
gossipd trinque: second, any node accessing qntra has the option of caching the data and making it available for their own
gossipd peers
trinque: felipelalli: were there a
gossipd, it'd subsume this whole
http mess entirely
ascii_field: incidentally, if we ever abandon this place for
gossipd etc. someone oughta leave shannonized (from logs) versions of us
ascii_field: mircea_popescu: i meant - the lot of us. nobody wants to really break a sweat for channel housekeeping because
gossipd is 'the right thing'
assbot: Logged on 10-07-2015 17:07:16; ascii_field: my current hypothesis is that all the relevant folks are waiting for '
gossipd' and don't give much of a fuck
ascii_field: my current hypothesis is that all the relevant folks are waiting for '
gossipd' and don't give much of a fuck
☟︎ trinque: trilema and loper I expect to be able to get via hamsterwheel shortwave
gossipd someday
ascii_field: decimation: you can do this, and sign the result with your
gossipd key, etc. sure
trinque: anybody ever looked at tinc as a starting place for a
gossipd?
ascii_field: i am looking forward to hearing of some answer other than soldering iron or
gossipd ascii_field: the only solution i can presently think of is a variant of ntp over
gossipd.