301500+ entries in 0.183s
mircea_popescu: and
the (otherwise spurious) concept of "a chan" reduces
to "a list of specified broadcast recipients".
☟︎ mircea_popescu: now something like a111 would run on an "observer" node, which is
to say with
the rule
that it doesn't publish/pass lines not reported by
the putative author. because
those constitute...gossip.
mircea_popescu: and
there's sure as fuck no way for u4
to prove
to u*
that u2 insulted u1's mom, or didn't.
mircea_popescu: meanwhile u4 has absolutely nfi wtf went down
there, u2 may be an asshole or fucking around or w/e.
mircea_popescu: at which point u1 can verify
that HE didn't get
the A5 or
the A9. And he couldn't care less
that it's a "network problem" or intentional withholding, much like MPEx doesn't gas what proxies do or don't do on
the way.
mircea_popescu: then in
the relationship u3-u1 whatever G may be A1(G) will be
mircea_popescu: whereas in
the relationship u3-u2 G is "Hello u2" and A1(G) is "Hi!" and B1(A1(G)) is "How goes ?" and C1(B1(A1(G)))) is
mircea_popescu: and in
the relationship u1-u2 G is "Hai u2 howru" and
then A1(G) is "fine
ty.", and A5(G) is "urmom's a ho!" and A9(G) is "lol j/k",
mircea_popescu: so,
to belabour
this point : if users are u1, u2, u3 and u4 ;
mircea_popescu: similarily,
the ACTUAL order of message, and
the presence or absence of "j/k" messages in
there can a) never be established beyond reasonable doubt and b) can only be usefully guessed by people. which is
to say, well connected, intelligent nodes.
☟︎ mircea_popescu: as a side-effect,
the ability
to extract political
time out of
this arrangement reduces
to being well connected
to well connected nodes
that are connected
to well connected nodes etc.
mircea_popescu: and if in
THAT conversation it uses C(B1(A1))
to
transmit {A(G1) ; A(G3) ; B(A3(G3)) ; B(A8(G1)) ; B(A5(G3))}
then
that's
that.
mircea_popescu: in any case, however
the list may be sorted by
the peer, ITS peers get
the list as it has it at
that point.
mircea_popescu: this, i believe, should be a setting exposed
to
the user. wherever
the fuck he wants it
to be sorted.
mircea_popescu: if at
this point a B(A8(G1)) should show up, it should be sorted somewhere in
the list. but where ?
mircea_popescu: once B(A3(G3)) shows up, it goes obviously before B(A5(G3)), so
the list is now {A(G1) ; A(G3) ; B(A3(G3)) ; B(A5(G3))}
mircea_popescu: then at
the next level, if B(A5(G3)) shows up, it goes in next.
mircea_popescu: message A (G3) shows up, if no A(G1) or A(G2) is present, it goes in first. if A(G1) shows up later, it goes before A(G3). which means
the list gets reordered.
mircea_popescu: so :
the genesis message, ie,
the first communication between a pair (gossipd is a pairwise comms mechanism) includes 10 keys.
mircea_popescu: actually, let me go into detail, why
the hell not. ("because derps stalk
the logs for examples of how x doesn't understand y" "and ?" "myeah.")
☟︎ phf: asciilifeform: ircv3 supports backdating (as well as "batch" message spool). i've hacked
together backdating for erc, since znc supports it for spooling missed messages, but clients like irssi support it out of
the box
mircea_popescu: if later i get a message with key 1, i put it before
the 3rd key one.
mircea_popescu: message contains 10 keys. rule is
that you use
the lowest.
mircea_popescu: note
that
there's no hard limit here. suppose, and again,
this is just a blackboard mock-up, not a
technical implementation :
mircea_popescu: "bitch, you wanna live in
the us, well... learn
to ham radio
then i guess ?"
mircea_popescu: it'll give
the "get
the fuck out of
the boondocks" some very real bite.
mircea_popescu: anyway. messages should be ordered by protocol,
through
the fact you can't decrypt
the latter without
the former.
mircea_popescu: asciilifeform hot damn, your idea of
the irc protocol as implemented'd better not require an irc client
the size of an os!
mircea_popescu: anyway. nothing wrong with
that split. i guess eventually "Sane irc client" will be a project.
mircea_popescu: one of
these days i should get a client
that's not retarded enough
to split urls.
mircea_popescu: can't believe how much hanbot 's shirt bothered
teh free &
teh brave interneteer community.
mircea_popescu: it doesn't ever speak
to
the PRINCIPLE of
the cellar, however. it's always a discussion of
the circumstances of
the house.
mircea_popescu: yet some general considerations are possible. like, if
the house's in a swamp, safe bet it has none. stuff like
that.
a111: Logged on 2016-05-15 00:03 asciilifeform: i love it when mircea_popescu
takes 'my'
traditional side in argument - e.g., as seen here, argument ad metacocksucknsa
deedbot: shinohai rated
trinque 1 at 3650630430 <<
Thanks for
the help w/ Gentoo
BingoBoingo ought
to just seriously start WoT lawncare service
BingoBoingo: or
there's commercial off
the shelf aerosol foams
that do
the same
thing, but
that's basically
the formula. Or do left coasters just let
the ground wasps be?
BingoBoingo: Anyways once you find
their holes pour a bit of gasoline down
there and
then plug
the hole with dirt.
BingoBoingo: asciilifeform: Do you know where
thier hole is?
a111: Logged on 2016-05-16 01:43 BingoBoingo: <asciilifeform> we even have something called a 'ground wasp' here << easy
to get rid of
BingoBoingo: mircea_popescu: Errored out on loading, but basically
they closed
the DAOscam bet early with a refund.
BingoBoingo: <asciilifeform> we even have something called a 'ground wasp' here << easy
to get rid of
☟︎ mircea_popescu: these "forensic experts" identifying "the wallet addresses used by
the hackers" are such experts indeed.
jurov: shinohai: you send me gpg signed request
to add all your qntra shares
to coinbr account X , and if you want,
to do it automatically from now on
shinohai: jurov btw i finally made account, how
to add qntra ?
danielpbarron: ;;later
tell jurov may I please get some coppers
today?
mircea_popescu: not
that
the idea is without merit, or
technologically uninteresting. but
the move is politically inexpedient.
mircea_popescu: point's been made, let em stew in denial now for a little bit, we come back
to it later.
mod6: was
thinking it'd be neat
to cross-ref
the email
to some live
twitter acct and
tag 'em
mircea_popescu: not
that much of a priority. let
them read
the logs, what.
mod6: im ciked
that phuctor is crushing keys while i sleep, ride
the
train, eat a sandwich
shinohai: Excellent work
this month
tho mod6
mod6: anyone outside can get
their vpatch "sponsored" after review & needsRebase returns false
mod6: it should be
that anyone in L1 can use it really.
mod6: if anyone has some interesting in working on a CI project with ole mod6, lemme know. it'll
take some doing, but it could be pretty worthwhile.
mod6: <+shinohai> just as a backup at least cuz i love
t so much << ok no worries
then, just wanna ensure we're not duplicating work and what not.
mod6: everyone should do
these
things on
their own.
a111: Logged on 2016-05-15 14:43 asciilifeform:
they ~will~ have one. it is,
to use mircea_popescutronic
terminology, simply a question of how
thick
the spiked club
they will be raped with until submitting.
mod6: yah, but
this verifies stuff for historical purposes and helps regression
test a bit.
mod6: yeah, bunch of stuff
there. something
to work
towards.
mod6: someone drops in
their vpatch somewhere, it verifies
that it doesn't have
to be rebased,
then spins up an instance somewhere and builds
the full orchastra, and
then runs
tests against it.
shinohai: just as a backup at least cuz i love
t so much