214400+ entries in 0.147s
trinque: the fuck are you people
talking about, blank
a111: Logged on 2017-02-23 23:53 asciilifeform: and holy mother of fuck, jurov , why does your
thing mangle '@' into 'at'
mircea_popescu: which burial should hopefully occur
this year, it's getting ridiculous already.
phf: i kind of want a way
to link useful assets
to a patch on btcbase. right now if you have a readme.txt inside a patches folder you get
that included as a prelude for a patchset (
http://btcbase.org/patches?patchset=fg). i
think it would be handy
to include wires over ssh link somewhere on
the wires patch page..
shinohai: Baloney exchange is on
there
too .... but majority of exchanges use crapflare so likely affected.
phf: asciilifeform:
that's some neat hackery
mircea_popescu: asciilifeform
the administration route is
terrible, how
the fuck do you dose
the ingestion
through splashing.
mircea_popescu: also,
the russians would have probably ventilated
their guy into ~survival.
mircea_popescu: i dunno
that guy who got splashed vx would go around complaining
that hey, i got splashed
ten minutes ago.
mod6: <+asciilifeform> [BTC-dev] (EXPERIMENTAL) A Recipe for
the use of Wires via SSHITunnels. << cool!
thanks alf.
mircea_popescu: what i'm more interested in is
this apparent limit on "what can be
thought about" based on some sort of i'm not even sure what. is it
the case
that i can't
think about women now ?
veen: forget it, it's a low value point i've already attempted
to make
a111: Logged on 2017-02-24 02:15 asciilifeform: without seeing
the rack, you don't know which ones -- if any -- are fg; and which one is my air conditioner; etc
veen: i suspect we'll never reconcile
the "must build snow-flake" paranoia with drive
to make architecture which can be reasoned about
a111: Logged on 2017-02-23 23:52 asciilifeform: [BTC-dev] (EXPERIMENTAL) A Recipe for
the use of Wires via SSH Tunnels.
mircea_popescu: it's worth keeping a farm of vartious nodes just
to watcdh
them struggle with
the chain. pretty interersting data.
veen: is
this new as of
the SHA1 nooze yesterday?
mircea_popescu: the only
thing is
that gpg is already obsoleted becauyse of its inane fingerprinting scheme
veen: sounds like solution for gpg is rip out `char shitprng();` implementation and replace it with one
that calls a
trusted noise source
mircea_popescu: listen
to
the words of he who suffers
this weekly with eulora.
veen: by
that
token s/\/dev\/random/\/dev\/fg/g again gpg src doesn't help us either
mircea_popescu: asciilifeform stronger argument
than it seems. leaving aside
the static issue, it's not even a given gpg compiles at all.
veen: userland shitprng is a wrinkle indeed, one i wasn't aware of until
this
thread
mircea_popescu: so it may make sense as a convenience
thing
to symlink it
to your fuckgoats
tty. but don't expect you've now
thereby fixed
the system
mircea_popescu: well
the only argument pro i'm aware of is "it's
there already"
veen: what is at issue at
this point is use of /dev/random at all, regardless of how it is implemented
mircea_popescu: yes. but if you
then
turn around and feed
that into userland shitprng, you've not impoved anything.
veen: surely we all agree
that linux-csprng < FUCKGOATS
mircea_popescu: that sameness allows
the enemy some levers it needn't have, first of all in its own safety's sake.
veen: proposal is replace output of linux csprng-crazy
that underlies /dev/random with
the output of FUCKGOATS
mircea_popescu: actually,
the sentiment here is
that ~all usg code is just reused scraps of
the same stale old crap.
this sentiment is fed by you know, us having cut up
the usg
toys numerous
times and having run into
the same bits.
veen: but it sounds like
the sentitment here is using a well-known noise fountain is a vulnerable one
veen: to explain my reasoning by way of donning my engineering hat, i saw an existing contract of
the form "noise comes out here" and a deficit default implementation, and proposed a better impl
mircea_popescu: because cryptography deals with absrtacts only,
the kerckhoffs standard makes sense. but whenever items with an associated mass are involved,
things change.
mircea_popescu: cryptographical security is one
thing ; operations security is another.
mircea_popescu: would you propose
the usg "invisible fighter" program is "security
through obscurity" ? because... it is.
veen: security
through obscurity eh?
veen: so you've got /dev/fg0
thru /dev/fg9
mircea_popescu: review
the
thread re centralization recently. it's centralized at
the sane place.
veen: seems gpg
tried
to sovereignty-wash a source of entropy and here it is bearing your criticism anyway
mircea_popescu: so, you may not ~care~
to compile
them all ; but whether you are interested in war or not -- war is interested in you.
mircea_popescu: and gpg is not very far off
this ; neither are ALL usg produced programs you are running.
veen: not sure
that recompiling gpg
to read from /dev/fg really frustrates enemy all
that much
a111: Logged on 2017-02-24 01:40 veen: oh it runs output of /dev/random
through it's own ('cs')prng?
veen: my reasoning is
that if
the semantics of /dev/random is
that,
taken over infinity, it emits flat spectrum of octects, and lots of binaries in
the wild have
that assumption baked it, why not coax kernel into allowing FUCKGOATS
to fulfill
the contract of /dev/random directly?