147000+ entries in 0.091s
shinohai: blockchain.info on
the phork: "To ensure
they remain safe while
the network is in
transition, we will
temporarily suspend bitcoin send, request, buy, sell, and exchange functionality roughly 12
to 24 hours before
the
time of
the fork until
the network stabilizes. "
jhvh1: BingoBoingo:
The operation succeeded.
jhvh1: danielpbarron:
The operation succeeded.
danielpbarron: !~later
tell BingoBoingo i don't
think your wotpaste url was correct. can you resend?
mircea_popescu: just about
the item i
thought was under discussion all along.
shinohai: What's
that a picture of, a fivver developer's meetup?
a111: Logged on 2017-10-26 16:59 mircea_popescu: anyway, back
to
the bait and switch
thing. reddit started life as a sort of 4chan with a jump, ie
the little girl's not on
the page, but you have
to click.
this wasn't accidental, because
their phase 2 ("growth phase" eh) consisted of
the exact value proposition : add your crap
to reddit, watch people click on it. once growth stopped
they dropped
the bait, switched
to "we will control
the conversation in
this ever-so-importan
mircea_popescu: in entirely unrelated lulz :
the "womenz in
tech" dynastry of aparatchicks over at wikipedia is pretty lulzi : sue gardner's lengthy
tenure ended (she was recycled
to nsa honeypot "tor project")
to be replaced by lila
tretikov,
two bit russki scammer gal who got sent off (over
the whole "knowledge engine" failed wikimedia
http://btcbase.org/log/2017-10-26#1729324 attempt)
to be replaced by katherine maher.
☝︎☟︎ mircea_popescu: gaussian is a decent "middle of
the way" representation ; but not so useful when discussing specifically stuff
to
the extreme one side.
mircea_popescu: yes it's
true
that heat in a rod, or whatever diffusion phenomena are relatively smooth in
this sense ; nevertheless em is
the least smooth of
the entire series.
mircea_popescu: anyway, i suppose
the correct reference i sohulkd have given re smoothness was weierstrass.
mircea_popescu: it's not indefensible
to suspect juche sparkplugs still fucking sparkplugs. perhaps.
mircea_popescu: ie,
the very substance of
the "fox can be found" argument. yes, if he uses sparkplugs you yourself sold him ; yes, if he lives in a space of gaussian noise ; yes if etc.
mircea_popescu: i suppose.
though really
the problem as far as interests here is
this : presented with an "aberration", in natural noise you really can't
tell whether relevant ; whereas in gaussian noise you generally can guess.
mircea_popescu: perhaps not
the best
term for it, as it's metaphorical rather
than descriptive.
mircea_popescu: or for
that matter, dja remeber
that old "encrypted" penguin logo ?
mircea_popescu: apeloyee let me
take
the matter
this way : are you familiar with how every "be a graphics artist" software package out
there has a coinvolve-with-gaussian "effect" ?
mircea_popescu: not ~entirely~ passive, he makes some assumptions on
the nature of "using hunter-provided equipment"
mircea_popescu: asciilifeform certainly ; but
this would
then match my criteria for "not by purely passive em methods" above.
mircea_popescu: however you define your function,
the principal distinction between gaussian and natural noise is
that
the former is smooth.
mircea_popescu: the problem with "natural noise" is
that
there's so many local maxima.
apeloyee: real noise is not so gaussian <<
true. but still means must solve
the q of optimal bandwidth each
time
mircea_popescu: (imo
the best model of "real noise" in
this sense is perusing
the SETI reports, or else any major, indiscriminate dataset of radiotelescopy.
THAT is how natural noise goes. with
the occasional neutron star and
the occasional "holy shit" scribbled in pen)
apeloyee: i meant optimal for
the sender and his expected recipient, not eavesdropper.
mircea_popescu: apeloyee but
this is relevant if you already know
there's a signal
there
apeloyee: if you use wide spectrums, you get back
to
the problem of, "nature has more energy
to burn
than you do" << I'm unsure. Shannon-Hartley
thorem sez
that at least in its conditions, using as wide spectrum as poosible is optimal.
mircea_popescu: if i actually
transmit fg output, do you expect you'll ever find ?
mircea_popescu: apeloyee suppose what is
transmitted is an equivalent of
the "number stations" of old. minus, of course,
the girly voice.
apeloyee: if you have more
transmitters
than enemy has receivers "in
the area" (precise meaning is a legit research q),
then you can hide from
the enemy, otherwise, you will be eventually found.
mircea_popescu: if you use wide spectrums, you get back
to
the problem of, "nature has more energy
to burn
than you do"
mircea_popescu: apeloyee narrowness of band was
the original attempt
to get out of
this.
the narrower
the band
the "better"
the reception in
the sense of fewer natural perturbations. which is what drove
the move
to diodes and
then ic.
apeloyee: if
there's
truly
that many perturbations,
they will also make difficult for
the intended recipient
to receive correctly.
mircea_popescu: asciilifeform
then i guess foxes use hunter-provided equipment.
mircea_popescu: apeloyee ammo is always expensive even if free, for a purely ecological reason. you don't want your space
too filled with ordnance.
apeloyee: it wastes all
the ammo.<< if ammo would be cheap
they wouldn't.
mircea_popescu: in general,
the only
thing a "find me an em shrubbery" produces reliably are false positives.
mircea_popescu: apeloyee
take an actually publicly-documented example : most crews
tend
to
turn off
the auto-return-fire
thingee because left
to its own devices it wastes all
the ammo.
mircea_popescu: all sorts of phenomena intercede. clouds is an example, but
the point is
that "noise floor" is not JUST a floor, but also randomly distributed perturbations, which during a whole day can reach significant amplitudes.
apeloyee: I concede I don't have a good estimate as
to
the amount of information a fixed location's good for
apeloyee: clouds move. otoh if
the suspected source keeps
transmitting...
mircea_popescu: (item got reliably measured because in
the 90s
they were derping about putting a d-fluorine mw laser on it.)
mircea_popescu: for instance :
the majority of low clouds passing over
this here observation point are em-louder
than CG-73.
apeloyee: can just as easily be, say, 17 sources geographically dispersed , working in concert << you would not particuraly need fancy modulations
then
apeloyee: by using stronger signal
than
the stoves. if you
try
to be below noise, you would
transmit slowly, and
thus still not evade max # of bits.
mircea_popescu: apeloyee maybe i'm missing something. how do you distinguish
this "something" from "people
turned on
their electric stoves" say ?
apeloyee: two stations won't let you recover
the signal, of course; just know something's
there as soon as it
transmitted enough bits (literally!)
apeloyee: asciilifeform: in one receiver case, one can detect a narrowband
transmission using autocorrelation; in case of "aperiodic" whatever, one can detect a spatially fixed
transmitter by measuring correlation between signals of
two stations. of course, in practice one would need
to use many pairs (or a clever algorithm I cannot now
think of), and filter known stations
shinohai is now
thoroughly depressed after reading October s.qntr statement
a111: Logged on 2017-11-04 14:18 asciilifeform: apeloyee: q was, is it possible
to
transmit in such a way
that for someone without
the key,
there ~is~ no 'signals received'
mircea_popescu: romanian expression is "taie frunza la ciini", ie, divide leaves among
the dogs.
mircea_popescu: same ol' disaster of commons. when it comes
to it, erryone's got something
to do at home.
a111: Logged on 2017-11-04 14:45 asciilifeform: ( incidentally
the other day asciilifeform finally met phf in meatland )
mircea_popescu: but
that practical consideration aside, obviously nothing wrong with doing
things right.