200300+ entries in 0.122s
mircea_popescu: BenBE well, while
the state of hardware generally is very poor, due
to
tmsr dilligence
this particular corner is actually solved, and so no, wrt rngs
the state is fine.
mircea_popescu: FG is, importantly, an ~auditable~ rng. you can audit
the
thing, and if you do, i'd be very interested
to hear
the results.
BenBE: It's a sad state related
to hardware
that we have.
BenBE: Any other FPGA should basically do: Original design was a CPLD. So you'd grab one FPGA you /can/ audit
the
toolchain for and compile
the design for
that FPGA. It's not
too much code.
mircea_popescu: last i heard he had enough of licking acids under
the microscope.
BenBE: And I did not say I'd fully
trust it --- far
too little actual audit work done.
mircea_popescu: well, we don't
trust xilinx for critical infreastructure for
the ~same reason we don'tr
trust windows.
BenBE: Trusting
trust. Yes.
mircea_popescu: so you are
trusting xilinx
to actually do what it says ? and
this with code
that you can't audit ?
BenBE: Multiple rings (3 or 5, would have
to lookup in
the VHDL file) are XOR'd
together when sampling.
Thus not one oscillator, but
the XOR of different onces.
mircea_popescu: as per
that ancient "pi digits are also random, especially if you don't know what pi is" lemma.
mircea_popescu: BenBE FG scores perfection not before seen in ent/dieharder, not
terribly sure how much
that convinces me.
BenBE: No, only did some
tests with dieharder on its output (at 2Mbps) which score several less WEAK in
the results.
mircea_popescu: obvious example : does monotonous
temperature variation result in more 1's ? something along
the lines of "batch 1 we kept at 20, batch 2 we
took from 0
to 40 over one hour, batch 3 we
took from 40
to 0 over one hour. out of
the 10gb worth of entropy recorded in
that hour, batch 1 is 50-50 split, batch 2 is 75% 0s, batch 3 is 74% 1s.
mircea_popescu: did you do a mapping of
temperature -> entropy or anything like
that ?
BenBE: In
the FPGA versions
they are built of unclocked rings of logic gates (transistors). With
the FPGA I use
they oscillate at about 150-200MHz (if I read
the information of
the synth
tool correct)
mircea_popescu: can't say i have any experience with said modulation. what are
the ring oscillators built out of ?
BenBE: With
the WhirlyGig it's
temperature-based modulation of ring oscillators.
☟︎ BenBE: What's
the entropy source used in
those Cardano RNG?
mircea_popescu: i needn't prove it. he doesn't want
to be ~guilty~ he keeps out of doing bad
things.
that simple.
BenBE: I am vary of
things he does, will also
tell others
to do
the same, but absent of proof I can't reasonably call him malicious.
BenBE: mircea_popescu: I
think
there's a difference between suspecting someone of malice and proving it. Cf. Occam's razor.
BenBE: Didn't know
that particular project, but have a
true RNG based on
the WhirlyGig design at warmcat
mircea_popescu: BenBE yes, but here's
the extra step : not only do i know koch is an evil shithead who dedicates his
time doing evil. i also say it.
mircea_popescu: asciilifeform incidentally,
the more i
think about it
the more i'm convinced
the ONLY "fingerprint" for rsa key may be...
the modulus. 4096 bits and fuck you, if you can't
take 32 chars you don't belong here.
☟︎☟︎☟︎☟︎ mircea_popescu: BenBE do you also know
the FUCKGOATS, so as not
to ever again use a "prng" for as long as you live ?
BenBE: I asked W.Koch about
the PRNG about 2 years prior
to
the break last year. And even
then it had been know for years before
that,
that
the PRNG is phishy but nobody cared
to actually step forward and rip it open.
BenBE: Depending on
the circles you worked in, you knew GnuPG is broken on several levels.
☟︎ BenBE: OT: I know
the guy who recently broke
the GnuPG PRNG. He has been working on a project with me for some
time
Framedragger: (well it's also
technically weak by using a 160 bit hash, etc) :p
mircea_popescu: but
to be perfectly clear : rfc 4880 is not
technically weak. it is politically subverside, and deliberately so. it consists of
the same material
that has schneider running around
trying
to convince people
turning
their computer in for an ipad is "the way
to bright future of socialism", or of
the uk cocksuckers
that covered up rotherham droning on as
to how
they have
to have everyone's keys
to "catch criminals".
BenBE: At least
those were
the aspects I mostly disliked about it. What were yours?
BenBE: Hard
to parse,
the RFC is somewhat strange, and possibly other
things.
BenBE: Can't claim I didn't learn anything while preparing
to setup my project.
BenBE: But getting everything into place helps alot figuring out how
to best optimise your DB, how
to perform imports, and so on.
BenBE: It is when you can't handle
the vast amount of data it involves (and yes, I know
that vast is an understatement here). Alone building a database
to manage all
the raw data for my KeyInfoDB/Kompromat project is ~500GB (compressed) keys. Automating stuff for grabbing
these at
their source for import is a chalenge of its own.
mircea_popescu: for a moment
there it read as "i don't like $newguy, he might do something stupid." which is silly, let him do
the stupid first, not like him after. wtf preemptive dislike.
mircea_popescu: tbh i'm not sure what fellow's
trying
to do, but anyway.
mircea_popescu: BenBE
trouble is we're moving away from
the inept koch standard.
BenBE: Suggstiong: you have all
the keys available
that make up
those dumps. What about using a batch job (once per day) creating a large .pgp file people can download.
That's both static and gives people all
the information (create UIDs for
the extra stuff if necessary)
mircea_popescu: BenBE
the problem is
that others may attempt
to use it and involve you
thereby. which is also
the whole of
their business plan, i'm sure you realise. anyway,
the whole
thing's moot : proper wot exists, we're using it,
the soup can go hang.
BenBE: asciilifeform: 1) No, can use it without (just need
to do stuff on
the command line
then) 2) not a feature you need
to use.
There's a CLI clint you can audit and use.
mircea_popescu: hopefully we don't end up with
too many people importing
this and
then wanting
to strangle us when
the obviously adhoc
tree ends up rebased.
Framedragger: and presumably sadmods as well in
the future (again, not yet cached)
Framedragger: mircea_popescu:
the only external-dump link
there is sadmods (not yet cached - again, unsure of impact on phuctor performance)
BenBE: asciilifeform: it's not
totally broken (as with GnuPGs normal WoT) in
that it makes
transparent
to not expect
too much from it, appart from account associations.
Framedragger: asciilifeform: mircea_popescu: ok, sorry for
that
then
Framedragger: mircea_popescu: if you check,
the hrefs on
that index page link
to phuctored.html on siphnos etc
Framedragger: asciilifeform: oh, so only
the deedbot announcer here on irc excludes some new keys?
mircea_popescu: i wasn'\t aware it does,
that's what deedbot uses
to pipe here also
Framedragger: mircea_popescu: i mean
that not *all* new entries are included, rsa
truncates,
them,
too. asciilifeform can confirm
tho.
mircea_popescu: Framedragger dja mean rss ?
that's
the fucking point,
to
truncate, so you don't keep getting ancient data
Framedragger: eh it's like facebook, i have a shitty facebook acct,
too
Framedragger: i don't
think rsa gives all, just 'first 10' or sth
Framedragger: rsa
truncates, etc etc, other methods (pg notify / etc as mentioned by
trinque) require asciilifeform's intervention
mircea_popescu: need a better method
than downloading
the whole list of shit you already know every
time.
Framedragger: mircea_popescu: everything except sadmods so as not
to upset phuctor
too much. need
to discuss frequency of caching with asciilifeform i
think
mircea_popescu: Framedragger do you currently cache
the whole phuctored dir ?
Framedragger: oh, nice
to have all of
them at one place, convenient
BenBE: Framedragger:
That's one of
the resources I will be sourcing my Key Information service from. Contains
the private keys for e.g.
the OpenSSL Debian Weak Keys (complete)
jhvh1: shinohai:
The operation succeeded.
shinohai: !~later
tell pete_dushenski Entry in bots directory for "Time since last block" has
transposed letter, should be `tslb` -
Thanks!
BenBE: mircea_popescu:
thx.
mircea_popescu: BenBE now you can self-voice (!!v ) ; do
that it'll save us
the
trouble
to up you.
a111: Logged on 2017-04-09 13:48 BenBE: I'm
the maintainer of
the GeSHi syntax highlighter for PHP, which is used in e.g. Wikipedia for source highlighting of articles. Also working on several crypto-related projects like my own
TLS/SSL
test, a collection of publicly-known set of compromised keys, an OpenSource management software for handling X.509 certificate issuance for a certificate authority.
mircea_popescu: shinohai dun knock it, might've been
the best answer
to date. except perhaps
that reuters chick.