234000+ entries in 0.157s

mod6: Food for
thought I guess. :/
mod6: Eh, I don't even
think I like my own idea.
mod6: Or
tb0t for
that matter.
mod6: Something just
to start
the
thought of what
to do about all of
this. Maybe
there is even some folding into a UCI with
this, not sure.
mod6: I did
that, as
to not spam
the channel here, and
to
try
to
think
through what my entire
thought might have been. If
there is any merit
to my
thoughts, I'll post it somewhere.
mod6: To start a bit deeper of a discussion surrounding what we
talked about with 'infrastructure' for V, if it's necessary or not, or whatever. I kinda put some
thoughts down;
http://dpaste.com/2S08M67.txt mircea_popescu: incidentally, is
there such a
thing as an "ai code commenter" ? specifically, rather
than
the hard problem of "human speech" or junk of
that nature, is
there any machine approach
to
turning compileable code into literate code (kuhn) ?
a111: Logged on 2016-04-10 13:15 adlai: ;;later
tell mircea_popescu submission for
trilema movie reviews: Inside Man
mircea_popescu wanders off
to add danielpbarron eu patch and see what happens.
jhvh1: shinohai:
The operation succeeded.
mircea_popescu: (in any case
the reason
that
the
two haven't merged sooner was
that in my head business (ie, dealings in between people) and code (ie, dealings in between machines) were fundamentally separated. but i'm very interested in whether
this notion actually stands
to scrutiny.)
a111: Logged on 2016-12-17 18:26 mircea_popescu: but
the point of deedbot is
to contain signed matter intended for human hands ; it is why
the signed matter intended for machine processing is a separate mechanism.
mircea_popescu: no see, nevermind
the practice. of major import here is
the possibility
that we've obtained a
theoretical challenge
through practice
shinohai: asciilifeform: is
this
the `blessed` method for use with fg at moment?
mircea_popescu: but
the point of deedbot is
to contain signed matter intended for human hands ; it is why
the signed matter intended for machine processing is a separate mechanism.
☟︎ mircea_popescu: asciilifeform what is
the matter with sending
to deedbot
the item #1 wrapped in a "This is a notarized version of
the original root of
the so and so project, prepared for
this and
that etc" ?
jhvh1: shinohai:
The operation succeeded.
mircea_popescu: come
to
the argument with your argument prepared
to be argued.
mircea_popescu: man... look here, you
throw
total shitfit, i repress poor Framedragger
to make some space for
the discussion, you seamlessly move into vintage lulz and log readings. wth is
this.
phf: i
think woodcoin is best altcoin. everyone else is going for
the whole cyberpunk fastcompany vibe, while
these guys are earnestly doing
their own
thing. it's like an elven
technology department, established by young visionaries despite
the disapproval of
the elders,
their servers all run on
thermal generators somewhere in
the cascadians, etc.
BingoBoingo: WHich is wrong, should be glueing, screwing, doweling, or mortise and
tenon
BingoBoingo: Woodcoin, blocks added
to chain by chopping
mircea_popescu: the dubious part is how exactly it'd pay,
the wallet encryption is strong as far as we know.
mircea_popescu: asciilifeform i imagine it's a global "search all accessible webservers connected
to bitcoin".
mircea_popescu: (Deliziosa creatura, vuoi le mie mutande sporche, le mie vecchie mutande? Sai che ciò è di una raffinatezza impareggiabile? Vedi come sono sensibile al valore delle cose? Ascolta, angelo mio, io ho il più gran desiderio del mondo di contentarti in questo poiché sai che rispetto i gusti. I capricci, per barocchi che essi siano, li
trovo
tutti rispettabili. Sia perché non ne siamo arbitri, sia perché anche il più singola
mod6: Once a solution is realized, I encourage you gentlemen
to work
together, putting previous disagreements aside.
mircea_popescu: this needs a proper discussion, and
the exact reforms/solutions aren't evident yet.
mod6: It could place
those in, just as it does regular deeds.. I'd say, if anyone is a good lisp dev (phf, whomever)
to work with
trinque
to get
this feature put into place.
mircea_popescu: mod6
the only q is whether
this isn't redundant ; and if it isn't, whether
the utility doesn't belie bad design etc.
mod6: I
think it would be really cool
to have something like !!deedbot <wotpaste_of_vpatch> <pgp_sig_of_wotpaste_vpatch>
mircea_popescu always found
the absence of
the range datatype a kind-of lulzy
thing in cs.
mod6: ok, cool!
then just leave
that part out.
mircea_popescu: mod6 hang on hang on, it's not even clear any actual blobs are involved
this
time around.
mod6: i'll even run a mirror of it myself if anyone would like me
to do so.
mod6: if I were him,
this is what I would do:
take
the schematic blob, encode it, clearsign it with a note at
the
top and a hash of its output value. submit
to deedbot. next, edit
the code somewhere or prefereably create a README.txt
that points
to
that deed, create a new fuckgoats_genesis.vpatch and fuckgoats_genesis.vpatch.alf.sig on nosuchlabs.com, point
to
them with your www.
mod6: now. if you need
to ship,
this instant. you either do what you can, as I did, or you do nothing until you can.
mod6: there may be some better solution
to
this with deedbot.
mod6: phf: oh i get it. i lead by example with
trb on
this.
there is no good solution here. we do not want blobs in V, and somewhere between a-z you end up with a blob
that must get signed along with
the vpatches.
mircea_popescu: the part where "please don't make me
turn ascii into
tarballs" is sensible, just, lets see what he actually is
talking about.
phf: mod6: we have
the 3rd party signed
tarballs,
that we can't convert
to v because of
the binary problem. i
think asciilifeform doesn't want
to fallback
to doing
the same with own releases. wants
to solve binary problem, or find some alternative solution.
there's also been a lot of sentiment here against using sha/sig combination. "sign
the
thing itself" etc.
mircea_popescu: thestringpuller neither of
these are "easy" in
the sense of windows. you'll have
to do research and work
the elbow grease
to see results.
mircea_popescu: and much cheaper b) make it encrypt
the same
thing a shitton of
times and see if
there's any key leakage.
Framedragger: (but i
think i got it why you don't want binary
tarball - it'd sure be nice
to be able
to host a plaintext package, with v sigs etc right
there. hm. but
then why deedbot
the *whole*
thing - why not deedbot only hashes of
tarball version of webthing. make webthing available
to view, but also for
those who wish
to verify via deedbot,
too, make
tarball available. not pretty, but worx for *now*.)
mircea_popescu: thestringpuller
the
two evident avenues would be a) power differential analysis (ie, see how much electricity it eats, and parasite waveforms it induces).
this is amply discussed on web re subverting gpg (eventually
they got it
to work with mere microphone, but same principle)
a111: Logged on 2016-09-15 18:10 asciilifeform: 'Can I re-seal after breaking
the center? No. A permanent change is made inside
the flash memory of
the processor.' << which i can't remove and replace with identical-looking one, containing same procedure, exactly why...?
mod6: <+asciilifeform> v should abolish
the idiocy of 'signed
tarballs' << it doesn't have
to. so far, we don't even have anything of
the kind. just vpatches + sigs. 'tis all.
mircea_popescu: if you've internally decided
that you won't ship
the items until "v demolishes
the idiocy of
tarballs" we have a problem.
mircea_popescu: asciilifeform which brings us back
to "what problem are you
trying
to solve"
Framedragger not sure why vgenesis must be used here, even
though in
the future
that's
the correct ideological pathway, sure. but i'm possibly underinformed and out of my element
mircea_popescu: you point
to deedbot which contains signed
tarball. what's
the problem ?