127300+ entries in 0.073s
shinohai: Welcome btcvixen .... hafta
to lurk until mircea_popescu comes online
diana_coman: asciilifeform, where
the fuck did I say gcd and lcm same??
diana_coman: (obv. no comments in
the code
to know why it uses lcm
there, so
the supposition above is entirely mine; if anyone sees a different reason for
that, let me know)
diana_coman: in other gnupg-isms: it uses lcm(p-1, q-1) instead of phi itself, supposedly
to make decryption faster but essentially
through making
the decryption exponent smaller; moreover,
this is a sort of buildup on
the whole pile of stink because it quite relies on p and q NOT being really random (if
they really are random, chances are gcd(p-1, q-1) relatively small and
therefore phi and lcm not really all
that different anyway)
☟︎ mircea_popescu: this
then allows
the emergence of software distribution networks, and so on.
a111: Logged on 2018-01-08 02:02 mircea_popescu: 5. client proceeds
to build what it can (if set
to, and as per signatures again) and present
the pile.
mircea_popescu:
http://btcbase.org/log/2018-01-08#1766927 << needless
to say re
this
that just because you
trust someone's signature it doesn't follow
that someone's obliged
to distribute his patches
to you ;
this allows him
to be paid for his work ; you're also more
than welcome
to obtain your patches from someone other
than
their author, provided... you
trust
their signature ; and obviously
the author's STILL not obliged
to distribute patc
☝︎ mircea_popescu: but if i were
to guess, i'd guess
this is where it's headed.
mircea_popescu: oh, obviously :
there's no possiblity of name conflicts, either.
the fact
that my key defines "usg" as x and
trump's key defines "usg" as y allows us both
to live our respective delusions happily.
mircea_popescu: in a sense moving
the problem away into where it belongs.
a111: Logged on 2016-11-15 19:06 asciilifeform: Framedragger: note
that
this was
the laughable piece, 'majority votes', i had nfi how
to cut
the knot of sybil etc.
mircea_popescu: at least
that way i have someone
to
tell his dentist sucked ; and he might negrate
the douche. or vice-versa.
mircea_popescu: rather
than
the current "google search" utterly useless bullshit.
mircea_popescu: item eminently doesn't deal with "programs" or "applications" or anything ; it's a proper url system, it locates resources. can put in
the gns
that you're a dentist in kennebunkport, maine, and if i need a dentist...
mircea_popescu: no more need
to run javascript (ie, unsecured remote code) ; no more need
to "ask people
to upgrade" ; no more anything of
the idiotic sort. if i want a cuntoo/mp-wp/we it is now i have
to ask here, and maybne someone produces one.
this could just as well be handled by
the machines, it's a machine
task.
mircea_popescu: so announcement of a new partch would be author changing his gns entry for X from press-to-q
to press-to-k ; if user is disinterested he jkust gets dragged along, otherwise if user requires he sign local copy he stays put.
mircea_popescu: and
that "bitcoin" can mean whatever
the fuick you choose it should mean.
mircea_popescu: obviously
the fact
that a bitcoin
txn container will come with "this
takes bitcoin" will not result in you recompiling bitcoin for each
tx -- you already have it ;
mircea_popescu: obviously we all do
this manually at present, but we really utterly don't want
to.
mircea_popescu: so in
the atomic case, practically overengineered but for
the sake of understanding each other : each v
tree should simply say it wants
to be pressed by a v presser, and
the client should inquire wtf
that is and get one.
mircea_popescu: you can make upon a private lan
the agreement
that "unicode" is how you call
the
tga decoder, and lo and behold... you read
tga with
the unicode displayer now.
trinque: this 1) agreeing ahead of
time with pantsuit council on what schemas exist or 2) infinitely nested lookup
tables for viewers is exactly what killed
the prior idiots
mircea_popescu: there's no requirement
that
the gns even be properly global ; it's your job
to make it global-locally and
that's
that.
mircea_popescu: i don't have
to have
the people who want
to use unicode
to agree
to anything in particular ahead of
time, and if
tomorrow
they decide
to call it fuck-bush instead,
they can and nothing gets broken
mircea_popescu: ie,
the whole
thing can be built out of parts without getting in anyone's way whatsoever.
mircea_popescu: in
this way, someone who wants
to read unicode will accept containers of
that
type and have in his seals sigs of people who wrote such a
tree ; and who wants
to see (what now is) .avi or .gif or etc similarily.
mircea_popescu: 5. client proceeds
to build what it can (if set
to, and as per signatures again) and present
the pile.
☟︎ mircea_popescu: 4. in all cases, a container consists of a
tag and binary data ;
the
tag section of
the gns
tells it what v
trees are acceptable applications upon
that data.
mircea_popescu: 3. open link
to machine, explain what it is it wants, get an answer in
the form of "here's a list of document containers, which do you want". it explains which it wants (such as "all" or w/e) and it gets a container containing
them. maybe it's gzipped or whatever.
mircea_popescu: 2.
the client queries whichever gns it is configured
to, and inquires as
to contacting machines which hold data related
to
this item. it receives a list of
them, on
the basis of what signatures mircea_popescu
told it
to
trust signed.
mircea_popescu: now let's look at how
things work. mircea_popescu decides one day he wants
to find out about sgml, and
therefore instructs his client.
mircea_popescu: v is basically just a lengthy pile of code
trees, made out of patches also signed by keys ; in practice its coinvolved with
the v-application, which can press etc, but let's here distinguish.
mircea_popescu: there's no obligation on anyone
to maintain any particular portion of it, or
to communicate it
to anyone -- can be run locally just as well (in practice, equiv rto "only
trust my signature").
mircea_popescu: so gns is basically just a lengthy registrar of assignments, a=b sorta
thing, signed by keys.
mircea_popescu: let's put
this in different
terms :
the happiest life is, empty planet, where man can go picking berries / quals/womenz as he feels inclined. everything fits well and all is good.
the necessary correlate is
that
the petri dish slowly fills, at which point incompatibles will result in perpetual war and strife. all is not so good.
mircea_popescu: trinque
the disaster i perceive is
this :
that a)
the worst world is
to have a lot of ad-hoc implementation of sexpr-datastructs but b)
the cheapest way is a and c)
the complete construction is unachievable.
trinque: dunno it ends up being any more complicated
than existing DHTs but for
the urge
to make a giant, complex, systemdtron with myriad ways of describing
the same
things.
trinque worked under a semantic-web-head for about 4-5yrs, built several incarnations of
this kind of meta-browser monstrosity
mircea_popescu: and indeed, management of
the implicit semantics in all
this will be
the end of
the world.
mircea_popescu: "and
the image is in
the list of
things you don't want so we omitted
that also"
mircea_popescu: then it could
tell you "well, here's
that
thing you wanted in a browser, because
that's where it says it goes, and we couldn't process
the song part because nobody you
trust signed an implementation of anything on
the list of song
things, but
the
text is
there as per so-and-so-textizer signed by x y and z"
mircea_popescu: anyway, so : a container could
tell you it's a document, and it contains a
text, a song and a picture. should
that container end up on a list of
things you're interested in, your client could
then look up what
to do with documents, and how
to parse
texts, songs and images.
a111: Logged on 2015-12-09 11:38 mircea_popescu:
http://dreamsongs.com/ <<< how
to be retarded.
today, richard "i don't have a name and i was born without a face, here's some salmon background" gabriel.
mircea_popescu: asciilifeform in ~principle~ a container could
tell you what class of procedure it wants
to be eaten by ; and gns could
tell you which v-trees implement it.
mircea_popescu: it's like watching middle class single parent 12 yos at work, mendacious manipulative little shits
that imagine
their nonsense is entirely opaque
to anyone and everyone.
mircea_popescu: well
thank you bitch, you ~could~ have said
this a year ago and saved everyone
the hassle. o wait, cuz needed
to appear all humble and shit for a year
to poach users / devs i see i see.
mircea_popescu: I
think
this is not a good outcome for
the Emacs user community, and I hate
to have
to criticize
the implementation of GNU Emacs 19, but at some point clean code and good structure counts more
than a false sense of parsimony, and
that point has been reached."
mircea_popescu: We believe
that
the merge in
the other direction is
the proper way
to do
things, and
this is
the course we shall pursue.
There are
things
that you have done
to GNU Emacs 19
that are good, and we believe
that
the best way
to proceed is
to clean
them up and put
them in Lucid Emacs 19. Maintainability is important
to us because even
though GNU Emacs 19 has no warranty, we must support a warranty on our commercial product which i
mircea_popescu: life for
the maintainer easy.
This is proven by
the fact
that Arceneaux was unable
to add
the extensions we needed
to an early version of GNU Emacs 19 given somewhat over a year's
time
to do it and by
the fact
that it has
taken you and Blandy so long
to get GNU Emacs 19 out.
mircea_popescu: lmao so after stringing along "oh, we have spent so much, we hope
to still work
together,
tell us how" for a year, gabriel is like "Unfortunately, GNU Emacs 19 is not constructed following modern programming practice - even
though
there are procedural interfaces
to some functionality groups, one could not say
that abstraction boundaries are maintained.
The mish-mash of data structures is just
too clumsy and incoherent
to make
mircea_popescu: i'm not asking how it'd be rendered, it god damn ed be rendered
the same. im asking...
mircea_popescu: so
tell me how is pstp going
to work without a gns
to reference container content names ?
mircea_popescu: isn't
this definitionally what naggum proposes one not do ?
mircea_popescu: so sgml is a flaming pile of shit, oda is a waste of paper and i even read french, what exactly are
these wheels one's supposed
to not reinvent ?
mircea_popescu: "oh but mp, before you reinvent
the wheel you should read
the other
things other people did before" "motherfucking hell what if I CAN NOT READ
THESE PEOPLE!"
mircea_popescu: basically,
technocucks so fucking petrified of
their emotions and
the structure of power,
they will retreat into bikeshedding as a comfortable substitute for sanity.