►
A
Good
morning,
good
evening
or
good
afternoon
or
wherever
you
are
good
night,
whatever
time
zone,
it
is,
I
hope,
you're
not
having
a
good
time.
So
where
are
we?
What
are
we
doing,
and
so
we
are
still
doing
what
we
were
up
to
where
we're
gonna
link.
A
You
know
we're
gonna,
take
the
mermaid
diagrams
for
p
tb
decks
and
for
dffml
and
we're
going
to
mesh
them,
and
then
that's
going
to
show
us
that's
going
to
hopefully
guide
us
a
little
more
as
we
go
about
our
about
our
journey
here
so
or
our
adventure
as
it
were,
or
alice's
adventure
eventually
all
right.
So
once
we
get
this
done,
what
are
we
going
to
do
so?
Why
is
this
important?
This
is
going
to
be
the
first.
A
A
We
form
a
record
of
past
executions,
because
the
first
execution
will
suggest
the
execution
of
the
next
execution
and
so
on
and
so
forth,
via
the
strategic
plans.
So,
and-
and
you
know
effectively,
what
we're
showing
here
is
how
you
know
alice
could
do
them
in
house
or
she
could
outsource
the
execution
of
bob
and
her
prioritizer
is
going
to
make
that
call
and
bob's
prioritizer
is
going
to
make
the
call
whether
he
wants
to
schedule
or
not
and
they're,
going
to
communicate
via
tbdx
all
right.
A
Well,
you
know,
let's
forget
about
where
we
were
and
approach
this
with
fresh
eyes,
so
this
looks
like
a
giant
mess.
So
what
are
we
doing
here?
Run
system
contacts
operation
get
bids
all
right.
I
think
that
we
were
in
the
middle
of
reworking
this
because
we
tried
it
like
three
or
three
three,
two,
two
or
three
weeks
now
yeah
so
execute
unpre,
okay,
okay,
let's
just
look
at
the
that
thing.
Let's
look
at
here.
This
is
the
truth.
A
A
A
And
then
alice
will
bid
on
this
herself,
so
we
don't
want
to
get
bids
we'll
fall
through,
but
you
can
also
have
a
situation
where
alex
you
know
maybe
has
some
you
know
will
bid
for
her
own
job,
because
that
way,
that
way,
we
we
the
goal
is
to
always
have
the
same
flow
right.
So
if,
if
we
see
a
new
bid
come
in
and
we're
alice
or
we're
bob
we're
gonna
bid
on
it
the
same
way
either
way
right.
A
So
we
might
take
in
to
the
prioritization
information
for
our
bidding
the
fact
that
we're
bidding
on
our
own
job
right-
and
so
you
know
the
price
will
be.
You
know,
there's
no
price
right
there,
the
the
the
it's
just.
You
know
it
is
what
it
is.
We're
gonna
communicate
the
fact
that
hey
you
know
it
may
cost
this
much
for
us
to
run
it
in
house,
but
you
know
you're
not
going
to
pay
any
extra.
So
you
know
here,
here's
the
here's
sort
of
you
know
we
we
can.
We
can.
A
A
A
All
right
prioritization,
so
if
alice
decides.
A
A
System
context,
self,
okay,
yeah,
we'll
just
say:
prioritization,
okay,
so
and
then
we'll
final
prioritization
will
funnel
prioritization.
A
See
what
happens
here?
Okay,
so
we
go
from
the
lock
network.
We
begin
prioritization,
we
hit
the
prioritizer.
We
have
okay,
so
execute
system
context
itself.
A
A
A
A
A
Okay,
so
there's
a
new
system
context
right,
so
this
is
when
we
have
yeah.
So
this
is,
you
know
our
input
context
or
our
input
yeah.
Our
input
set
input
set,
which
has
an
input
side
context
right,
and
so
you
know
a
new
one
of
these
comes
in
we're
being
asked
to
execute
it
right
it
it
hit.
We
went
through
the
lock
network
and
we're
going
to
execute
now,
so
we
may,
let's
see.
A
A
All
right
so
now,
when
we
say
on
chain
really
what
we're
talking
about
is
just
sending
this
did
blob
somewhere
that
someone
else
can
get
to
it.
Whoever
the
intended
recipient
is
right.
So
that's
sort
of
the
beauty
about
this.
That's
the
real
beauty
about
this
did
stuff.
This
distributed
web
node.
All
of
this
is
like
chain
is
a
very
it's
not
like
the
previous
blockchain
stuff.
So
much
right,
it's
more
of
like
a
distributed
messaging
system,
all
it
actually,
all
it.
I
mean
it's
really
just
a
secure
messaging
interface.
A
It
happens
to
be
that
that
messages
are
you
know
when
you
send
one
message:
it
links
to
the
previous
messages
right
or
whatever
previous
messages.
You
wanted.
It's
just
that's
the
chain
right,
so
it's
not
like.
I
mean
it's
just
it's
a
linked
list
right,
it's
a
cryptographically,
secured
linked
list
and
that's
how
bob's
gonna
know
what's
really
coming
from
alice
and
that
bob
is
actually
going
to
offer
to
execute
this
thing,
and
it's
not
just
coming
from
someone
random,
because
why
the
hell
would
bob
do
that
right.
A
You
know
he's
not
just
going
to
go,
execute
random
people's
code
on
his
computer
unless
he
knows
exactly
what
he's
been
asked
to
execute
and
who
is
asking
him
to
execute
it,
and
you
know
what
assurances
does
he
have
that
this
is
not
going
to
hose
his
whole
setup
right
or
or
cost
him
a
lot
of
money
right,
so
he
he
thinks
that
he
can
do
this
for
alice
right.
Maybe
maybe
bob's
got
some
server.
A
That's
just
sitting
idle
right
or
you
know
yeah
so,
and
you
know
he'd
like
to
be
able
to,
for
example,
this
machine
right
now
that
we're
on
this
is
not
taking
a
lot
of
look.
Look
let's
like.
A
Really
100
cpu,
oh
okay,
whatever
way
to
disprove
my
point
computer!
Well,
now
it's
15!
So
you
know
we
have
a
whole.
A
Whatever
100
minus
16
is
84
cpu
that
could
be
doing
things
right
and
I'm
sure
there's
a
lot
of
other
computers
on
the
network.
Well,
hopefully,
they're
not
doing
a
lot
of
things,
but
you
know
there's
probably
a
lot
of
other
computers
that
could
use
some
compute.
A
You
know
because
they
are
at
a
hundred
percent,
and
you
know
it
would
just
be
nice
if,
if
it
was
a
reasonable
request,
we
could
just
do
it
for
them.
You
know:
there's
really
no
reason
to
have
all
these
computers
on
and
not
doing
anything
they're
sucking
up
power
anyways.
So
it's
like
a
distributed
back
end.
So
basically.
A
How
do
I
get
off
track?
How
do
I
get
off
track?
Okay,
so
new
system
contacts
explore
outsourcing
opportunities,
alice,
get
bids
and
then
so
then
she
should
ensure.
So
basically,
if
she's
gonna
get
bids,
she's
gonna
need
to
ask
get
bids.
A
A
A
So
then
you
know
that
goes
to
send
the
id
doctor
chain,
and
this
is
going
to
this
web
3
will
want
to
put
at
the
top
here.
Then
it
looks
like
so
we're
going
to
put
this
on
top
of
tv
decks.
A
And
maybe
we'll
put
this
below
tv
decks
so
that
that
way,
the
thing
that
we're
copying
from
an
upstream
is-
and
actually
we
should.
Oh
there's
no
comments
in
mermaid,
but
we
need
to
remember
what
we
have
the
video
that
we
source
this
from
tbdx
and
I
think
we
have
the
links
too.
So.
A
A
There
we
go
get
bids
on
system
contacts,
execution
boom,
send
the
id
doctor
chain
boom
over
to
new
inputs
boom
down
to
bob
all
right.
So
when
we
may
need
another,
so
so
bob
may
just
see
this
system
context
fly
onto
the
chain
and
go
ahead
and
bid
on
it
right,
but
we
may
want
to
have
he
may
he
may
do
that
right.
So
he
may
do
that
yeah.
He
may
do
that.
So
he
might
just
do
that
right.
We're
just
gonna
assume,
maybe
yeah
we're
gonna
leave
that
there.
A
So
the
moment
this
system
context
gets
suggested
by
strategic
plan
the
it
should,
you
know,
get
added
as
input
the
input
network
should
transparently
add
it
to
the
chain,
and
bob
should
pick
it
up.
Just
as
alice
did
right
and
it
should
go
for
prioritization
right.
So
in
the
same
moment,
that
alice
is
seeing
it
right
well
about
the
same
moment
right
we
could
make
it
we
can
we
can
we
can.
We
can
gate
it
with
async
io.
This
would
be
fun,
so
we
can.
A
This
is
gonna,
be
this
is
async.
Io
is
great
for
this.
We
can
have
async
io
events
we
can.
This
is
just
fantastic
okay,
we
can
run
it
all
in
the
same
thread:
no
multi-threading,
thank
god.
A
Okay,
you
can
spin
things
out
to
threads
if
you
want,
but
all
this
all
of
this
is
I
o,
which
is
why
this
is
so
great,
and
then
you
can
spin
out
cpu,
bound
or
gpu
bound
to
threads
and
and
we'll
be
able
to
we'll
be
able
to
do
that
automatically
so
that'll
be
sweet
and
we'll
be
able
to
assess
locality
so
that
we
can
optimize
for
what
is
in
cpu-bound
threads
and
what's
in,
we
can
throw
whole
sub-sections
of
data
flows
and
cpu-bound
threads,
because
we
understand
the
locality
of
their
inputs
and
we're
collecting
data
on
their
runs,
and
we
can
optimize
and
then
pick
out
what
things,
what
what
subflows,
what
things
should
be
subflows
in
what
orchestrators
it's
going
to
be
sweet
analysis
is
going
to
do
it
all
for
us
after
we
build
up.
A
A
A
All
right
there
we
go
bob
all
right
boom
watch
that
boom
bob
boom.
This
is
ridiculous.
Okay,
wow
bob
a
lot
of
stuff
here;
okay,
so
get
beats;
okay,
so
get
beats
okay.
So
now,
let's
see
so
bob
bob
prioritizer
new
system
contacts
executed
in-house
bob
execute
system
context,
new
system
context
executed;
okay,
so
okay,
so
bob's
gonna
have
a
new
system
context
from
external
now,
external
neural
external.
A
A
Wow,
okay,
why
is
there
another
thing
going
from
ask
okay,
so.
A
A
A
A
So
yeah
so
we're
going
to
we're
going
to
run
so
we're
going
to
generate
tb,
dex
messages
and
we're
going
to
they're
going
to
be
themselves
within
the
ids,
so
yeah.
So
when
we,
when
we
talk
about
creating
all
this
stuff,
we're
just
going
to
create
an
input
object
and
that
input
object
is
going
to
be.
A
We're
just
doing
this
for
simplicity
right
now,
we're
just
sending.
I
know
I'm
well
aware
this
is
terrible
performance
wise,
but
for
the
simplicity,
stake
of
using
a
single
transport
mechanism
for
our
poc
here,
we're
we're
just
gonna,
send
everything
to
the
chain
right
that
way
we
serialize
to
a
flat
file.
We
pick
up
another
machine
all
as
well,
so
tp
dex,
oh
tb,
let's
put
I'm
putting
it
in
the
diagram,
so
message
is
communicated
via
chain
for
pocket.
A
A
A
A
Respond
with
proposed
did
of
executed
system
context,
given
as
source
currency
all
right.
So
what
is
bob
gonna
need
to
do
for
that?
Well,
he's
going
to
have
to
create
a
did
right,
so
he'll
have
to
take
the
system
context
now
build
a
new
system
context
using
his
top
level
system
context
right
and
if
you
can
do
that,
you
know
if
you
can
generate
a
valid
system.
Context
like
you
know,
hey,
you
know.
This
thing
requires
vms
right.
A
Well,
does
bob
acts
have
this
thing's
gonna
we're
asking
you
to
spin
up
three
vms
and
give
us
ssh
access
to
all
of
them
right?
So
do
you
have
access
to
something
that
will
let
you
spin
up
vms
bob?
Okay?
Well,
you
don't
all
right!
Well,
that's
gonna
be
a
problem
right,
so
you
can't
generate
a
valid
system
contacts.
Therefore,
you
can't
generate
a
bid
you
could,
but
alice
is
going
to
validate
your
bid
right
and
she's
going
to
say
well,
you're.
A
You
know
this
is
not
a
valid
system
context
right,
and
how
do
you
tell
that?
Well,
you
look
at
all
the
inputs
and
make
sure
all
the
inputs
mapped
somewhere
within
the
architecture
right
or
so
so
yeah.
So
basically
bob
should
determine
the
validity.
You
know
his
his
ability
to
to
generate
a
valid
and
we
should
probably
write
this
down
so
determine.
A
A
A
And
then
he's
going
to
go
ahead
and
post
a
a
ask
message.
You
know
with
this
new
did
that
he's
created
with
the
proposed
system
context
you
know
which,
which
basically
this
is
going
to
have
all
the
you
know
more
specifics
on
hey.
You
know
this
is
this
is
what
I'm
saying
is
an
equivalent
system
content,
and
this
is
what
bob's
saying
is,
is
basically
hey.
You
know
here's
the
best
I
can
do
right
and
then
alice
will
get
this
bid
right
and
she's
gonna
she's
gonna
take
a
look
at
it
and
she's
gonna.
A
Yes,
explore
outside
so
get
bids
right
so
ensure
context
on
chain
input
to
chain.
So
right
now
so
new
system
contacts
in
house.
So
now
we
have
what
we're
going
to
see.
This
is
an
operation
right.
So
this
is
an
operation
executing
so,
let's
see,
see
alice,
says,
give
bids
and
then
she
she
fire
and
forgets
or
well.
She
doesn't
forget
she
needs
to
set
some
kind
of
wake
up,
so
she'll
wake
up.
A
So
let's
see
us
check
on
bids.
A
A
And
then
you
know,
maybe
there's
also
like
a
timeout
right,
so
system
contacts
been
retreat
received
prior
to
this
timeout
for
bid
system
context
bid
selection
right.
So
basically,
you
know:
hey
she'll
have
some,
and
this
goes
into,
like
you
know,
how
long
does
she
have
to
think
about
it?
Right
I
mean
this
is
where
we're
kind
of
changing
changing
the
way
we
think
about.
How
do
we
program?
You
know
at
least
alice
right
is?
Is
you
know
it
doesn't
have
to
get
done
immediately?
A
You
know,
take
your
time
find
the
best.
You
know
the
best
solution
for
the
job,
and
then
you
know
get
back
to
us
when
you,
when
you
figured
it
out
right,
don't
just
basically
don't
don't!
No,
no!
No
more,
like
you
know.
Well,
we'll
have
static
decision
trees.
Obviously,
but
you
know
basically
we're
we're
we're
going
to
assess
our
options.
A
First
right
we're
going
to
build
all
the
static
decision,
trees
that
are
available
to
us
and
then
we're
going
to
whittle
down
our
choices
and
and
really
assess
our
options
and
then
we're
going
to
execute
that
right.
So
then,
that's
basically
our
timeout
right
now
or
we're
basically
going
to
see
this
bid
come
in
we're
going
to
say
hey.
This
looks
great,
you
know,
let's
just
go
for
it
right.
A
Okay,
so
alice
other
alice
in-house
on
a
different
machine,
so
she
can
execute
it,
for
you
know
no
cost
right
on
on
a
machine.
That's
already
running.
Basically,
this
is
a.
This
is
a
no-cost
thing
to
run
this
on
this
one
computer
within
alice's
internal.
You
know,
asset
pool,
so
then
she
would
just
you
know
if
she
saw
that
bid
come
in,
she
would
just
accept
it.
Maybe
right
we
might
have
a
strategic
plan
that
would-
or
you
know
we
might
yeah.
A
We
might
have
this-
let's
see
so
that's
gonna
go
into.