►
From YouTube: Workshop: Engineering Filecoin's Economy with Juan Benet
Description
In this live zoom presentation, we discuss the mechanisms and models presented in the recent report Engineering Filecoin's Economy. We will also discuss a number of questions that have been submitted by the community. To have your questions covered, please submit your questions in the linked form below.
Read the report:
https://filecoin.io/blog/introducing-the-filecoin-economy/
Submit your questions:
https://docs.google.com/forms/d/e/1FAIpQLSd0n3kBOa7szVIn20yEGtTqi7vtv9PhtsxGqnJOve4t2Ju1ZQ/viewform
A
Hey
everyone.
I
see
the
first
few
people
trickling
into
the
room.
We're
gonna
wait
just
a
little
while
for
more
people
to
show
up
and
then
we'll
get
started.
With
this
workshop,
seeing
25
people
so
far
and.
A
Counting
how's
everyone
doing
assuming
a
lot
of
you
all.
Are
miners
have
space
race
going
or
almost
halfway
there?
It's
kind
of
crazy
to
think.
A
A
A
Very
north
america-centric
so
far,
but
I
suspect
some
people
are
being
a
little
quiet.
All
right,
we're
three
minutes
past
our
start
time
more
people
might
trickle
in,
but
for
now
we
can
get
started
hello!
Everyone!
Thank
you
for
joining
this
workshop.
A
A
I
know
there's
been
a
lot
of
demand
to
understand
how
some
of
our
crypto
economic
decisions
have
been
made
and
really
really
hoping
that
that
we'll
be
providing
a
lot
of
clarity,
a
lot
of
insight
into
all
of
the
decision
making
and
work
that
has
gone
into
this
for
our
miners,
we're
about
a
week
and
a
half
into
space
race,
and
I
hope,
that's
exciting
for
developers
and
other
supporters
of
the
community.
A
We've
never
been
closer
to
mainnet
launch
and
it's
a
it's
an
exciting
time
in
in
the
final
coin
project
so
yeah,
I
today's
workshop
will
be
presented
by
zx
and
juan
zx
has
been
absolutely
spearheading.
The
crypto
econ
work
for
the
file
plan
project.
A
So
there's
no
better
person
to
be
telling
you
how
a
lot
of
these
decisions
were
made
and
kind
of
what
the
ideas
behind
them
were
and
juan
is
hopefully
a
familiar
face
to
all
of
you
he's
the
the
founder
of
the
filecoin
project.
So
I
will,
I
will
turn
it
over
to
them
and
I
hope
I
hope
everyone
has
a
good
time
and
has
a
lot
of
the
questions
answered.
Juan
will
be
kicking
things
off
so
over
to
you.
One.
B
Thank
you
very
much.
Hopefully
I
can
do
justice
to
the
to
that
green
intro.
So
welcome
everyone.
So
this
is
what
we
think
will
be.
You
know
the
first
of
of
you
know
a
number
of
workshops
over
time
that
we'll
have
about
the
falcon
economy.
B
We
hope
to
kind
of
kick
things
off
with
with
some
good
discussion
and
then
just
we
hope
to
also
have
some
continued
discussion
on
slack
as
as
we
have
been
having
for
for
many
weeks
now,
and
then
I
hope
that
you
know
this.
One
will
be
introductory
we'll
kind
of
go
through
kind
of
the
high
level
goals
of
the
of
the
economy.
Why
and
kind
of
cover
various
different
sections
and
then
we'll
we'll
go
through
kind
of
the
report
that
we
released
recently
in
some
detail.
B
We
we
will
again
given
that
the
timing
here
you
know
it's
not
like
a
five
hour
workshop.
We
only
have
one
hour
here,
so
we'll
kind
of
go
through
each
section.
To
give
you
a
good
understanding
at
a
high
level,
then
we'll
have
then
we'll
go
into
a
number
of
questions
that
we've
received
so
far.
So
we
put
out
a
kind
of
a
call
for
questions
so
that
we
could
aggregate
the
most
common
questions
and
the
most
important
questions
and
give
more
detailed
answer
answers
in
those.
B
So
we'll
go
through
those
questions
and
then,
if
there's
time
that
we
can
maybe
take
some
questions
from
from
chat.
Definitely
as
we
go
through
ask
questions
as
you
have
them,
we
probably
won't
answer
them
right
away
in
the
moment,
but
if
you,
if
you
ask
your
questions,
we'll
be
able
to
kind
of
touch
on
them
as
we
either
as
we
go
or
at
the
very
end
and
yeah.
This
is
this:
is
an
experiment
we'll
see
how
this
workshop
goes?
B
If
it's
very
useful
and
valuable
for
people,
then
we
can
do
more
more
over
time
and
diving
into
into
certain
areas
to
explain
them.
Explain
them
in
more
detail
great.
So
with
that
I'll
share
my
screen
and
go.
B
So
this
is
like
a
just
a
again
a
super
high
level,
high
level
intro.
B
The
the
thing
to
remember-
and
we
kind
of
chose
this
diagram
to
to
emphasize
it-
is
that
we're
in
this
period
of
time,
where
we're
creating
these
internet
native
systems
and
economies
that
you
can
think
of
as
public
utilities
in
a
sense,
but
that
are
basically
as
large
as
cities
or
potentially
even
small
nations
and
those
kinds
of
economies
and
systems
are
being
run
by
people
across
across
the
world,
and
these
systems
have
a
whole
bunch
of
features
that
effectively
make
them.
B
You
know
full-on
economic
structures
with
many
different
kinds
of
participants
that
have
different
incentives,
different
goals,
but
that
together
are
providing
some
some
service
or
some
utility
to
the
rest
of
the
world.
So
one
of
the
long-standing
models
that
people
have
been
using
to
to
describe
these
systems
is
to
model
them
as
a
as
kind
of
export
economy.
B
Since
you
can
think
of
the
the
whole
system,
as
as
a
network
with
various
different
participants
and
different
roles
and
then
using
the
token
of
the
system
to
kind
of
transact
within
the
system
and
then
having
some
kind
of
export
structure
to
the
rest
of
the
to
the
rest
of
the
world,
and
so
you
can
think
of
the
of
the
network,
the
falco
network,
as
it's
kind
of
buying
raw
materials
and
kind
of
potentially
importing
raw
materials
or
potentially
producing
them
at
some
levels,
especially
once
you
get
into
markets
that
are
able
to
trade
in
powerpoint
for
for
buying
drives
and
so
on.
B
But
then,
for
the
most
part
exporting
the
service
out
to
the
rest
of
the
world
of
of
the
cloud
storage
service,
it's
kind
of
like
the
high
level
model.
So
we
yeah
we
put
out
this
this
report,
we'll
kind
of
go
into
it
in
a
bit
more
detail
later,
I
want
to
reference
another
talk
that
I've
given
in
the
past
that
kind
of
goes
into
detail
on
other
parts
of
the
protocol
at
different
levels
and
looks
at
it
from
a
protocol
perspective
or
from
a
blockchain
perspective,
and
so
on.
B
So
a
couple
important
data
points
for
everyone
to
remember
and
kind
of
why
why
falcone
is
here?
We
are
in
the
space
transition
as
a
species
where
billions
of
humans
and
trillions
of
computers
now
live
and
work
together
and
we're
deeply
integrated
and
we
are
generating
tons
of
data
and
computation
constantly,
and
you
can
think
of
most
of
our
lives
now
being
and
most
of
our
activity
being
assessed
by
computers
in
some
way
or
another.
B
The
internet
is
constantly
granting
us
new
superpowers
and
all
of
that
comes
in
form
in
the
form
of
applications
that
we're
running
and
those
applications
tend
to
again
generate
tons
of
data
that
needs
to
be
stored
and
backed
up
somewhere
and
beyond
being
stored
somewhere.
We
want
to
have
control
and
proper
encryption
and
sharing
controls
over
over
how
that
data
is
distributed
and
shared
to
to
other
parties.
B
So
it
has
become
an
extremely
important
part
of
our
of
our
of
our
species
and
our
civilization
to
be
able
to
have
an
internet
native
system
for
storing
and
safeguarding
safeguarding
all
this
information.
B
This
is
only
going
to
accelerate
and
get
more
intense,
as
many
new
forms
of
computation
arrive,
so
things
like
augmented
reality,
virtual
reality,
pretty
much
interfaces
and
so
on.
I'm
sure
that
many
folks,
so
that
so
the
the
the
recent
kind
of
a
demo
of
of
the
neural
link
system
like
that
systems
like
that
are
going
to
be
become
more
and
more
and
more
commonplace,
and
suddenly
our
our
interaction
as
humans
with
the
computing
medium
is
going
to
get
higher
and
higher
bandwidth
and
more
and
more
data.
B
That's
highly
sensitive
that
we
need
to
have
kind
of
individual
control
over
is
going
to
increase,
so
you
know
the
the
amount
of
data
in
the
world
is
just
increasing
tremendously
and-
and
these
are
kind
of
rough
projections
made
by
by
various
parties
that
kind
of
plotted,
in
terms
of
of
you,
know,
numbers
of
zetabytes
that
we're
getting
into
all
of
these
are,
of
course,
you
know
super
unclear
some
looking
at
the
past,
we
can
roughly
estimate
and
know
the
rough
amount
of
data
looking
at
the
features
is
highly
unpredictable,
but
we
know
that
we're
going
to
be
generating
an
enormous
amount
of
amount
of
data,
and
so
we
want
systems
that
are
able
to
cope
with
that
level
of
of
information
at
the
end
of
the
day
like
these
are
the
I
want,
of
course,
I
repeat
this
a
lot,
so
I
won't
go
into
detail
here,
but
the
long-term
goals
is
to
build
a
decentralized,
efficient
and
robust
foundation
for
humanities.
B
Information
is
that
each
of
these
words
have
a
lot
of
the
features
of
the
protocol
on
the
network
and
the
economy
relate
to
each
one.
Of
these
words,
decentralization
is
achieved
by
building
an
economic
system,
with
incentives
for
all
kinds
of
different
parties
in
different
nations
and
so
on.
The
efficiency
that
the
economy
supports
efficiency
by
creating
market
structures
that
allow
us
to
optimize
and
and
provide
services
at
better
and
better
cost.
B
We
can
achieve
robustness
by
enabling
many
different
parties
across
the
world
to
enter
and
not
creating
moats
or
centralized
systems
that
prevent
parties
from
entering
and
providing
their
services,
and
we
can
really
make
sure
this
is
a
proper
foundation
by
by
at
the
end
of
the
day,
providing
a
system
that
that
that
everyone
can
rely
on
and
trust
without
having
to
trust
any
one.
B
Individual
party,
where
the
system
as
a
whole
will
continue
going
on
and
survive
all
kinds
of
events,
so
you
can
use
it
truly
as
a
foundational
system
for
application
building
without
having
to
to
trust
in
the
continued
good
behavior
of
anyone.
Any
one
party,
so
yeah,
that's
kind
of
the
you
know.
An
important
detail
here
as
well
is,
at
the
end
of
the
day,
the
data
that
we're
that
we're
collecting
and
that
we're
storing
in
all.
B
In
all
these
systems,
it
is
about
you
know,
all
kinds
of
applications,
so
certainly
there's
a
lot
of
organizations
with
a
lot
of
data
they
want
to
store.
But
ultimately
the
thing
that's
going
to
drive,
the
the
most
usage
is
going
to
be
applications
and
so
having
deep
integrations
between
falcon
and
all
kinds
of
applications
is
going
to
matter
a
lot.
B
So
you
can
think
of
the
you
know,
kind
of
the
whole
growing
falcon
ecosystem
as
building
the
storage
services.
Infrastructure
of
you
know
the
the
next
decade
to
be
able
to
provide
this,
the
storage
facilities
for
for
all
kinds
of
applications,
and
so
this
you
know
the
the
market
economy
that
we're
trying
to
create
is
one
where
we
can
provide.
You
know.
B
On
the
one
side
you
have
clients,
with
with
all
kinds
of
requests,
being
able
to
to
use
the
storage
services
of
storage
providers,
and
so
this
is,
you
know
you
can
start
thinking
of
the
different
groups
that
that
are
participating
in
the
in
the
network.
Then
we've
roughly
kind
of
formulated
these,
these
five
groups,
clients,
miners
developers,
token
holders
and
ecosystem
partners,
we'll
look
at
them.
B
A
little
bit
more
details
detail
later,
it
looks
like
here
in
the
central
part
that
you
know
we're
in
the
middle
of
the
space
race.
I
think
roughly
we've
hit
88
petabytes
in
around
nine
days,
or
so
I
think
in
the
latest
last
days,
we're
kind
of
growing
at
a
around
10
petabyte
pippy
bites
a
day
which
is
enormous
amount
of
storage
like
this.
B
Is
this
truly
tremendous
storage
rate
source
growth
rate
and
and
if
we're
looking
at
it
from
from
a
yearly
perspective,
if
we,
if
we
can
keep
that
that
up,
you
know
we
can,
we
can
really
start
hitting
in
the
kind
of
like
three
and
a
half
exabyte
range,
so
that
would
be,
you
know,
truly
phenomenal
if
we
could,
if
we
could
do
that
and
and
just
to
kind
of
to
benchmark
there
like
that,
that
starts
being
kind
of
a
significant
significant
player
in
the
in
the
in
the
global
scale
of
storage.
B
You
know
huge
kudos
and
thanks
to
all
the
folks
that
are
running
blockchain
explorers
and
all
the
participants
in
the
in
the
space
race
really
exciting.
One
of
my
favorite
graphs
here
is
just
like
how
well
distributed
the
storage
power
is
across
a
lot
of
different
folks.
That's
phenomenal.
This
is
one
of
the
best
best
power
distributions
that
I've
seen
in
any
blockchain
and,
like
that's,
really
phenomenal,
to
see
this
level
of
decentralization
great.
B
So
with
that,
let
me
switch
over
to
to
the
report
and
then
we'll
we'll
dive
in
you
can
see
that
correct
the
page
sweet,
so
I
kind
of
covered
like
the
high
level
here
you
know
there's
a
lot
of
you
know.
One
of
the
key
introductory
points
is
just
how
much
how
much
data
we're
storing
in
the
world
but-
and
we
chose
like
one
one
simple
graph-
to
show
that
there's,
of
course,
all
kinds
of
other
ways
of
of
segmenting.
B
This
data,
looking
at
different
kinds
of
distributions
that
are
coming
from
different
kinds
of
applications
and
systems
and
so
on,
and
also
the
the
the
amount
of
data
growth
in
different
geographies
matters,
a
lot
through
the
network
or
or
other
kinds
of
systems
like
that.
So,
at
the
end
of
the
day,
the
popcorn
network
must
be
able
to
adapt
to
the
demand
of
storage
and
storage
services
around
the
world,
but
just
kind
of
the
the
scale
here
is
the
staggering
starting
at
staggering
amount.
B
See,
if
I
don't
know,
if
you
want
to
add
anything
to
that
part,
otherwise
we
can
yeah.
I
guess
yeah
one.
One
key
point
here
is
you
know
the
combination
of
different
features
around
you
know
falcon,
bringing
together
verifiable
storage
where
you
you
can
guarantee
the
the
correct
storage
of
data
in
the
long
term.
B
It's
an
extremely
valuable
and
useful
property
that
hasn't
been
around
at
this
kind
of
scale
before
and
so
we'll
see
what
kinds
of
use
cases
and
applications
come
come
from
that
one
important
property
is
open
participation
being
able
to
having
anybody
be
able
to
enter
the
network
without
having
to
know
anybody
else
and
being
able
to
join
and
provide
storage
power
in
the
network
and
provide
a
storage
service
and
so
on.
B
This
will
get,
of
course,
easier
and
easier,
as
we
as
things
like
the
retrieval
market
and
so
on,
become
part
of
the
part
of
the
system
right
now,
there's
kind
of
like
a
high
barrier
or
entry
for
people
to
be
to
be
storage
miners,
but
as
we
improve
the
proofs
and
as
we
get
other
kinds
of
mining
like
retrieval
mining,
that
bar
barrier
for
participation
is
going
to
drop,
lower
and
lower
and
lower,
and
so
that
that
will
enable
a
whole
host
of
other
participants
to
join.
B
One
of
the
really
key
things
here
is
that
we
try
to
design
this
whole
system
all
over
the
place
to
be
to
really
enable
people
to
optimize
locally
for
their
their
regions
and
so
and
to
kind
of
be
able
to
provide
direct,
direct
access
to
content
and
data
directly
from
minor
to
to
participant
and
so
clients
being
able
to
select
which
miners
to
store
data
with
in
the
geographies
that
they
care
about
in
the
in
the
regions
they
care
about
or
with
the
kinds
of
facilities,
and
so
on
that
they
care
about
enables
kind
of
an
optimization
in
the
marketplace
that
that
may
not
be
be
able
and
may
not
be
possible
in
other
networks
and
other
designs,
and
one
of
the
the
the
hopes
and
goals
over
time
is
that,
of
course,
miners
will
will
specialize
over
time
and
and
perhaps
we
might
see
that
certain
kinds
of
use
cases
will
will
start,
preferring
certain
kinds
of
minors
and
so
on.
B
B
Years,
cx,
if
you
want
to
add
anything,
feel
free
to
jump
in
otherwise
I'll
keep
going.
C
Yeah,
I
think
it's
important
to
highlight
that
it's
like
we're,
building
the
basic
infrastructure
to
allow
for
other
things
to
evolve
and
emerge.
As
long
as
everything
is
serving
contributing
back
to
the
same
goal
of
the
network,
then
we
are
achieving
something
very
successful
here.
B
Yeah,
so
once
we
start
looking
at
the
the
economic
structure,
you
know
the
at
the
highest
level.
We
have
to
remember
we're
building
a
a
marketplace
for
data
and
for
data
storage
in
the
data
service,
and
so
that
you
know,
there's
a
few
basic
set
of
kind
of
commands
or
or
actions
that
the
different
parties
take
in
order
to
be
able
to
do
that.
B
A
lot
of
you
who
have
participated
in
the
space
race
or
have
played
around
with
with
the
test
nets
and
so
on,
are
probably
familiar
with
the
idea
of
being
able
to
propose
deals.
Then
miners
being
able
to
accept
or
reject
deals,
then
clients
being
able
to
transfer
the
data
directly
to
miners,
which
can
happen
online
or
offline.
So
you
know
you
could.
If
the
data
is
big
enough,
you
could
send
it
through
through
offline
means,
like
packaging
and
so
on,
and
then,
once
the
data
arrives
with
miners
miners,
then
putting
that
data.
B
The
data
for
that
deal
into
sectors
sealing
those
sectors,
so
they
can
become
provable,
prove
that
into
the
chain
and
start
earning
not
only
the
deal
rewards,
but
also
the
block
reward
from
it
from
an
increase
in
the
in
their
power
and
then
later,
of
course,
clients
being
able
to
retrieve
retrieve
the
file.
So
right
now.
B
One
important
caveat
here
is
because
the
proofs-
because
we
don't
have
fast
enough
proofs
to
provide
super
fast,
retrieval
straight
from
the
sealed
seal
storage
which
we
hope
to
have
in
the
future,
because
we're
not
there
yet
we're
in
the
mode
right
now,
where
we're
kind
of
spending
more
storage
in
order
to
provide
really
fast
retrieval
and
that's
kind
of
a
requirement
now,
and
so
we,
the
the
ceiling,
amount
ceiling
and
storage,
allows
you
to
to
prove
it
to
the
to
the
network
for
kind
of
like
the
the
keeping
an
extra
copy
around.
B
That's
unsealed
allows
the
fast
retrieval
over
time
that
that
requirement
will,
we
hope,
will
go
away
as
we
get
dramatically
faster
process
replication
that
allow
you
to
unseal
data
directly
from
the
source.
That's
sort
of
the
goal
where,
where
we're.
B
B
And
the
we
already
covered
a
bit
of
you
know
kind
of
how
to
model
the
the
expert
economy
in
general,
but
you
can
think
of
all
these
different
parties
that
we
we
talked
about
the
miners,
the
clients,
the
token
holders,
the
developers
and
the
ecosystem
participants
as
taking
various
different
roles
in
the
in
the
economy.
And
you
know
those
larger
segments
are
not
uniform.
You
can
think
of
storage
miners
and
retrieval
miners
and
other
kinds
of
miners
that
might
emerge
in
the
future
as
providing
specialized
services.
B
You
can
also
think
of
developers
as
fulfilling
various
different
kinds
of
roles
like
developers
of
the
protocol
and
researchers
and
so
on,
or
developers
of
applications
that
are
building
ways
to
use
the
storage
right
so
developers
and
participants
that
join
the
economy
and
are
building
applications
that
use
the
use.
The
network
will
then
strengthen
the
whole
economy
and
will
help
everybody
operate
better.
B
So
the
you
know,
one
big
part
of
this
is
to
really
think
of
this
whole
system,
as,
as
you
know,
the
whole
economy
there's
all
kinds
of
competition,
internal
competition
within
different
agents
inside
of
the
economy,
but
in
a
kind
of
a
macro
scale,
the
whole
economy
either
works
together
either
succeeds
together
or
fails
to
get
together.
B
So
that's
a
really
important
piece
here,
where
you
want
these
infrastructures
all
over
the
place
where
you
have
both
a
kind
of
the
internal
competition
to
be
able
to
optimize
and
create
a
more
efficient
system.
But
you
also
want
us
the
larger,
omega
micro
structure,
structures
that
enable
the
entire
community
as
a
whole
to
succeed
together
exactly.
C
And
it's
not
like,
I
think
the
mindset
here
is:
it's
not
zero
sum
anything
like
for
us
to
create
a
very
useful
and
a
reliable
network.
People
do
need
to
come
together
and
collaborate
and
then
part
of
the
cornerstone
of
blockchain
technology
is.
It
provides
interesting
way
for
us
to
coordinate
and
collaborate
together
as
a.
B
Species
great
so
jumping
into
the
the
partition
participants
in
the
economy.
We've
talked
a
bit
about
miners
and
clients,
but
let's
talk
about
some
of
the
others,
so
yeah
with
developers
we
can.
We
can
have
protocol
developers,
application
developers,
tool,
developers
and
service
developers
so
think
of
people
being
able
to
provide
better
systems
and
better
tooling
for
miners
and
other
search
providers
or
potentially
for
clients.
B
So
imagine
developers
coming
and
creating
something
like
a
kind
of
like
a
marketplace
ui
where
you
can
see
different,
all
the
different
miners,
their
facilities
and
you
can
see
pictures
of
their
of
their
setups
and
they
can
kind
of
advertise
all
their
added
services.
So
all
of
that
would
be
extremely
useful
and
valuable.
B
A
really
useful
and
valuable
addition
to
the
to
the
network,
in
addition
to
the
economy
that
will
come
from
from
from
the
developer
group,
then
the
kind
of
the
group
of
of
token
holders-
that's
one
of
you-
know
really
key
component
to
all
of
these
systems-
is
parties
being
able
to
you,
use
the
currency
and
be
able
to
buy
and
sell
tokens
from
all
the
other
parties
to
to
help
those
parties
offset
their
costs
right.
B
We
also
think
that
there
there
might
be
all
kinds
of
possibilities
for,
for
the
whole
defy
world
to
to
connect
to
falcon
in
a
very
interesting
way,
to
provide
things
like
loan
structures
and
so
on,
for
for
parties
being
seeking
to
to
participate
in
the
economy,
whether
it's
to
finance
things
like
larger
setups
or
you
know.
Maybe
a
new
party
wants
to
open
a
whole
new
miner
and
wants
to
raise
capital
in
order
to
be
able
to
do
that
and
so
being
able
to
do
kind
of
loans
through
the
defy
system.
B
Ecosystem
becomes
a
really
valuable,
useful
thing
or
potentially
being
able
to
do
systems
where
people
are
buying
and
selling
the
actual
raw
materials
directly
with
with
these
stock
concerns
so,
for
example,
being
able
to
buy
and
sell
disks
or-
or
you
know,
cpus
and
gpus,
and
so
on.
B
All
the
stuff
that
goes
into
actually
providing
the
service
doing
that
without
so
on,
will
require
a
very
thriving
economy
of
of
token
holders,
as
well
as
we've
seen
with
ethereum
and
other
and
other
systems,
and
then
ecosystem
partners
are
a
key
piece
here,
where
this
is
kind
of
like
a
larger
catch-all
term
for
larger
groups
who
are
providing
kind
of
some
some
differentiated
or
unique
service
into
the
ecosystem
and
are
kind
of
you
know
you
can
think
of
vertical
labs
or
or
or
the
platform
foundation
or
or
even
new
new
and
exciting
groups
that
are
joining
like.
B
B
So,
for
example,
grant
programs
are
a
good
good
good
example
for
the
kind
of
thing
that
ecosystem
partners
might
do
or,
for
example,
suppose
that
a
very
large
manufacturer
of
of
storage,
like
think
of,
like
disc
manufacturers
like
seagate
and
western
digital
and
and
so
on,
and
get
involved
in
the
network
and
are
starting
to
participate
directly
with
the
network.
Then
they
would
sort
of
become
these.
B
These
high-level
ecosystem
partners-
and
so
that's
that's,
a
really
important
role
in
the
whole
system
for
people
to
be
able
to
to
to
to
participate
in.
I
would
think
of
kind
of
like
the
verifier
network,
as
part
of
as
part
of
that.
A
B
Right
great,
so,
let's
move
on
to
storage
services
on
filecoin,
so
you
know
once
so.
We
talked
a
little
bit
about
kind
of
things
from
the
from
the
the
high
level
client
minor
operation,
but
looking
at
it
a
bit
from
the
notion
of
the
life
cycle
of
minor.
B
B
You
you
need
to
commit
capacity
into
the
network
and
and
the
commitment
of
capacity
requires
a
collateral
and
that's
so
that
when
once
you
start
earning
power
in
the
network,
if
you
don't
provide
the
proper
power
in
the
network
and
you
with
the
reward
that
you're
getting,
you
may
need
to
be
slashing.
So
that's
that's
where
that
is
coming
from.
We.
We
need
a
a
a
portion,
an
economic
structure
there
and
a
mechanism
that
allows
the
network
to
slash
participants
that
are
that
are
signing
up
to
commit
and
and
store
capacity.
B
But
actually
don't
do
so
once
kind
of
enough
capacity
is
reached
to
to
become
kind
of
the
minimum
miner.
At
that
point,
a
miner
can
start
minting
blocks
and
creating
blocks,
and
at
that
point
they
can
start
earning
the
block
reward,
which
is
one
of
the
strongest
incentives
in
the
entire
system,
and
so
you
know
being
able
to
commit
a
lot
of
capacity.
Get
to
a
certain
reach.
B
A
certain
threshold
to
become
to
mint
blocks
is
a
key
part
once
you're
there
as
a
miner,
you
can
continue
adding
capacity
and
potentially
start
sealing
useful
storage
into
into
that
capacity,
and
so
that's
kind
of
when
miners
start
making
deals
with
users
and
with
clients,
and
so
clients
and
minors
find
each
other
arrange
for
a
deal
to
happen,
transfer
the
data
either
online
or
offline,
and
and
then
that
data
gets
sealed
into
into
storage.
B
So
the
the
clients
at
that
point
put
up
a
a
a
collateral,
and
so
when
they
do
so,
both
the
clients
and
the
miners
put
up
collaterals
associated
with
that
deal.
If,
if
the
miner
you
know
fails
to
to
succeed
in
the
in
in
keeping
around
that
deal,
then
that
that
collateral
is
is
learned
if
the
client
there
are
certain
conditions
where
ideal
collateral
on
the
client
side
is
useful,
we
had
other
mechanisms
that
we're
going
to
make
use
of
it
in
the
right.
B
Now
we
don't
have
a
minimum
at
the
moment,
but
but
there
are
all
kinds
of
structures
that
might
actually
use
that
more
effectively
and
then
and
and
for
example,
one
of
the
the
key
things
here.
That
is
often
a
question.
Is
you
know
what
happens
if,
if
somebody's
trying
to
store
something
that
that
that
miners
don't
want
to
store
anymore,
can
they
get
away?
Can
they
get
they
get
out
of
the
deal?
B
And
so
one
of
the
goals
in
the
long
term
is
to
be
able
to
have
a
collateral
to
slash
there
for
for
clients,
that's
not
in
place
right
now,
but
in
the
future
we'll
be
able
to
to
have
that
kind
of
that
kind
of
system.
We
expect
that
to
come
through
a
falcon
improvement
proposal.
B
Then,
after
that,
you
can,
you
can
think
of
kind
of
the
full
life
cycle
of
miner
once
they
have
a
lot
of
capacity
or
or
sealed
sectors
with
storage
in
there.
Miners
need
to
prove
the
storage
that's
in
there
and
that
you
know
if,
for
whatever
reason
they
have
a
fault
and
and
don't
submit
the
person
at
the
right
time
or
lose
some
of
the
storage,
then
a
fault
will
happen.
Might
as
well
have
a
have
a
predetermined
amount
of
time
to
be
able
to
recover
that
fault.
B
Anything
to
add
tx
or
I'll
jump
into
talking
about
sectors
great,
so
you
know
kind
of
the
the
base
unit
of
storage
in
the
entire
falco
network
is
the
is
the
sector
and
the
sector
is
a
certain
predetermined
amount
of
storage
space
right
now.
It's
32
gigabytes,
32,
gigabytes
and
64
gigabytes
different.
This
is
tied
closely
to
the
proof
of
replication.
B
So
as
the
proof
of
replication
changes,
the
the
sector
size
might
change
as
well,
so
falcon
supports
having
multiples
exercises
because
of
that
at
the
moment
we're
only
using
those
two
right
now.
I
believe
it's
more
efficient
to
use
the
32k
detector,
but
as
the
chain
throughput
gets
saturated,
then
it
will
become
more
potentially
more
valuable
to
use
at
64
gigabyte
sectors.
So
we
might
see
this
this
period
where
at
the
beginning,
is
a
lot
of
32
gigabyte
sectors,
eventually
at
some
point,
transitioning
over
to
64
gigabyte
sectors,.
B
And
here
you
can
sort
of
see
how
how
sector
quality
works,
and
so
one
of
the
you
know,
important
parts
about
a
sector
is
that
a
sector
is
putting
together
either
has
no
deals
in
it.
So
it's
just
capacity,
that's
been
committed
or
might
have
some
some
amount
of
deals
with
with
useful
storage
in
there
and
the
deals
themselves
can
carry
a
multiplier
and
that
multiplier
is
an
important
signal
to
be
able
to
sort
of
increase
the
quality
of
the
deal
and
therefore
increase
the
quality
of
the
sector.
B
So
when
we
think
about
the
the
duration
of
times
that
the
deals
live
for
and
the
quality
of
those
deals
and
so
on,
we
can
use
the
the
the
the
quality
of
the
deals
themselves
to
then
add
up
into
the
quality
of
the
sector,
and
that
way
that
quality
is,
is
what
gives
the
the
kind
of
quality
adjusted
power.
So
you've,
you've,
probably
seen
that
blockchain
explorers
and
so
on.
B
Have
you
know,
rob
eye
power
and
quality
adjusted
power,
and
so
the
the
robot
power
is
just
the
raw
notion
of
you
know:
bytes
of
storage,
but
you
know,
like
capacity
storage,
is
really
useful,
but
it's
not
actually
providing
a
service.
Yet
it's
good
to
have
actual
real
storage
in
there.
Now.
B
There's
this,
you
know
fundamental
problem
between
you
know
granting
an
incentive
for
for
for
adding
storage
where,
if
you
can't
tell
that's
actually
useful,
then
you're,
potentially
incentivizing,
the
storage
of
garbage
and
the
storage
of
like
of
zeros
and
so
on,
and
that's
not
useful
to
the
network
and
that's
actually
really
bad
for
the
network
in
the
in
the
long
term.
B
So
so
this
is,
where
kind
of
like
the
whole
verifier
network
and
so
on,
has
to
have
to
come
in
and
participate,
we'll
touch
on
that
a
little
bit
later
in
the
in
the
conversation,
but
that's
sort
of
where
this
quality
notion
is
coming
from
the
quality
notion
is
coming
from
being
able
to
to
to
look
at
the
different
quality
of
the
deals
and
then
add
that
up
into
a
different
quality
of
the
sector.
B
One
important
thing
here
is
that
this
whole
area
could
be
could
it
could
be
subject
to
all
kinds
of
improvement
and
research
right?
So,
for
example,
if
we
found
a
magical
automated
way
of
checking,
you
know
raw
data
and
check
that
it
was.
You
know,
good
and
useful,
without
having
to
need
something
like
the
peripheral
network,
then
maybe
we
could
have
a
quality
multiplier.
B
That
applies
that
way
directly
to
all
kinds
of
regular
deals,
but
that's
kind
of
like
a
very
hard
open
research
problem
that
it
might
take
years
to
solve,
or
maybe
never
will
be
solved.
There's
maybe
an
impossibility
result,
but
the
point
here
is
this:
this
kind
of
structure
and
framework
enables
the
network
to
to
classify
different
kinds
of
of
qualities
and
and
quality
multipliers
for
different
kinds
of
deals
in
the
long
term.
In
the
short
term,
we
sort
of
expect
only
these.
B
These
three
options
committed
capacity,
regular
deals
and
verified
client
deals,
and
they
have.
You
know
these
three
parameters
over
time.
If
we
can
find
a
way
of
distinguishing
between
a
regular
deal
and
and
garbage,
then
maybe
that
quality
multiplier
can
go
up,
and
maybe
we
don't
need
the
verified
client
deal
or
something
like
that
or
maybe
in
the
future.
We
might
find
that
the
quality
is
actually
really
useful
as
a
mechanism,
and
we
might
end
up
with
many
other
kind
of
categories
of
things
in
there.
So
we'll
see
how
that
that
changes
over
time.
B
But
for
now
what
you?
What
you
should
think
about
is
just
there's
rabbi
power
and
it's
really
useful
to
provide
capacity
to
the
network.
But
it's
much
more
useful
and
valuable
to
provide
useful
storage
and,
to
the
extent
that
we
can,
you
know,
really
enable
all
kinds
of
useful
applications
across
the
world
across
geographies
and
across
use
cases
and
applications
and
so
on.
To
get
verified
client
deal
quality.
B
Then
that
will
be
a
very,
very
useful
way
to
direct
the
falcon
network
to
store
really
valuable
data
for
all
of
humanity
and
so
like
that's
kind
of
like
the
long-term
long-term
goal.
But
you
know
for
a
while,
we'll
we'll
be
fine
with
just
a
massing
capacity
for
a
while.
So
at
the
moment
space
race
has,
you
know:
88
petabytes
of
capacity,
there's
no
useful
storage
in
there
yet,
but
once
we
once
we
kind
of
amass
that
capacity,
also
in
kind
of
a
geographically
redundant
way
and
so
on.
B
We
can
then
start
gathering
the
the
sets
of
clients
and
so
on
that
will
come
in
and
use
all
of
that
capacity.
C
Add
yeah-
and
I
think
part
of
the
coverage
here
that
we
need
to
assign
a
deal
quality
for
a
sector
is
changing.
Blockchain
states
is
pretty
hard
so
that
we
can't
really
keep
track
of
the
dynamic
quality
of
a
sector
as
there's
an
expire
which
we
would
love
to.
So
as
another
pragmatic
trade-off.
Here,
we
assign
quality
to
the
sector
and
make
sure
that
the
seller
quality
doesn't
change
throughout
its
lifetime.
That's
why
we
see
the
weightage
it's
on
the
space
time
occupied
by
the
sector,
not
just
space
or
time.
B
Yeah
yeah
and
one
one
you
know
important
piece
at
the
end
of
the
day-
is
that
miners
are
providing
space
time
right,
you're,
providing
a
certain
amount
of
storage
space
for
a
duration
of
time.
So
the
actual
unit
that
people
should
be
thinking
about
is
providing
that
space,
the
quality
of
the
space
time,
and
so,
as
dealers
expire,
maybe
less
in
a
shorter
time
frame
than
in
the
sector.
Then
the
quality
of
the
sector
might
be
might
lower
over
over
time.
C
B
Exactly
I
went
too
far
great
so
so,
let's
talk
about
stealing
sectors
for
for
a
moment.
So,
as
we
know,
the
proof
of
replication
is
kind
of
an
expensive
computational
operation
over
over
raw
data
and
that's
what
enables
the
the
data
of
a
sector
to
be
able
to
be
sealed
into
into
something
we
can
actually
prove
improving.
B
The
network
that
algorithm
that
we
use
to
go
from
a
raw
unsealed
sector
into
a
sealed
sector
is
the
preferred
replication
and
right
now
we're
using
sdr
stack
depth,
robust
graphs
stock,
stacked
depth,
robust
poor
apps.
In
order
to
be
able
to
to
do
this
over
time.
You
know
the
community
has
also
seen
the
the
nse
poor
app.
That's
another
really
great
candidate
we've
evaluated
many
poor
up
candidates
over
over
the
last
two
three
years.
B
B
You
know
proof
size
into
the
chain
ceiling
and
unsealing
times
and
so
on
and
kind
of,
and
also
make
sure
that
it's
oriented
towards
commodity
hardware
and
not
asics
and
all
that
kind
of
stuff.
So
currently
we
are
in
scr
having
we,
we
hope
to
upgrade
to
things
like
nfc
in
the
future,
and
there
might
be
even
like
we,
we
already
know
of
potentially
many
better
poor
apps.
That
would
be
much
faster
to
operate
in,
and
so
we
expect
all
of
that
kind
of
kind
of
functionality
will
improve
over
time.
B
Now,
of
course,
those
upgrades
will
usually
come
in
the
form
of
kind
of
this
optional
upgrade
where,
as
we
come
up
with
a
new
as
the
network
and
the
whole
community
comes
up
with
new
new
algorithms,
we
can
suggest
them
review
them
analyze
them
security,
audit
them
and
and
test
them
once
we're
confident
it's
a
good
one.
We
can
add
it
into
the
network
through
a
through
a
falcon
improvement
proposal
and
add
it
as
an
option.
B
So
then
some
miners
can
start
storing
with
that
and
if
that
was
useful
and
valuable,
then
more
miners
will
choose
it.
But
so
we
hope
that
when
we,
when
we
add
the
next
upgrade,
you
know,
miners
can
still
choose
whether
to
store
an
sdr
or
sorting
or
nsc
or
something
else.
So
we
think
of
like
that
kind
of
of
an
evolutionary
landscape
or
an
upgrade
landscape,
to
be
the
right.
B
The
right
pathway,
where
the
systems
as
a
whole
will
get
better
and
faster
as
new
and
better
poor
algorithms
come
out,
and
so
you
building
in
kind
of
the
notion
of
being
able
to
upgrade
over
time
without
having
to
to
to
to
to
lose
anything.
There
is
important
right
so
making
those
upgrades
optional
is
important
for
miners,
because
you
want
all
of
your
sdr
storage
to
continue
being
being
useful
and
valuable
through
an
upgrade
like
that.
B
And
maybe
one
one
important
thing
there
is
the
you
can
think
of
the
cost
of
sealing
the
the
the
raw
cost
of
sealing
as
a
big
expenditure
for
the
five
point
network
right.
So
all
of
the
all
of
the
cost
of
the
of
the
production
of
the
whole
network
is
going
into
into
the
cost
of
sealing
so
making
sure
that
the
ceiling
cost
goes
down
over
time
from
from
one
algorithm
to
another
is
really
really
valuable,
because
that
means
that
the
whole
economy
is
paying
less
for
for
that
same
function.
B
So
this
kind
of
research
and
is
extremely
valuable
and
useful
to
to
to
the
whole
network.
C
Yes
hundred
percent,
I
think
that
goes
back
to
the
analogy
of
how
can
we
make
that
conversion
process
and
raw
material
into
exportable
services
are
good,
more
efficient
and
robust,
and
I
think
that
will
go
a
long
way,
both
in
terms
of
the
proof
upgrade
and
many
other
kinds
of
upgrades,
including
consensus
that
will
play
a
role
in
doing
that.
B
Great,
so,
let's
dive
into
the
the
life
cycle
of
a
sector.
So
when
I
left,
when
a
sector
is
added
to
to
the
network-
and
maybe
I'll
use
yeah
this
diagram
here
so
the
very
beginning,
the
sector
starts
by
a
minor
just
unilaterally
packing
some
amount
of
either
just
raw
capacity
or
or
a
set
of
deals,
and
so
on
into
a
sector
and
there's
some
amount
of
of
you
can
think
of
sort
of
the
the
deal
payments
that
the
clients
put
up
as
being
part
of
that
associated
with
that
sector.
B
So
one
that
that
packing
is
done,
then
a
miner
seals
the
sector
into
into
this
into
a
seal
sector.
In
order
to
do
so
has
to
make
some
commitments
into
the
chain
and
then
do
a
proof,
so
there's
a
whole
kind
of
flow
for
for
proving
to
the
network.
B
How
that
detector
was
sealed
correctly
once
the
sector
has
been
has
been
proved
correctly,
then
it
joins
part
of
the
the
large
amount
of
storage
that
the
miner
is
providing
and
that's,
where
a
proof
of
space
time
just
checks
over
time
that
that
all
of
that
storage
is
being
stored
correctly
and
so
miners,
in
kind
of
a
steady
state,
are
proving
a
large
amount
of
storage
all
at
once
of
all
this,
the
seals
all
of
these
steel
sectors.
Now
those
individual
sectors
have
expiration
dates
depending
on
when
they
they
were.
B
They
were
started
and
depending
on
the
lifetime
of
that
sector,
and
so
that
sector
over
time
will
will
potentially
expire,
which
is
the
normal
flow,
and
then,
at
that
point
you
know
after
that,
sector
expires,
then
that
that
all
of
that
is
garbage
collected
and
that
space
can
now
be
used
for
for
new
sectors
or
potentially
asking
for
maybe
terminate
it's
terminated
as
a
case
where
miners
might
lose
the
data
storage,
and
so
there's
a
that
termination
is
is,
is
really
bad
for
the
network,
and
so
that
comes
with
a
with
a
high
high
cost
of
slashing.
B
So
so
the
good
case
is
when,
when
sectors
and
the
lifetime
detector
expires
now,
if
a
miner
has
suffered
faults
or
something
like
that,
then
those
sectors
might
be
terminated
or
potentially,
if,
if
there's
some
downtime
in
a
large
scale,
those
sectors
might
go
into
a
faulty
state
for
a
while
and
then
there's
kind
of
this
recovery
period
to
be
able
to
to
be
able
to
recover
that
storage
before
they're
fully
terminated,
and
so
here
you
know,
there's
a
there's
a
set
of
fees
associated
with
with
these
kinds
of
operations.
B
So
when,
when
there's
some
commitments
put
in
place
to
to
be
able
to
seal
a
sector
and
and
and
then
some
kind
of
clutch
collateral
associated
with
with
having
a
sector
in
the
network,
that
then
enables
you
to
start
earning
block
rewards
which
are
pretty
significant.
B
But
then,
if
there
are
faults,
then
there
are
certain
fees
associated
with
that.
There
are
fees
for
reporting
and
faults
correctly,
and
so
you
know
if
miners
have
a
have
a
just,
an
operational
failure
and
they
have
kind
of
an
honest
mistake
and
kind
of
report,
a
fault
then
there's
kind
of
a
certain
fee
associated
with
that.
B
If,
if
there
is
a
fault
but
there's
no
kind
of
honest
reporting
and
then
the
chain
has
to
catch
it,
then
there's
a
higher
fee
for
higher
fee.
For
that
and
then,
of
course,
if
the,
if
the
sector
ends
up
being
being
being
terminated,
then
it's
a
much
larger
fee
for
for
losing
all
of
the
storage.
At
the
end
of
the
day,
the
whole
falcon
network
should
look
at
sex
determination
as
a
really
bad
event.
B
If
sector
was
terminated,
it
means
that
clients,
data
was
lost
and
if
clients
data
was
lost,
that's
a
really
terrible
situation.
So
imagine
you
know
if
you,
if
you
go
to
the
bank
and
you
store
money
or
you
go
to
any
kind
of
on
your
person,
any
website
that
is
storing
your
photos
or
messages
and
so
on
and
they
lose
the
data.
If
that
service
loses
the
data
or
the
bank
loses
your
money,
it's
like
really
really
bad.
So
we
should
look
at
sector
termination
as
that
kind
of
an
event.
B
Miners
are
agreeing
to
to
store
data
in
the
long
term,
are
earning
a
lot
of
block
rewards
for
doing
so
and,
like
that's,
that's
the
valuable
service
that
the
storage
miners
are
providing.
If
storage
miners
are
not
properly
providing
that
that
service
and
and
fail
for
the
network
and
fail
for
all
the
clients,
then
that
termination
fee
is
that's
why
it's
that
pretty
significant.
C
Yeah,
this
ties
in
nicely
with
the
the
space-time
concept
that
we
brought
up
earlier
when
we
talk
about
sector.
So
when
my
reminders
add
a
sector,
it's
a
commitment
in
terms
of
space
and
time
that
they
need
to
provide
and
maintain
to
the
network
and
the
clients,
and
the
rewards
should
go
exactly
to
this
reliable
and
useful
storage.
And
if
you
terminate
early,
the
the
sector,
storage
is
not
very
reliable
and
it's
not
very
useful.
So
therefore,
the
reward
shouldn't
go
there.
A
B
All
right
so
miners
have
in
general
right
now.
Storage
miners
have
a
few
different
collaterals,
there's
the
there's
the
pledge
collateral
and
then
there's
the
the
storage
steel
collateral,
and
you
can
think
of
these
collaterals.
The
same
way
that
you
should
think
about
proof
of
stake
systems
and
proof
of
stake
systems
with
collaterals.
B
These
are
systems
that
enable
the
right
mechanisms
to
incentivize,
correct
participation
in
the
protocol
in
the
system,
and
so
those
collaterals
are
sized
accordingly
to
provide
good
good
behavior
in
the
whole
system
and
to
achieve
like
the
good,
good,
correct
economic
output.
So,
at
the
end
of
the
day,
you
have
to
look
at
a
large
system
where
parties
are
competing
with
each
other,
potentially.
B
Working
with
each
other
or
against
each
other,
and
so
on,
to
try
and
provide
this
this
global
service.
And
when
that
happens,
you
need
certain
kinds
of
structures
to
make
sure
that
the
economic
activity
is
good
and
and
positive
and
is
helping
the
whole
network
grow.
And
so
that's
where,
where
kind
of
the
sizing
of
these
of
these
collaterals
comes
from
the
initial
platform,
for
example,
is
you
know
early
on?
We
had
a.
B
We
had
a
bit
of
a
different
design
where
there
was
a
much
larger
upfront
collateral
and
we
kind
of
heard
a
lot
of
feedback
from
the
community
coming
saying:
hey
what
if
we
could
use
the
the
block
rewards
as
as
as
part
of
the
collateral
and
so
kind
of
like
the
block,
rewards,
come
in
and
become
part
of
the
collateral,
as
opposed
to
just
having
to
put
up
all
that
up
front,
and
so
we
redesigned
a
lot
of
the
system
in
order
to
satisfy
that.
B
Where
kind
of
like
the
the
the
requirement
and
the
collateral
is
actually
a
lot
smaller
now
and
just
kind
of
covers
like
an
initial
period
and
then
the
block
rewards
will
then
accrue
as
collateral
for
for
some
period
of
time.
That
initial
pledge
has
to
satisfy
a
few
things.
So
when
the
network
is
young
and
small,
the
collateral
has
to
also
help
increase
the
the
attack
costs
into
the
system.
B
So,
at
the
end
of
the
day,
the
the
whole
consensus
protocol
here
relies
on
on
the
storage
power
being
working
correctly
and
not
having
a
51
attack,
and
so
if,
when
the
network
is
kind
of
smaller
and
it's
relatively
cheap
or
easy
to
to
to
you
know
get
enough
storage
to
mount
a
mount
to
a
a
consensus
attack.
Then
we
need
something
like
the
pledge
collateral
to
increase
the
cost
of
that
attack.
B
So
you
can
think
of
this
exactly
the
same
as
as
that's
kind
of
how,
in
proof
of
stake
protocols,
a
a
consensus
attacker
would
then
have
to
buy
up
a
lot
of
the
a
lot
of
the
tokens
to
mount
that
attack
and
then,
and
that
makes
the
attack
extremely
expensive.
You
can
think
of
that.
The
same
way
here
now
that
initial
pledge
collateral
will
will
decrease.
That
amount
will
decrease,
as
the
network
gets
larger.
B
So
if,
if,
if
the
network
gets
really
large,
then
you
know
kind
of
the
contribution
of
each
particular
sector
to
the
whole
network.
It's
actually
much
smaller.
Therefore,
the
the
black
collateral
associated
with
that
sector
goes
down,
so
you
can
think
of
at
the
beginning
in
the
early
days
of
the
network,
because
it's
a
lot
smaller
each
additional
each
sector
has
represents
a
much
larger
percentage
of
the
total
storage.
So
the
blatch
collateral
is
larger.
B
As
the
network
gets
bigger,
then
that
collateral
goes
down,
hopefully
that
that
helps
explain
some
of
the
confusion
around
collateral.
B
Sorry
yeah,
so
so,
then
we
also
have
a
deal
collateral
associated
with
miners
where,
as
part
of
of
entering
into
a
deal,
there's
a
certain
amount
of
collateral
that
needs
to
be
associated
with
that.
Because
again,
if
the,
if
the
miner
is
creating,
we
want
to
disincentivize
a
couple
of
things.
One
is
the
creation
of
fake
deals
and
also
making
sure
that
those
deals
are
correctly
served.
It's
it's
a
very
important
market
mechanism
to
enable
clients
to
be
able
to
sort
minors
by
by
collateral.
B
So
we
we
believe
that
minor
clients
are
going
to
be
selecting
minors
that
offer
larger
deal
collaterals,
because
a
deal
collateral
is
a
statement
from
the
minor
betting
that
they
can
indeed
store
the
data
for
longer
periods
of
time.
So
as
a
client
looking
at
the
market
and
trying
to
hire
storage
providers,
the
client
would
select
miners
that
that
offer
a
much
higher
deal
collateral.
That's
why
this
is
variable
to
enable
a
market
mechanism
to
to
to
find
what
the
right
value
for
the
collateral
is.
C
I
think
one
thing
I
may
want
to
add
here
is:
when
people
talk
about
collateral,
really
it's
incurring
a
capital
cost
like.
Sometimes
we
may
think
of
it
as
like.
It's
an
upfront
commitment,
but
really
that
you
still
get
that
collateral
back
after
the
commitment
is
fulfilled.
So
really
is
the
capital
cost
for
a
small
period
of
time,
and
it
would
be
it'll,
be
a
conceptual
mistake
to
think
that
the
entire
thing
is
the
cause
or,
or
that
say
something
about
the
expectation.
B
One
important
piece
here
is
also
to
think
about
redundancy,
so
the
the
whole
system
make
sure
that
any
one
participant
in
the
network,
any
one
miner
that
is
storing
data-
has
a
very
high.
You
know
cost
to
losing
that
data,
but
clients
themselves
have
to
account
for
the
fact
that
that
might
happen,
and
so
clients
need
to
be
able
to
make
higher
multiple
miners
in
order
to
store
their
data.
And
so
this
could
happen
a
number
of
ways.
B
B
Another
way
is
to
use
erasure
coding
where
you
have
where,
when
you
really
care
about
the
long
term
survival
of
the
of
the
data
you
have
a
a
client
can
take
their
data
erasure
code,
it
into
into
a
set
of
pieces
and
kind
of
break
it
up
into
into
a
lot
of
little
pieces
and
then
sort
each
of
those
pieces
with
different
minors,
and
that
creates
a
level
of
redundancy
and
reliability.
That
is
extremely
strong,
but
it's
very
expensive
to
be
able
to
retrieve
that
and
reconstitute
the
data.
B
So
there's
this
trade-off.
That
is
entirely
available
to
clients
to
explore
between
full,
creating
full
replicas
of
the
data
like
full
individual
copies
of
the
data
and
store
them
with
different
miners,
or
using
erasure
coding
and
breaking
it
up
into
individual
pieces
and
storing
each
of
those
pieces
with
different
different
miners.
So
all
of
that
is
available
available
to
clients.
C
I
think
this
is
a
classic
example
of
the
protocol,
defining
a
very
small
subset
of
things
that
we
want
incentive
and
and
rewards
and
penalties
hold
true
and
everything
else.
We
leave
it
to
the
ecosystem
and
the
market
to
evolve
and
emerge,
and
this
will
be
a
perfect
example
of
like
how
another
marketplace
of
different
features
can
just
be
built.
On
top
of
that,.
B
Another
component
here
is
being
able
to
retrieve
things
quickly,
so
I
covered
earlier
that
for
now
we're
we're
doing
fast,
retrieval
by
by
maintaining
an
additional
copy
over
time.
The
fast
retrieval
will
be
able
to
happen
over
over
just
a
properly
sealed
copy
once
we
get
fast
enough
preserve
replication,
but
there
are
other
kinds
of
retrieval
that
might
be
offered
so
fast.
Retrieval
is
not
just
about
being
able
to
sort
the
data
quickly,
but
also
a
minor
being
close
to
the
close
to
the
to
the
to
the
user.
B
So
that's
where
the
entire
retrieval
market
comes
in,
so
a
storage
provider
can
storage,
miner
might
be
located
in
one
geography
or
they
might
start
participating,
as
well
as
a
retrieval
miner
and
getting
copies
of
the
data
in
specific
locations
to
be
able
to
serve
it
very
quickly.
This
is
the
this
is
the
trade-off
between
long-term
storage
and
content
delivery
networks,
and
so
we
want
to
be
able
to.
B
We
want
miners
in
falcon
to
be
able
to
select
where
they
want
to
be
in
that
continuum,
whether
they
want
to
be
just
storage
miners,
backing
up
data
in
the
long
term
or
being
able
to
serve
it
really
quickly,
closer
and
closer
and
closer
to
the
user
and
participating
in
all
of
the
retrieval
retrieval
side
of
things
we'll
cover
the
result
market
a
little
bit
more
later
great.
So
one
important
component
here
is
the
ability
to
do
offline
data
transfer.
B
B
B
So
one
very
important
part
of
the
whole
system
is
that
it
needs
to
be
able
to
work
with
offline
data
transfer,
so
participants,
clients
are
able
to
hire
a
miner.
Make
a
deal
then
send
the
data
in
disks
in
hard
drives
through
the
mail
and
once
they
arrive,
miners
are
able
to
take
that
data
check
it's
correct
and
then
add
it
to
to
their
sectors
and
clan.
So
it's
a
really
kind
of
of
a
an
unusual
sort
of
requirement
for
an
online
blockchain
protocol.
B
But
when
you're
dealing
with
data
and
we're
dealing
with
storing
large
amounts
of
data,
you
encounter
limits
that
just
the
internet
can't
satisfy.
So
we
need
to
take
the
internet
and
move
it
into
the
mail,
and
so,
like
you,
you
think
of
think
of
this,
as
like
the
first
one
of
the
first
systems,
where
we're
kind
of
integrating,
not
just
the
internet
data
flow.
But
data
flow
over
over
the
packaging
system.
C
B
So
one
of
the
the
the
important
parts
of
all
of
this
is
that
the
the
the
amount
of
data
that
the
whole
network
stores
is
is
kind
of
related
to
the
utility
that
the
economy
provides
right.
So
you
can
think
of
the
falcon
economy
as
having
a
notion
of
utility
or
value
provided
to
the
world
and
that
utility
increases
as
as
more
useful
and
valuable
data
is
stored
by
clients.
B
So
the
whole
network
and
the
whole
economy
really
wants
to
make
sure
that
the
clients
are
have
a
great
user
experience
that
have
a
a
really
useful,
useful
service
and
and
therefore
we
want
a
market
structure
with
where
clients
have
a
lot
of
power
being
able
to
kind
of
select
the
miners
that
they
want
to
deal
with
and
and
they
want
to
prefer
being
able
to
to
get
strong
assurances
about
the
the
quality
of
the
of
the
storage
that
that
they're
getting
being
able
to
count
on
the
data
being
verifiable
and
improvable,
and
so
on
and
and
so
on.
B
B
So
the
whole
life
cycle
of
of
of
clients
is
one
where,
in
order
to
to
you
know,
they
have
initially
some
data,
they
want
to
store
in
the
network
and
they
find
out
about
falcoid,
and
the
first
thing
they
do
is
is
they
want
to
hire
a
miner,
and
so
they
might
start
in
investigating
what
what
potential
minors
they
want
to.
They
want
to
store
with,
and
so
they
have
to
be
able
to
to
make
a.
B
They
have
to
make
a
client
account,
and
so
they
make
an
account
in
the
network.
They
may
get
verified
we'll
talk
about
this
a
little
bit
later,
but
they
may
go
through
this
verification
process
and
then
once
they're
trying
to
store
data,
they
find
the
right
miners
and
they
go
to
the
right,
and
they
might
do
this
through
through
marketplaces
or
dll
aggregators
or
similar.
B
B
Most
of
it
is
really
important
data
that
needs
to
exist,
and
only
a
small
part
of
it
might
need
to
be
retrieved
so
think
of
all
kinds
of
important
archives
and
records,
and
and
and
all
this
kind
of
information
is,
is
stored,
stored,
but
potentially
read
only
once
or
a
few
times
so
so
we
don't
necessarily
need
to
make
all
of
that
data
accessible
super
quickly.
What
we
need
to
make
accessible
really
quickly
is
is
a
much
smaller
amount
of
data
that
that
clients
are
going
to
request
constantly.
B
So
you
know,
a
very
small
amount
of
the
data
on
the
internet
is
request
most
of
the
time,
so
the
large,
the
vast
majority
of
traffic
on
the
internet
is,
is
based
on
a
small
amount
of
the
data
that
that
is
stored
in
the
network,
so
so
that
kind
of
distribution
has
to
be.
It's
part
of
this
whole
picture
where,
at
the
end
of
the
day,
clients
are
storing
their
data
and
sorting
various
different
amounts.
They
might
be
replicating
in
different
areas.
B
They
might
be
asking
they
might
be
hiring
minors
in
different
geographies
to
then
be
able
to
to
get
the
kind
of
jira
replication
they
want
and
then
once
they
have
that
that
in
place
they
might
retrieve
that
data
at
different
points
in
time.
B
You
can
go
into
deal
proposals
in
a
moment,
but
anything
to
add
about
the
lifecycle,
cx.
C
No,
which
is
one
last
bit,
I
think,
like
we
verify
client,
I
think
we're
really
placing
a
lot
of
power
in
the
hands
of
storage
clients
they're,
not
just
going
to
sit
back
and
relax,
they're
actually
going
to
be
an
active
participant
in
the
economy
and
they
and
how
they
participate
interact
with
the
economy.
We
drive
the
product
offering
we
drive
how
the
utility
off
the
network,
so
I
think
it's
super
important
to
be
cognizant
of
that,
and
we
also
come
together
as
a
common
to
put
that
into
good
use.
B
Great,
so
the
deals
that
clients
can
make
you
know
this,
and
these
are
primarily
storage
client
deals.
There
are.
There
are
other
kinds
of
deals
that
that
may
be
for
retrieval
market
and
so
on.
But
here
for
for
for
storage
deals
we're
dealing
with
a
few
parameters.
You
can
think
of
the
the
amount
that
that
of
data
that
the
deal
represents.
You
can
think
of
the
specific
cid
of
the
data
that
the
deal
is
asking
to
store.
B
Then
there's
a
cert
there's
an
amount
of
of
file
coin
associated
with
with
that
deal,
and
then
there's
other
parameters
like
that.
Might
you
know
kind
of
in
this
kind
of
a
certain
amount
of
time
that
the
deal
has
to
has
to
last?
For
so
there's
there's
an
expiration
associated
with
that
deal
over
time.
We
might.
We
might.
B
We
have
thought
about
a
bunch
of
different
kinds
of
of
improvements
to
the
whole
protocol,
including
the
ability
to
extend
deals,
migrate,
deals
from
one
minor
to
another,
enable
miners
to
transfer
deals
themselves
being
able
to
to
upgrade
deals
so
being
able
to
substitute
the
data
that
that
a
deal
has
so
a
client
might
need
to
kind
of
upgrade
that
data
inside
of
a
deal
and
so
on.
So
all
of
that
kind
of
stuff
just
sort
of
improvements
in
the
long
term.
B
For
now
the
deals
are
very
simple:
it's
a
certain
it's
it's!
They
identify
a
specific
amount
of
storage
stored
for
a
specific
amount
of
time
for
a
specific
price,
with
with
a
collateral
associated
with
it.
B
But
this
is
about
making
sure
that
that
that
clients
that
provide
really
useful,
valuable
data
to
the
network
and
increase
the
utility
of
the
network
that
are
going
to
attract
more
clients
that
are
going
to
bring
a
lot
more
business
to
the
whole
system
can
do
so
effectively
by
by
making
sure
that
their
data
is
truly
is
stored
correctly
and
stored
well
by
by
by
the
miners.
B
And
so
this
is
where
the
the
verified,
client
and
quality
amount
comes
in
comes
in
so
verified,
clients
are
give
that
deal
quality
multiplier,
and
so
the
idea
there
is
to
create
a
very
a
very
strong
lever
for
the
clients
that
are
verified
to
make
that
data
extremely
valuable
to
miners.
And
so
we
wanted
the
condition
where,
if
between
you
know,
storing
capacity,
more
capacity
or
a
verified
client
deal,
miners
are
always
going
for
the
verified
client
deal,
and
so
that's
a
very
important
piece
of
the
piece
of
the
puzzle.
B
Now,
of
course,
we
would
love
to
be
able
to
do
it
without
the
verification
process
and
we
would
love
to
be
able
to
do
it
with
just
marking,
marking
deals
but
again,
there's
this
hard
incentive
problem
there,
where,
if
you
just
give
a
reward
for
deals,
then
you're
just
incentivizing,
the
storage
of
garbage
and
and
we
need
to
make
sure
that
that
is
really
actually
useful,
useful
storage,
so
useful
storage,
then
the
if
the
utility
of
the
network
increases
over
time.
B
B
They
help
improve
it.
They
make
it
better
and
they
attract
other
other
clients
and
so
on.
So
think
of
it
also,
as
kind
of
like
a
a
whole
business
development.
Part
of
the
part
of
the
picture,
so
verified
clients
are
really
key
component
of
the
economy
to
attract
a
lot
of
users
and
a
lot
of
clients
to
to
the
network.
C
I
think
a
common
concern
there
will
potentially
relate
to
like
abuse
and
so
on
and
so
forth.
I
think
like
as
long
as
the
process
is
transparent
with
accountability,
and
then
we
start
with
limited
amount
of
trust
and
that
abuse
can
be
greatly
minimized
and
the
net
effect
is
a
very
pragmatic
approach
to
actually
solve
a
very
difficult
technical
problem.
B
Yeah,
exactly
and-
and
so
I
think
you
know
we'll
we'll
we'll
write
more
and
and
talk
more
about
that
whole
very
verifier,
verified
client
component
and
and
we're
working
there
to
to
find
like
a
really
good,
open,
open
structure,
we
really
want
to
make
sure
that
verifiers
are
well
distributed
across
geography.
You
know
in
different
countries.
B
We
want
to
make
sure
that
they
are
in
have
a
lot
of
familiarity
with
different
use
cases,
so
so
different
kinds
of
data
storage,
so
consumer
data
or
business
data
or
scientific
data.
You
know
all
of
that
kind
of
that
kind
of
area
is
well
covered
and
we
want
to
make
sure
that
there's
like
good
redundancy
of
verifiers.
So
so
you
know
if
one
verifier
is
not
working
super
correctly,
people
can
go
to
go
to
another.
B
So
there's
a
whole
kind
of
a
whole
network
associated
with
this,
and
we
we'll
be
writing
more
about
this
process
and
yeah
people
are
interested
in
being
verifiers,
definitely
definitely
reach
out.
As
yet
said
at
the
beginning,
we're
gonna
we're
gonna
start
with
a
smaller
number
and
over
time
kind
of
grow
grow.
That-
and
you
know
it's
one-
a
really
important
function
for
the
entire
network
to
to
to
help
make
sure
that
we're
storing,
really
high
quality
and
useful
and
valuable
data
right.
B
So
if,
if
verifiers
are
not
doing
their
work
correctly
and
are
allowing
garbage
to
be
stored,
then
that
would
be
really
really
bad
for
for
the
economy
long
term,
it
might
seem
like
it's
good
for
minus
the
beginning,
but
it's
really
not
it's
going
to
to
not
be
a
useful
incentive
for
the
long
term.
It's
it's
not
going
to
attract
better
users,
it's
not
going
to
attract
more
business
and
so
on.
B
So
this
is
the
part
where
we
talk
about
sort
of
the
retrieval,
the
retrieval
mining.
So
you
know
one.
B
One
really
important
part
of
the
of
the
whole
network
is
being
able
to
store
and
retrieve
the
data,
so
you
wanna
not
only
store
it
but
serve
it
really
quickly,
and-
and
this
is
where
we
start
looking
at-
not
just
the
storage
services
but
start
looking
at
kind
of
cdn
style
delivery
networks,
and
this
is
where
what
we
really
want
is
is
a
network
that
fluidly
kind
of
adapts
to
the
demand
of
clients
trying
to
retrieve
data.
B
So
we
want
a
very
rich
and
successful
retrieval
market,
where
miners
that
are
serving
data
to
to
users
can
do
so
as
quickly
as
possible
and
and
for
you
know,
cheaper
and
cheaper
prices,
and
so
the
content
delivery
flow
as
it
as
it
works
now
is
that
clients
need
to
find
which
miners
are
storing
the
content
and
and
then
retrieve
the
content
directly
from
that
miner
over
time,
as
we
add
more
and
more
facilities
to
the
retrieval
market
side-
and
this
is
entirely
off
chain,
you
can
think
of
this
as
all
layer.
B
Two
solutions
this
can
happen
in
with
more
and
more
interesting
components,
you
can
think
of
being
able
to
have
gossip
oriented
protocols
that
find
different,
different
pieces
of
the
different
pieces
that
are
that
are
stored
and
advertised
in
the
network
and
then
being
able
to
kind
of
serve
to
the
user
really
quickly.
B
So
we
want
to
to
have
a
whole
ecosystem
here
of
of
different
approaches
and
different
systems
that
can
provide
retrieval
and,
at
the
end
of
the
day,
optimize
the
content
delivery
to
to
work
like
like
really
high
quality
cdns.
There
was
a
there
was
a
retro
market
summit
a
few
a
couple
weeks
ago
and
and
we'll
link
it
in
the
in
the
description
of
this
video.
So
you
can
also
dive
deep
into
where
that's
that
is
at
the
moment
and
and
the
flows
there.
C
And
some
basic
work
and
some
I
want
to
highlight
to
so
to
the
audience
that
some
part
of
this
can
be
done
today
on
the
protocol
and
for
sure
we
can
do
there's
a
little
feature.
That'll
be
available
at
network
launch.
C
Now
we
don't
have
to,
we
don't
have
to
wait
to
like
a
much
like
later
later,
iteration
of
the
protocol,
to
enable
dissolve
that
it's
here
today,
even
though
it
may
not
be
very
smooth
and
so
and
so
forth,
but
like
it's
there
and
it's
possible
and
different,
like
teams
and
companies
can
be
developed
just
on
this
retrieval
market.
B
Itself,
yeah,
so
the
you
know
one
of
the
the
key
this
is
that
is
entirely
layer,
two
right,
so
so
it
doesn't
require
any
hooks
into
into
the
chain
we
already
have.
We
have
all
the
facilities
we
want,
which
are.
We
know
what
the
content
is
and
we
have
payment
channels
exactly.
B
And
and
again
like
here,
we
have
a
diagram
that
is
kind
of
geographically
distributed.
That's
the
kind
of
thing
that
we
want
and
we
need
for
for
a
good
cdn.
We
want
the
data
storage
to
be
well
distributed
around
the
world
and
to
match
the
demand,
and
so,
for
example,
if
there's
really
high
demand
in
one
region,
so
suppose
that
I
don't
know,
maybe
shanghai
suddenly
has
a
large
spike
in
falcon
usage
and
a
lot
of
people
start
storing
and
retrieving
data
in
shanghai.
B
Then
we
want
a
lot
of
the
of
the
of
the.
We
want
the
retail
market
to
kind
of
adapt
to
that
and
suddenly
a
lot
more
virtual
miners
appear
in
that
region
and
provide
and
serve
the
data
quickly.
So
we
want
that
kind
of
adaptive,
cdn
cdn
mechanism
and
so
there's
a
whole.
You
know
range
of
tools
that
can
that
can
help
support
this,
and
it's
entirely.
You
know
layer,
two.
B
We
don't
have
to
do
this
through
to
500
improvement
proposals,
which
means
that
we
expect
a
lot
of
different
teams
to
come
up
with
many
different
strategies.
There's
a
number
of
good
good
projects
and
protocols
that
we're
coming
out
of
hack
fest
that
approach
this
in
a
couple
different
ways:
there's
a
there's
a
two
or
three,
so
it
runs
through
yeah.
Three,
three
different
approaches
to
doing
this
over
gossip
sub
that
people
are
experimenting
with.
So
we
we
expected.
B
This
is
going
to
be
a
a
pretty
rich
area
for
for
a
lot
of
manners
to
participate
in,
but
that'll
come
that'll
come
with
time.
B
Great
so
one
of
the
one
of
the
probably
most
important
improvements
that
we've
made
in
the
last
last
couple
of
years
to
the
to
the
whole
economy
is,
you
know,
a
really
really
important
change
in
in
the
distribution
and
minting
of
the
of
the
of
the
network
of
the
token,
and
so
this
is.
This
relates
to
the
the
block
reward
and
how
the
block
reward
is
is
is
distributed.
B
So
most
blockchains
started
start
with
like
an
exponential,
mending
minting
approach
and
that
exponential
minting
just
comes
from
bitcoin
and
it's
pretty
much
what
everybody
sort
of
started.
Following
now,
a
few
a
few
groups
started
exploring
with
different
things,
so
ethereum
famously
changed
that
and
then
started
doing
a
different
kind
of
minting
with
a
certain
kind
of
fixed
amount,
which
is
kind
of
it
was
kind
of
labeled
as
inflationary
time,
but
it
was
kind
of
deflationary
anyway,
because
it
was
the
the
certain
constant
was
decreasing
over
time.
B
Oh
sorry,
the
consonant
was
the
same,
but
the
proportion
was
decreasing
over
over
a
long
period
of
time.
B
So
as
we
as
we
explored
whole
different
designs
in
the
falcon
economy,
we're
let
we
realize
a
very
important
problem,
which
is
the
the
growth
of
the
network
and
the
and
the
utility
of
the
network
needs
to
which
we're
trying
to
subsidize,
with,
with
a
token
will
come
over
long
periods
of
time.
But
we
not
will
not
all
come
at
the
very
beginning
or
or
said
another
way.
B
What
we
really
need
is
to
be
able
to
tie
the
success
of
the
economy,
to
the
succ
to
the
to
the
minting
rate,
so
you
want
the
minting
rate
to
relate
to
the
success
of
the
economy.
So
this
is
where
the
network
baseline
comes
from.
What
the
network
baseline
is
trying
to
achieve
is
to
say
that
there's
a
there's,
an
ideal
goal
of
the
amount
of
storage
that
the
network
should
be
providing.
B
B
To
work
together
to
reach
that
network
baseline
and
kind
of
pushes
off
the
competition
for
the
block
reward
until
kind
of
passing
that
baseline,
and
so
this
is,
we
think,
is
like
one
of
the
most
probably
most
important
improvements
that
we've
made
to
to
the
to
the
to
the
economic
design
of
the
whole
system,
and
we
think
it's
going
to
it's
going
to
greatly
reward
the
the
miners
that
are
long-term
oriented
that
are
providing
really
high
quality
services,
and
that
are
here
for
for
for
the
long
term.
B
Now,
of
course,
that's
gonna
not
reward
miners
that
are
for
the
short
term
that
are
just
trying
to
like
mine
some
tokens
and
leave
and
not
actually
provide
useful
quality
of
service,
and
so
it's
that
important
mechanism
that
creates
a
difference
between
the
parties
that
are
here
for
the
long
term
and
the
parties
that
are
not.
If
people
are
aligned
with
creating
a
really
useful
and
high
quality
and
high
value
service
in
the
long
term,
then
the
network
baseline
is
going
to
benefit
them
significantly.
C
B
So
you
can
think
of
this,
this
image
as
a
really
good
way
to
describing
different
stages
of
the
economy
at
the
very
beginning.
What
we
want
is
to
build
a
lot
of
capacity.
We
want
to
optimize
the
entire
system.
We
want
to
improve
the
we
want
to
arrive
with
really
strong
and
secure
proofs,
but
it's
okay.
B
If
they're
not
well
optimized
for
retrieval,
we
want
to
have
like
basic
retrieval
in
the
protocol,
and
then
we
want
a
lot
of
the
and
and
so
in
order
to
to
support
all
of
this,
then
a
lot
of
the
the
the
the
revenue
of
minus
is
going
to
come
from
come
from
simple
minting.
As
we
reach
stage
two
of
the
network,
we
want
a
drastically
more
scalable
chain.
We
want
a
greatly
reduced
cost
for
ceiling
and
for
all
operations.
B
We
want
much
more
efficient
proofs
and
retrieval,
and
at
that
point
we're
starting
to
see,
we
want
to
see
kind
of
the
quality
of
service
of
the
data,
storage
and
retrieval
to
be
much
higher,
and
that
means
that
suddenly
deal
payments
become
a
stronger
part
of
the
revenue
for
miners,
and
at
that
point
we
now
should
be
hitting
the
baseline
and
at
that
point
kind
of
the
the
the
revenue
of
miners
are
is
coming
from
baseline,
maintaining
and
supplementing
and
kind
of
as
we,
and
so
you
can
think
of
of
of
the
baseline
minting
there
as
being
a
really
key
part
of
it,
of
achieving
that
high
quality
of
storage
and
so
on.
B
And
you
know
the
level
of
scale
that
we
want
the
level
of
scale
of
storage
that
the
network
wants.
Once
we
reach
stage
three,
then
at
that
point
the
whole
network
can
establish
a
very
strong
kind
of
branded
reputation,
as
a
storage
system
can
have
a
much
larger
suite
of
services
and
products
can
go,
can
be
in
many
markets,
and
so
on
can
have
like
really
high
quality,
cdn
style
retrieval
and
at
that
point
kind
of
like
the
deal.
B
Payments
and
the
quality
service
should
be
excellent
and
be
able
to
compete
with
any
other
storage
network
in
the
world
and
be
ideally
the
best,
and
at
that
point
it's
the
deal.
Payment
is
going
to
to
be
a
large
fraction,
potentially
dominate
the
the
the
revenue
of
miners.
So
so
it's
it's
it's
these
stages
that
the
baseline
is
helping
to
achieve.
The
the
baseline
optimization
is
allows
us
to
start
with
us
with
a
lower
stage
where
we're
mostly
building
capacity
and
we're
not
setting.
We
don't
know
how
much
storage
is.
B
Then
the
the
the
economy
is
thriving
so
well,
and
it's
doing
being
so
successful
that
the
client
payments
flowing
into
the
economy
is
what's
the
main
source
of
revenue.
So
this
is
how
we
translate
these
stages.
You
can
think
of
these
stages
as
the
stages
of
the
economy
stage.
One
is,
is
early,
the
the
economies
of
the
nascent
a
lot
of
people
are
coming
and
there's
a
lot
of
promise
in
the
future,
and
so
we're
kind
of
subsidizing
and
bankrolling
the
entire
effort.
B
The
block
reward
is,
is
paying
for
miners
to
come
and
join
the
join
the
economy.
Then,
in
stage
two
we're
talking
about
optimize,
optimizing
and
scaling
the
system,
a
thriving
economy
that
starts
providing
really
high
quality,
but
it's
not
maybe
not
the
best
in
the
world
yet
and
then
in
stage
three.
We
want
to
be
the
best
in
the
world
and
and
source.
The
vast
majority
of
revenue
coming
in
from
deal
payment
and.
B
Yeah,
and
so
in
the
you
know
this,
the
amount
of
time
here
could
be
is
variable.
It
really
depends
on
the
quality
of
service
and
the
scale
of
the
network,
but
we
think
of
this,
as
taking
many
years,
you
can
think
of
you
know,
think
of
of
a
system
like
bitcoin
or
ethereum.
B
It's
been
many
many
years
in
the
in
the
improvement
and
quality
of
a
service,
so
we
it
may
take
multiple
years
to
to
kind
of
make
it
through
each
of
these
stages,
and
these
stages
are,
are
you
know,
kind
of
entirely
about
the
quality
of
the
service
as
a
whole?.
B
B
At
the
end
of
the
day,
this
is
a
decentralized
network
and
it's
up
to
up
to
us
all
together
as
a
community,
to
make
this
work
to
really
put
in
the
the
work
to
create
something
of
great
value
that
that
builds
a
really
great
network
that
clients
want
to
use
and
want
to
want
to
use
for
their
applications
want
to
use
for
the
storage
of
their
of
their
data
and
so
on.
B
So
think
of
you
know,
if
you
step
back
for
a
moment
and
you
think
of
like
what
are
the
most
successful
blockchain
systems
in
the
world
and
they're
the
ones
that
provide
the
highest
quality
of
service,
you
can
think
of
bitcoin
as
a
as
the
first
one
that
provided
the
ability
to
move
around
money
in
the
world
in
a
very
reliable
and
robust
way.
And
you
know
it's
one
of
the
you
know
the
most
valuable,
the
most
valuable
blockchain
so
far,
ethereum
is
the
second.
B
By
and
and
many
would
argue,
the
whole
economy
on
top
of
ethereum
and
all
of
that
put
together
is
much
more
valuable
than
than
bitcoin.
All
of
that
put
together
is
coming
from
the
high
quality
of
the
utility
of
ethereum
as
an
application
platform,
and
all
these
other
systems
and
economies
that
are
built
around
it.
So
we
want
filecoin
to
be
to
be
an
extremely
valuable
system,
an
extremely
valuable
blockchain
and
that'll
come
from
the
quality
of
service
of
the
whole
system.
B
B
Great,
let's
keep
moving
so
token
allocation,
so
you
know,
I
think,
two
two
really
important
updates
here
that
that
that
are
part
of
the
part
of
the
protocol.
So
one
is
we
we
adjust.
We
clarify
we.
The
amount
of
of
allocation
for
miners
was
around
70
of
the
whole
of
the
whole
allocation
of
tokens
on
falcoin,
and
you
know
we
were
for
a
long
time
discussing
and
debating
how
much
you'd
go
to
storage
miners
for
trivial
miners
and
other
types
of
miners.
B
In
the
long
term,
we
had
a
lot
of
a
lot
of
discussions
for
with
a
lot
of
folks
around
what
is
the
right
amount
in
the
long
term,
and
we
think
that
the
subsidy
of
the
network
is
most
important
earlier
on,
and
it's
most
important
earlier
on
for
for
storage
miners
that
are
providing
the
large
capacity
of
the
network.
So
we
try
to
to
give
as
much
as
as
we
as
much
as
we
could
to
to
storage
miners,
and
you
know,
there's
a
strong
argument.
That
would
say
hey
really.
B
But
we
think
that
if
we
can
create
a
lot
of
value
in
the
next
two
or
three
years
and
make
it
an
extremely
successful
and
high
value
network
that
then
this
this
large
amount
of
the
subsidy
going
to
storage
miners
for
the
service
of
both
maintaining
the
blockchain
and
doing
consensus
plus
storing
all
of
the
data.
Those
two
services
kind
of
are
worth
this,
this
kind
of
minting
rate
and
so
we'll
see.
B
If
that
turns
out
to
be
the
case,
we
this
is
pretty
pretty
aggressive
and
a
lot
and-
and
so
I
think
you
know-
maybe
retrieval
miners
in
the
future
that
are
watching.
This
are
going
to
be
angry
at
me
for
for
for
presenting
a
a
so
generous
thing
for
for
storage
miners
and
not
for
retro
miners.
But
this
is
our
best
guess
today.
We
think
that
this
is
the
right,
the
brighter
location.
B
We
think
that
this
reserves
a
lot
for
for
trail,
miners
and
other
types
of
miners
in
the
future.
We
hope
it's
is
good
enough,
but
at
the
end
of
the
day
this
is
gonna
be
a
you
know.
B
You
know
three
years
from
now
the
network
can
look
at
the
the
distribution
and
the
allocations
of
what's
remaining
on
on
minted
and
and
if
the
community
thinks
that
it
that
it
should
be
apportioned
differently,
then
then
you
know
that
can
be
that
that
is
a
community
decision
in
the
long
term
in
the
future.
B
There's
als
a
whole
set
of
parameters
here
around
different,
so
this
this
table
includes
a
lot
of
oh
yeah.
A
couple
couple
important
details:
one
is
there's
a
lot
of
different
variables
associated
with
token
supplies.
B
One
important
one
is,
you
know:
there's
there's
a
number
of
funds
that
are
that
are
burnt
because
they're
part
of
collaterals
that
need
to
be
slashed
or
they're
part
of
fees
as
part
of
the
the
network,
transaction
fee
and
kind
of
gas
being
burnt,
and
so
on.
That
is
going
into
the
total
burned
funds
and
that's
kind
of
burnt
burnt
forever.
B
The
other
important
detail
is
is
market
cap,
so
one
one
important
thing
to
note
is:
there's
a
lot
of
different
ways
to
try
and
and
estimate
kind
of,
like
the
value
of
of
cryptocurrencies
and
so
on.
We
think
our
position
is
that
most
of
them
are
pretty
flawed.
All
of
them
should
be
taken
with
a
huge
grains
of
salt.
No
one
should
really
kind
of
take
a
look
at
those
numbers
and
just
kind
of
believe
them
out
right.
B
Most
of
them
are,
are
you
know,
very
complex
markets
with
order
books
that
are
not
kind
of
infinitely?
You
know
infinitely
large
and
so
on.
So
you
have
to
be
really
careful
when
you
kind
of
try
and
estimate
the
value
of
the
whole
of
the
whole
network,
we
feel
a
lot
more
comfortable
about
estimating
the
value
of
the
utility
of
the
network,
like
how
useful
is
it,
how
much
value
is
it
giving
to
the
world,
but
in
terms
of
kind
of
thinking
about
a
market
cap
and
so
on?
That's
we.
B
We
want
to
tell
the
community
to
to
exercise
a
lot
of
caution
when
trying
to
estimate
the
value
of
the
network.
That
way,
we
we've
seen
some
some
groups,
using
like
some
extremely
aggressive
and
inflated
numbers
like
like
just
using
the
total
supply
ever
and
saying
that
that's
the
the
market
cap
that
we
think
that
is
totally
wrong
and
totally
inflating
the
value
and
and
really
really
off
and
kind
of
our
it's
kind
of
dangerous
to
do
that.
So
please
don't
do
that.
B
We
think
that
the
most
sober
amount,
the
most
like
kind
of
correct
amount,
is
to
use
the
circulating
supply
and
then
that's
that's
like
the
right,
the
right
market
cap
size,
but
that's
the
suggestion.
We
are
not
we're,
not
the
the
best.
We
we
don't
make
decisions
for
other
people.
B
People
should
actually
caution
with
how
they
make
their
calculations,
but
we
think
that
the
right,
the
most
the
most
accurate
measure
is
to
is
to
kind
of
use
the
circulating
supply
and
in
in
general,
we
think
that
the
the
whole
exercise
and
effort
of
valuing
valuing
networks
and
and
so
on,
is
is
in
its
infancy,
and
we
think
a
lot
of
that
will
will
grow
and
improve
over
time,
and
people
will
make
make
much
better
guesses
as
to
as
to
how
to
properly
value
these
things
and,
and
so
on.
B
That
kind
of
section
7
is
about
a
whole
bunch
of
other
details
that
are
part
of
the
part
of
the
part
of
the
network,
part
of
the
part
of
the
protocol
in
in
and
a
bunch
of
different
mechanisms,
and
so
these
this
are
kind
of
the
initial
network
parameters
that
that
that
we
recommend
these
are
not
final,
we're
still
refining
some
of
them,
and
so
they
will
likely
have
an
update
to
these
coming
in
in
the
next
weeks,
and
we
also
expect
the
community
as
a
whole
to
tune
these
over
time.
B
So
we
we
expect
that
this
will
be
we'll
have
some
strong
initial
parameters,
but
over
time,
as
the
network
upgrades
as
many
new
features
come
online
as
the
chain
becomes
more
scalable
as
we
get
new
personal
replication,
these
parameters
will
change
and
adjust
with
time,
and
that
is
entirely
a
community
governance
question
similar
to
to
kind
of
ethereum
improvement
proposals.
Faculty
improvement
proposals
will
kind
of
drive
adjustments
to
these
parameters
and
we
expect
a
lot
of
the
community
to
participate
in
making
those
really
difficult,
difficult,
long-term
decisions.
C
Nope
sounds
good,
maybe
it's
time
for
us
to
do
some
faqs
or.
B
Yeah
yeah
sounds
good,
so
let's
move
from
to
your.
C
Screen
well,
thank
you
for
watching
us
through
the
report.
I
hope
audiences
understand
better,
like
the
intention
and
and
why
we
design
things
a
certain
way,
and
I
think
I
want
to
just
reiterate
on
the
the
goal
of
the
system
and
one
of
the
many
mental
models
that
we
use
to
think
about
the
falcon
economy.
C
We
are
all
here
together
as
a
common
asset
community
to
make
to
produce
valuable
storage
related
goods
and
services
and
export
that
to
the
outside
world,
and
we
have
a
list
of
like
common
questions
about
the
report
that
we
collected
in
the
last
week,
and
I
will
be
happy
to
address
all
these
some
of
these
like
one
by
one
here
so
for
the
first
one.
How
does
baseline
maintain?
I
think
what
they're
asking
is
exactly
work.
C
There
are
different
interpretations
around
how
they
work,
so
I
hope
to
hope
we
can
provide
greater
details
here.
So
taking
a
step
back,
we
I
think
we
went
through
the
goal
just
now,
with
juan
the
minting
rate
should
really
tie
to
the
rate
of
inc.
C
The
network
utility
rate
in
terms
of
how
valuable
the
network
is
in
providing
the
goods
and
services,
and
we
really
want
to
incentivize
and
reward
storage,
not
just
ceiling,
speed
and
hardware,
even
though
they
are
important,
but
maybe
not
in
not
more
important
in
an
exponential
way,
and
we
also
want
to
provide
certain
counter
pressure
to
shocks
or
in
the
event
that
the
shotgun
happened.
C
There's
this
anti-fragile
element
in
the
system
to
bring
people
back
and
remove
back
and
and
and
contribute
back
to
the
and
ensure
the
the
running
of
the
network,
and
therefore
we
propose
this
a
hybrid
minting
structure
in
terms
of
the
falcon
storage
mining
allocation
with
30
in
simple
exponential
decay
and
the
remaining
70
in
baseline
minting.
C
And
then
we
talk
about
what
the
baseline
is,
which
is
some
kind
of
a
network
network
level.
Kpi
which
is
like
this
is
an
ideal
case.
How
fast
network
will
grow
starting
at
what
initial
value?
And
then
we
define
a
progress
of
how
all
of
us
as
a
community
in
achieving
this
kpi,
which,
which
is
what
we
call
the
network
time,
and
this
is
in
contrast
with
in
simple
x
and
simple
exponential
decay.
C
We
have
this
notion
of
the
simple
time
or
the
wall
clock
time
that
would
always
advance,
and
then
we
mean
rewards
based
on
the
progress
of
the
network,
how
far
the
network
has
achieved
in
terms
of
the
space
time,
in
terms
of
the
space
time
that
we
expect
and
as
I
mentioned
earlier,
if
the
baseline
is
consistently
met,
then
the
minting
would
be
exactly
the
same
with
the
original,
simple
exponential
decay.
Otherwise
some
early
rewards
are
delayed
and
the
minting
rate
will
be
spread
out
into
the
future.
C
So
on
the
right
here
we
have
the
mathematical
specification,
and
here
is
just
an
example
of
what
the
improved
meeting
rate
would
be.
So,
as
you
can
see,
the
initial
meeting
rate
is
still
high,
so
sizable,
but
we
actually
delay
some
of
this
mean
rate
towards
the
future.
When,
let's
say,
there's
new,
there's
no
improvement
to
the
protocol
and
now
we
can
unlock
greater
potential,
we
can
actually
unlock
a
greater
proportion
of
the
kpi
in
a
later
date
and
then
then
we
provide
greater
incentive.
One.
B
Yeah
one
one
important
thing
here
like
that
that
example,
improvement
rate
is,
is
again
just
one
possible
example.
That
depends
entirely
on
the
amount
of
storage
in
the
network.
So
if
we
hit
the
baseline,
if
we
hit
the
baseline
amount,
then
it
goes
back
to
the
original
mint
rate
right.
So
so,
if
we
hit
the
baseline,
then
we
go
back
to
the
mint
rate
if
we
are
lower
than
the
baseline
that
we're
deferring
some
of
that
that
currency,
just
to
subsidize
in
the
future.
C
Yes-
and
this
is
just
one
example
and
and
how
you
actually
and
the
exact
realization
is
still
in
the
hands
of
the
community,.
B
Yeah
sorry,
one
important
detail
here
is
like
this.
This
baseline
construction
really
gives
a
very
strong
that
strong
sense
of
like
deferring
the
strong
computation
until
past
the
past.
The
baseline
is
an
important
feature
to
make
sure
that
miners
in
the
early
days,
aren't
kind
of
like
just
measured
in
terms
of
their
results
against
each
other,
but
they're,
measured
also
in
the
results
against
against
the
baseline.
C
Right
in
in
the
way
that
bitcoin
other
networks
work,
but
they
also
compare
each
other
when
we
are
below
the
baseline
as
well.
So
it's
like
a
complete
and
collaborative
relationship
of
all
the
miners,
yeah
and
one
thing
to
call
out
here
too.
I
think
there
is
a
very
subtle
incentive
for
sector
of
longer
promise
duration,
because
something
that
we
realize
is
as
new
sectors
get
added,
some
old
sector
might
be
expiring.
So
it's
actually
the
best
interest
for
minor
to
commit
sectors
of
longer
lifetime.
C
Cool
next
is
how
is
initial
pledge
exactly
calculated,
so
I
think
so,
let's
revisit
revisit
all
the
goals
and
constraints.
It's
not
meant
to
be
a
barrier
of
entry
to
entry
for
small
miners.
It
must
be
small
enough
to
be
feasible
for
miners
to
join
the
network,
or
it's
easy
to
reason
about
whether
it
they
should
join
the
network,
and
it
must
be
large
enough
to
secure
the
network
and
and
provide
an
incentive
to
improve
the
quality
of
service.
So,
therefore,
there
are
two
components
that
one
touched
on
earlier.
C
One
is
the
initial
storage
plan,
which
is
based
on
the
expected
reward
that
the
sector
will
earn
in
the
future,
and
this
provides
some
initial
collateral
for
for
slashing
against,
like
penalties
when
the
sector
before
the
sector
earns
any
reward
and
there's
also
the
the
component,
which
is
called
the
initial
consensus
pledge.
C
So
the
initial
storage
pledge
is
based
on
the
block
reward
and,
as
boiler
goes
down
over
time
as
a
general
trend,
it
may
not
be
enough
to
provide
a
sufficient
guarantee
to
cause
a
consensus
attack
or
to
consensual
attacks.
At
the
same
time,
it's
also
not
proportional
to
the
storage
that
you
add
so
in
an
absolute
term,
in
a
sense
that
if
let's
say
the
network
has
a
lot
of
storage
showing
up
the
total
pledge
doesn't
change
when
the
network
has
much
smaller
amount
of
storage
showing
up.
C
So
we
want
it
to
be
proportional
to
the
storage
committed
and
we
also
want
to
make
a
consensus
attack
costly.
So
therefore,
these
two
components
they
are
combined
in
this
particular
way
and
the
initial
recommendation
for
initial
storage
pledge
is
an
estimated
20
days
worth
of
blood
reward
that
will
be
earned
by
the
sector.
So
if
the
sector
has
to
be
around
180
days,
it
feels
like
20
days
of
bloody
war.
Initial
pledge
seems
to
be
a
reasonable
number
in
terms
in
terms
of
the
token
and
then
the
initial
consensus
pledge.
C
I
understand,
there's
a
confusion
here
about
the
30
percent
and
I
think
what
I
mentioned
earlier
is
it's
a
fraction
that
will
help
the
whole
network
achieve
that.
Try
to
achieve
the
target
when
we
are
at
the
baseline
or
above
the
baseline,
but
in
reality
for
a
particular
sector.
It
should
just
be
a
very
small
amount
of
alcohol
and
and
that
specific
relationship
is
defined
in
the
equation
at
the
bottom.
B
Yeah,
I
think
yeah-
that's
really
well
explained,
I
think
maybe
one
one
other
component
in
the
long
term
kind
of
like
maybe
years
years
in
the
long
term,
people
may
want
to
revisit
these
these
calculations
and
that
you
might
find
other
as
as
we
reach
much
higher
scalability
and
throughput
in
the
chain,
as
we
do,
sharding
and
and
regions
and
all
that
kind
of
stuff.
We
might
end
up
in
a
world
where
these
equations
need
to
be
revisited.
B
B
So
we
would
expect,
in
the
long
term,
say
five
five
years
out
five
ten
years
out
to
to
see
changes
here
from
the
community,
and
you
know
that
that
is
the
kind
of
thing
that
that
over
time
will
as
a
whole
group
we'll
see
what
is
the
the
highest
likelihood
likelihood
of
success
for
for
the
network
as
a
whole.
B
At
the
end
of
the
day,
it's
about
making
sure
the
network
is
secure,
that
we
don't
have
any
kind
of
consensus
problem
and
that
the
the
the
the
storage
is
well
is
well
kept
for
the
entire
duration
and
that's
not
at
all
going
to
be
going
to
be
lead
to
bad
incentives
and
bad
incentive
structures.
And
we
also
want
to
make
sure
that
there's.
B
C
And
exactly
so,
I
think
it
has
become
a
common
theme
in
our
presentation
today,
like
it's,
a
it's,
an
evolving,
it's
involving
system,
and
many
of
we
have
new
discovery
and
new
findings
along
the
way,
and
I
think,
as
long
as
it's
a
better
analysis,
they
also
respect
the
same
constraint
and
making
sure
all
the
subsystems
work
together.
We're
happy
to
look
at.
We
can
review
the
proposal
together
as
a
community
and
make
decisions
from
there.
C
The
next
question,
which
also
happened
to
be
asked
in
the
chat
room,
is
what
is
a
sector
quality,
adjusted
power
or
sector
qap
and
appears
like
as
motivated
earlier.
We
want
to
reward
sectors
of
different
content
slightly
differently
depending
on
what's
inside
the
sector,
and
we
also
have
the
construct
constraint
that
we
can't
really
change
this.
We
can't
really
change
it
through
what
percentage
all
the
time
a
sector
expires,
come
and
go,
there
will
be
a
lot
of
changes
to
the
to
the
blockchain
state.
C
So
therefore,
we
invented
this
notion
of
a
sector
quality
and
also
the
multiplier
to
reward
different
sector
sectors
of
different
content
differently
and
the
deal
weight
here
is
basically
the
space
time
occupied
by
a
geotype
in
a
sector
and
a
sector
is
defined
by
by
its
promise,
duration
and
its
size.
And
then
basically,
it
would
be
the
deal
quality
multiplier
of
the
dual
types
in
a
sector
weighted
by
the
space
time.
So
we
have
like
two
examples
here
of
how
the
calculation
is
broken
down.
C
So
the
first
one
when
we
have
like
half
of
the
sector
is
filled
with
verified
client
deals,
and
so
we
have
16
multiplied
by
10
and
then
16,
which
is
the
committed
capacity
deal
type
divided
by
the
space
time
of
the
of
the
sector,
which
will
give
us
5.5
for
the
sector,
quality
and
and
there's
another
example
down
here
and
this
sector
quality
which
convert
get
converted
into
sector
quality.
Adjusted
power
is
also
the
consensus
power
in
the
in
the
system.
C
Great
next,
let's
see
this
is
a
very
common
question.
How
can
I
become
a
verifier
or
verify
client
so
right
now,
I
think
the
process
is
still
being
issued
under
development,
but
we
can
talk
briefly
about
the
principles
and
I
think
juan
mentioned
saw
that
earlier.
The
verification
process
should
be
easy
and
plentiful
and
the
verified
clients
are
intended
to
improve
the
utility
and
product
offering
off
the
network.
C
It's
it's.
They
play
a
very,
they
have
a
much
stronger
level
and
they
also
play
a
very
important
role,
especially
in
early
days
in
shaping
what
the
network
has
to
offer
and
what
the
network
is
about,
and
it
should
also
make
the
network
more
decentralized
by
adding
by
adding
a
layer
of
social
trust
on
a
purely
resource-driven
network
and,
as
mentioned
earlier,
with
limited
trust,
transparency
and
countability
ability.
Should
we
can
minimize
the
views
of
this
mechanism.
B
I'll
add
here
that
yeah,
this
is
one
of
the
one
of
the
things
that
we
need
to
to
describe
and
write
a
lot
more
about,
and
a
lot
of
this
program
is
still
still
in
development.
The
principles
of
cx
are
are
totally
spot-on
and
you
know
some
of
the
things
that
that
that
we're
thinking
about
and
we're
thinking
about
this,
not
just
in
the
falcon
team,
but
with
a
lot
of
other
all
other
participants
that
are
starting
to
get
involved
in
in
shaping
shaping
the
network
and
shipping.
B
The
the
ecosystem
is
to
really
make
sure
that
there
is
good
geographical
diversity.
B
Good
use
case
diversity
that
we're
looking
at
many
different
kinds
of
application,
use
cases
and
use
case
domains
that
we
have
multiple
organizations
that
we
have
organizations
tuned
to
different
sizes,
verifiers
that
verify
for
small
amounts
of
data
like
say
a
gigabyte
or
two
gigabytes,
or
something
look
very
different
from
verifiers
that
verify
for
petabytes,
and
so
we
all
of
those
are
are
factored
in
at
the
end
of
the
day.
B
All
of
this
is,
is:
is
a
community
oriented
governance
process
similar
to
the
falcon
improvement
proposals,
so
we're
going
to
end
up
with
some
mechanism
where
there's
kind
of
a
community
oriented
application
process
and
a
community
oriented
review
and
and
and
and
approval
of
verifiers.
So
you
can
think
of
of
that,
as
kind
of
like
some
of
the
principles
over
time.
We
we,
we
think
we're
going
to
start
with
a
network
in
kind
of
like
a
smaller
size
and
over
time
grow.
So
we
we.
B
We
don't
want
to
immediately
jump
to
having,
like
you,
know,
50
verifiers,
or
something
like
that,
because
that
we
don't
yet
know
for
sure
how
the
entire
system
is
going
to
work.
So
we
think,
starting
with
something
like
10
or
maybe
15,
or
something
like
that
is,
is
the
right
amount
early
on
and
over
time
scale
it
out
and
grow
it
as
we
understand
how
the
shape
of
the
systems
have
to
work.
B
So
we
might
go
for
say
at
the
beginning
kind
of
geographic
diversity,
maybe
one
or
two
verifiers
in
different
in
various
continents
and
then
and
then
over
time,
scale
it
out
to
more
and
more
more
parties.
But
it
really
depends
on
on
how
these
things
work,
on
how
the
the
the
network
operates
and
and
so
on.
But
but
again
we
don't.
We
don't
need
this
to
be
in
effect
from
day
one
we
this
can.
This
can
land
and
develop
over
time.
C
Next
so
we
mentioned
a
bit
about
the
daily
blood
reward
and
use
that
in
to
compute
calculate
fees
and
also
initial
pledge.
So
then
the
question
is
how
exactly
is
estimated.
So
the
detail
here
is
the
blood
reward
per
day
is
for.
The
first
principle
is
the
like,
liquid
of
a
for
likelihood
for
a
sector
mining,
a
block
and
multiplied
by
the
size
of
the
block
reward
and
the
likelihood
of
mining.
C
A
block
is
dependent
on
its
it's
weighted
by
so
it's
rated
by
the
network,
quality,
adjusted
power
and
and
both
in
terms
of
its
what
it
is
and
and
the
rate
of
which
that
is
changing
and
the
size
of
blob
reward
is
also
recorded
on
chain
which
is
never
reward
for
epoch,
and
we
use
both
of
these
values
are
smooth
on
chain
and
then
we're
making
extrapolation
to
estimate
how
much
a
sector
would
earn
in
a
day,
and
that
would
give
us
a
much
more
accurate
estimation
than
some
of
the
earlier
iterations.
C
This
is
in
a
similar
calculation
to
the
initial
consensus
pledge.
So
the
reason
why
we
have
this
in
the
first
place,
if
we
want
to
provide
some
minimum
level
of
sla
for
storage
clients
when
entering
into
deals
and
also
like
cloud,
also
reflect
the
fact
that
deals
consume
on-chain
resources,
and
this
would
be
an
extra
additional
cause
to
self
dealing
and
and
just
want
to
call
out.
This
is
still
specific.
So
it's
not
it's
separate
from
pledge
which
is
per
sector
and
miners
don't
really
have
to.
C
Even
though
I
don't,
we
don't
encourage
miners
not
taking
deals
but
like
this
is
not
incurred
when
miners
are
not
engaging
in
storage
deals
and
the
calculation
is
similar
to
what
we
suggest
the
equation
before
five
percent
of
the
circulating
supply
and
then
normalized
by
the
narrow
baseline,
which
will
amount
to
a
pretty
small
amount
per
deal
per
sector.
C
Next,
how
does
bloody
well
vesting
work,
so
we
use
the
blood
reward
to
reduce
us
collateral
to
reduce
the
initial
pledge
requirement,
even
though
there
is
still
some
initial
pledge
and
the
reward
earned
by
the
sector
is
to
subject
to
slashing
until
its
promise
is
fulfilled,
which
is
in
terms
of
both
the
space
and
the
time
and
reward
vesting.
C
It's
a
self-linear
to
provide
this
strong
incentive
for
minors
to
maintain
their
to
maintain
their
promise
towards
the
end
of
the
sector's
lifetime,
and
one
of
the
many
proposals
that
we
consider
and
the
one
they're
recommending
at
the
moment
is:
we
have
a
reward
delay
and
then
the
linear.
So
when
the
ball
is
mine,
the
blow
reward
is
armed.
C
C
C
Well,
the
short
answer
is
a
minor
need
to
acquire
tokens
for
initial
pledge.
It
can
come
from
many
places,
exchanges
landing
space
race
and
there
are
many
use
cases
that
are
being
developed
about
landing
at
the
moment,
and
I
just
want
to
call.
I
think,
a
lot
of
this
anxiety
that
we
sometimes
may
have
is
with
the
network
initialization,
and
I
think
it's
a
separate
problem
from
the
design
principles
and
the
rationale-
and
we
also
also
other
work
stream
to
explore
ways
to
make
the
initial
onboarding
smoother
for
storage
miners.
B
I
should
add
here
yeah
a
lot
of
couple
things
here.
One
of
the
things
we're
exploring
right
now
is
is
how
to
use
the
the
the
sectors
and
the
storage
that
miners
have
already
added
to
space
race
in
the
mainnet.
So
now
we
can't
guarantee
that
we're
going
to
be
able
to
do
that,
but
that's
something
that
we
are
exploring
so
an
important
thing
and
take
away
from
miners
right
now
is
hey.
B
B
So
hang
on
to
those
sectors,
and
that
might
be
that
might
help
solve
this
initialization
problem
or
we
can
start
with
a
pretty
sizable
network
and
then
miners,
don't
you
know,
could
could
just
kind
of
sit
there
at
that
amount
and
then
just
you
know
you
don't
have
to
add
more
until
we
start
getting
some
of
the
block,
rewards
and
whatnot.
So
that's
also
available
to
you.
You
can
just
wait.
B
Another
option
is
like
you:
don't
have
to
buy
anything,
you
can
just
wait,
and
so
that's
one
of
the
the
things
that
we're
hoping
to
to
add
in.
So
this
is
a
really
important
question
we're
very
sensitive
to
it.
We
want
to
make
sure
that
the
entire
economic
flow
of
the
system
works.
B
Part
of
the
space
race
goal
was
to
enable
the
initialization
of
the
network
and
being
able
to
provide
some
initial
initial
file
coin
to
to
miners,
to
be
able
to
actually
use
in
in
running
the
network,
and
so
yeah
look
out
for
these
options
and
then
yeah
at
the
end
of
the
day,
lending
exchanges
and
so
on.
We
think
are
going
to
be
a
big
part
of
this
once
five
coin
and
ethereum
bridges
work
better.
C
Great,
the
next
question
is:
what
happens
if
price
goes
down
after
a
miner's
purchase
token
for
initial
flash,
which
is
a
question
that
we
also
get
pretty.
We
also
see
a
lot
in
the
chat
room,
so
rather
we
don't
control
price
or
can't
predict
it,
and
prices
are
result
of
collective
belief
and
action
of
all
participants
in
in
the
economy
in
the
market,
and
it's
also
meaningless
to
consider
prices
without
looking
at
the
time
frame
that
you're
talking
about.
C
Are
you
talking
the
next
day,
three
months
or
six
months
out,
and
it's
just
not
a
very
meaningful
discussion
and
all
participants
in
the
economy
will
make
that
decision
based
on
their
own
horizon
and
their
own
expectations,
which
is
what
I
mentioned.
B
Earlier
you
could
yeah.
I
think
I
think
one
important
thing
here
is
like
all
of
these
economies
and
markets
are
large,
complex
systems
with
many
participants.
The
best
advice
is
like:
if
you're,
if
you're,
not
sure
about
your
own
things,
don't
participate
just
hold
back
just
wait.
Let
let
wait
until
until
the
network
advances
and
things
stabilize.
B
We,
we
wanted
the
entire
network
to
exercise
a
lot
of
caution
and
be
careful
about
about
the
participation,
we're
finding
we're
trying
to
create
good
and
useful
ways
for
for
parties
to
enter
the
network
through
through
yeah
through
things
like
space,
race
and
so
on.
But
we
also
know
that
there's
a
lot
of
miners
and
a
lot
of
participants
that
are
that
have
formed
companies
and
are
building
organizations
that
are
waiting
for
falcon
to
to
to
be
live.
B
That
are
burning
a
lot
of
money
right
now,
and
so
we
we
know
all
of
those
folks
want
the
network
to
arrive
sooner
and
all
that
kind
of
stuff.
All
that
kind
of
thing
at
the
end
of
the
day,
exercise
caution,
make
your
own
best
decisions,
be
careful
and
just
wait
if
you're
uncertain
about
anything,
just
just
wait
and
don't
don't
don't
make
any
any
rash.
C
All
right
next
is
a
bit
more
back
to
the
technical,
which
is
what
will
trigger
slashing
and
how
does
it
work?
So
there
are
many
different
kinds
of
slashing,
but
I
just
want
to
focus
on
those
related
to
the
storage
reliability.
C
So
when
the
sector
is
in
the
40
states,
a
miner
pays
a
sector
fall
fee
for
every
day
that
the
sector
is
is
in
that
state
and
the
direction
of
setting
the
number
is
once
reliability
exceeds
a
reasonable
threshold.
The
risk
posed
by
these
fees
should
be
actually
decreased
pretty
rapidly.
So
if
your
storage
is
liable,
this
shouldn't
be
something
there
will
be
a
huge
concern
and
the
second
fee
would
be
so.
We
want
to
encourage
incentivized
matter
for
reporting
their
thoughts
as
early
as
they
can
and
be.
C
So
if
a
minor
does
not
announce
the
faulty
state
and
instead
the
network
detects
that
and
then
there
will
be
a
separate
fee
in
in
case
of
potential
malicious
or
dishonest
behavior,
given
that
it
is
still
a
permissionless
network,
and
if
the
sector
is
in
the
faulty
state
for
too
long,
then
the
sector
is
considered
terminated
and
the
miner
may
have
to
pay
a
termination
fee.
C
C
Topics:
next
storage,
storage,
clients,
compensated
for
in
the
event
of
data
loss
or,
and
how
will
token
volatility
impact
storage,
clients
so
search
clients
are
not
directly
compensated
for
each
loss
file,
even
though
the
unpaid
eu
payment
will
go
back
to
the
client.
Part
of
the
reason
here
is:
we
are
in
a
semi-super
anonymous,
pseudo-anonymous
network
and
the
protocol
has
no
data
a
way
of
telling
whether
distinguishing
between
like
clients
and
miners.
C
And
so
they
want,
given
all
the
slashing
and
penalty,
what
happens
to
the
burn
power
point
right
now
when
they're
burned,
they
descend
to
an
unstandable
and
standard
unsuspendable
address,
and
this
reduces
the
circulating
supply
of
the
token.
It's
also
possible
that
in
the
future,
when
a
community
decide
to
reallocate
these
tokens
through
a
falcon
improvement
proposal,
but
that
will
be
that
will
be
up
to
the
community
as
well.
B
Or
yeah,
I
would
say,
yeah
no.
I
think
we
should
count
on
all
these
all
these
tokens
being
burned
forever
and
in
the
long
term
it
is
possible
like,
like
anything
in
the
protocol.
It
is
possible
in
the
long
term
people
think
people
decide.
B
The
community
decides
that
they
should
be
used,
but
we
expect
that
that
won't
be
the
case
that
that
they
will
just
be
burned,
burnt
forever
and-
and
so
I
think
at
least
kind
of
in
the
next
three
three
to
five
years,
and
so
on
that
we
should
assume.
That's
the
case.
That,
however,
again
like
none
of
us
control
the
network
right
as
soon
as
we
release
the
network,
it's
a
governance
process
that
it
that
is,
network
oriented,
and
so
the
community
decides
to
do
something
like
that's
the
community's
decision.
B
C
B
Could
you
refresh
your
your
your
browser?
I
think
I
added
another
another
question
in
there
that
came
in
let's.
C
B
Yeah,
so
what
uptime
and
slas
are
expected
for
minors?
And
so
slas
are
you
know,
service
level
agreements?
What
kind
of
kind
of
op
time
is
required?
B
B
Larger
miners
have
to
be
much
more
reliable,
especially
for
consensus.
Consensus
is
the
really
important
one
now
you
you
should
consider
that
being
a
storage
miner
on
file
coin
is
running
and
participating
in
a
cloud
storage
service.
That
means
it
should
be
in
operation
all
the
time
we
want
to
achieve
reliability
to
be
able
to
compete
with
other
clouds
and
so
think
about,
like
you
know,
five
nights
over
five
or
six
nights
of
reliability.
That's
really
hard
to
get
to
do.
B
B
Now
this
is
especially
critical
for
consensus
power.
So
if
you're,
if
you
are
participating
in
one
of
the
you
know
the
top
20
or
top
30
miners,
that
means
you
have
a
significant
fraction
of
the
consensus
power.
That
means
that
you
need
to
participate
actively
and
be
ready
to
to
to
take
action
if
the
chain
starts
doing
something
something
problematic.
B
Blockchains
are
living
systems
that
are
under
attack
all
the
time,
so
you
can
think
of
bitcoin
and
ethereum
and
many
other
chains
as
being
under
attack
by
a
lot
of
participants
all
the
time,
so
think
of
like
all
the
different
hacks
that
people
have
tried
or
dos
attacks
and
so
on.
So
this
means
that
storage
miners
that
have
a
significant
amount
of
power
need
to
be
part
of
need
to
be
able
to
respond
to
emergencies,
so
problems
can
and
will
happen
and
miners
need
to
be
alert
to
fix
those
problems.
B
So
this
means,
like
larger
facilities
and
larger
operations,
will
have
many
people
involved
and
they'll
be
kind
of
like
on-call
rotations.
It's
really
important
that
certain
kinds
of
attacks
on
the
network
require
responses
in
very
short
time,
meaning
one
or
two
hours.
We've
we've
seen
responses
in
bitcoin
and
ethereum
that
require
that
kind
of
speeder
response.
It's
very
rare.
We
hope
that
this
is
a
very
rare
thing,
but
it's
it's
we
do.
B
We
do
need
a
significant
fraction
of
the
consensus
power
being
able
to
interact
in
that
and
and
and
and
and
operate
at
that
at
that
speed,
so
we're
working
to
provide
some
some
channels
for
communication
for
storage
miners
that
that
represent
a
large
amount
of
the
power
and
so
on
to
be
able
to
coordinate
an
event
of
these
kinds
of
problems,
and
so
we'll
have
more
to
say
there.
B
But
it's
really
important
that
that
source
miners
really
consider
the
the
uptime
guarantees
that
are
required
to
satisfy
the
economic
requirements
of
the
network.
If
parties
can't,
if
we
can't
provide
that
as
a
whole
network,
then
then
you
know
slashing
happens.
Then
the
network
stops
and
the
block,
roars
and
aren't
getting
released.
All
kinds
of
economic
activity
doesn't
happen.
This
is
really
critical
of
the
network.
The
network
continues
right.
C
And
just
to
just
remind
ourselves
again,
I
think
we
said
that
as
well.
I
think.
Currently,
we
don't
assume
that
in
the
penalty
rules,
it's
a
much
more
reasonable
and
relaxed
a
much
more
relaxed
storage,
reliability.
C
Requirement
cool
next:
when
will
final
parameters
be
released,
so
I
think
it
has.
It
should
be
clear
by
now
there
will
be.
I
mean
there
isn't
really
a
final
parameter.
There
are
no
really
final
parameters,
there
are
no
final
parameters,
since
the
system
continuously
evolved
and
improvement
can
always
be
made
and
new
features
and
understanding
will
emerge,
and
this
is
very
interesting
in
terms
of
the
economic
problem
that
we
are
dealing
with.
We
are
basically
an
evolutionary
economic
system,
but
we
should
be
making
the
mainnet
recommend
a
parameter
recommendation
available
pretty
soon.
C
This
is
one
question.
The
question
is
specific
to
the
gas
decalculation,
so
the
question
was,
I
believe,
is
about.
Do
we
consider
charging
different
base
fee
based
on
different
message
types?
Actually,
we
did.
We
actually
recommended
accounting
based
on
the
different
message
types,
because
some
of
the
messages
like
proof
commit
or
like?
Oh,
it's,
a
simple
transaction.
C
They
might
have
different
economic
meaning
to
the
network,
but
there
are
other
constraints
that
we
have
to
deal
with
here
so
like
this
will
be
considered
as
an
as
a
file
coin
improvement
proposal
and
just
want
to
make
a
note.
The
current
model
actually
charge
charges.
Different
fees
based
on
the
message
type
because
it's
based
on
the
the
burn
fee
is
based
on
the
gas
usage
and
gas
usage,
varies
across
different
messages,
different
message
types
and
just
want
to
make
sure
we're
on
the
same
page
on
the
goal
and
constraint.
C
So
there
should
be
a
network-wide
base,
feed
that
is
burned
for
the
usage
of
gas
as
messages
consume,
on-chain
resources
and
it
makes
sense
to
pay
to
the
network.
Otherwise,
we
create
a
bunch
of
other
incentive
issues
that
we
mentioned
below,
such
as
the
externality
associated
with
a
pretty
expensive
message
that,
let's
say
the
miner
can
potentially
put
in
for
free
or
some
denial
of
service
attack.
That
we
also
observed
during
testnet.
C
B
Yeah,
I
think
I
would
add
here
that
that
we've
seen
just
in
the
space
race
so
far
a
lot
of
a
lot
of
fees
burned
and
so
we're
working
on
the
parameters
there
too,
to
make
sure
that
less
is
less
sprint.
B
Yeah
yeah,
a
lot
of
that
is
just
software
like
having
having
just
not
immediately
sending
and
waiting
until
the
chain
is
less
congested
and
so
on.
Yeah.
C
Exactly
which
I
took
some,
I
kept
saying
it's
actually
a
good
learning
point
for
all
of
us
that
it's
it's
not
about
getting
your
chain
in
asap.
Getting
messaging
asap
is
important,
but
it's
not
zero.
C
I
think
this
is.
This
should
be
the
last
question
on
the
slide.
How
is
participation
incentivizing
foul
point,
and
is
there
a
tension
between
client
desire
for
low
price
storage
and
minus
interest
in
profit,
so
at
very
important
question,
but
going
to
tying
back
to
our
mental
model
as
falcon
as
a
storage
economy
where
we
all
come
together
for
the
use
like
structural
goods
and
services
will
benefit
when
the
economy
thrives
and
produces
value
or
utility?
C
And
we
consider
like
a
country
with
like
currency
and
this
export
industry.
We
have
raw
materials
coming
in
and
then
we
are
converting
that
into
goods
and
services
to
be
exported
and
whether
this
value
really
get
translated
into
the
exchange
rate
with
other
currency
pairs
with
other
currency.
C
A
really
base
is
really
based
on
supply
and
demand
on
the
market
itself,
but
something
that
we
do
for
know
in
terms
of
like
fundamental
forces
is
like
generating
tokens
very
quickly
when
there's
little
utility
on
the
network
might
lead
to
a
depreciation
and,
as
we
mentioned
earlier
in
one's
presentation
for
this
s-curve
minor
revenue
in
the
long
run
should
be
driven
by
the
storage
dual
payments,
and
these
deals
are
initially
heavily
subsidized
by
the
network.
With
the
goal
that
we
actually
set.
C
I
believe
that's
all
here
and
I
think,
like
let
me
just
exit
my
screen
share.
Can
you
pass.
B
Just
wanna,
you
know
one
one
really
big
important
disclaimer.
You
know
this
stuff
is
subject
to
change.
Everything
here
is
a
preview
like
we're
very
close
already.
We
think
a
lot
of
the
parameters
here
are
really
sad,
but
just
you
know
don't
make,
don't
make
a
lot
of
decisions.
Just
you
know
it's
a
preview.
This
is
working,
development
and
and
and
so
on,
so
you
know
really
strong
disclaimer
on
on
on
the
on
the
information
yep.
C
And,
as
I
said,
I
think
we're
the
frontier
of
networks,
social
system,
economic
system,
technical
system,
technical
network
system,
and
I
think
it
required
all
of
us
to
come
together
to
pull
our
knowledge
and
experience
and
resources
to
build
something,
valuable
and
and
meaningful
together.
C
B
We're
really
really
excited
about
where
this
is
headed.
We're
really
excited
about
the
entire
network,
we're
we're
looking
at
the
space
race
numbers
and
and
we're
it's
amazing
that
we're
hitting
this
much
storage.
Let's
look
at
it
right
now,
we're
at
89
petabytes,
just
in
in
the
amount
of
time
that
we've
we've
talked
another
petabyte
has
been
added
to
the
network.
This
is
amazing.
We're
about
to
hit
90.,
it's
a
really
huge
amount
of
storage.
B
B
We
we
just
need
to
to
to
you,
know,
get
through
get
through
the
space
race
and-
and
you
know
you
use
all
of
the
all
of
the
learnings
that
we're
getting
from
space
race
to
make
improvements
and
find
good
ways
of
of
dealing
with
this
kind
of
network
initialization
question
and
and
we're
almost
there.
So
it's
really
exciting
to
be
here
with
everybody.
The
entire
team
is
super
stoked
and
excited
about
about
where
we
are.
It's
been
for
many
of
us.
B
It's
been
many
many
years
of
work
to
to
reach
this
moment,
so
we've
been
working
super
hard
for
many
years
to
to
be
here.
We're
super
excited
about
the
economy
that
we're
building
together.
We're
super
excited
for
for
your
participation
in
in
providing
this.
This
you
know
really
new
kind
of
service,
a
new
kind
of
utility
and
you
kind
of
economic
economic
structure.
We
hope
that
the
you
know
the
conversation
today
has
been
has
been
really
useful.
B
B
We
tried
to
answer
a
lot
of
them
as
we
went,
but
but
we
we,
if
there
are
a
few
more
definitely
go
to
go
to
slack
and
and
ask
them
there
and
we'll
make
sure
to
to
come
back
to
some
of
this
later
on.
We
might
see,
I
don't
know
if
you
have
a
bit
more
time
to
take
a
few
more
questions,
but
maybe
if
we
take
maybe
five
questions
from
chad
and
then
then
we
can,
then
we
can
call
it
a
day.
B
So
so,
let's
see
question
I'm
going
to
go
in
in
in
maybe
I'll
go
in
in
order
from
from
the
top
down.
Since
people
have
been
asked,
questions
for
a
long
time
before.
B
C
Right,
I
think
we're
still
evaluating
some
of
this.
We
will
have
updates
on
the
the
main
recommendation,
pretty.
C
B
Yeah,
I
think,
and
then
you
know
kind
of
like
the
big
question
is
like
hey
how's,
the
mania
timing
and
so
on.
We
feel
really
good
about
it.
We
feel
the
the
whole
network
is
working
really
well.
We
see
this
amount
of
of
storage
online
stay
tuned
for
for
more
updates
on
this,
but
but
yeah
we're
looking
we're
looking
to
be
in
really
good
shape.
B
How
how
after
main
launch
should
the
discover
data
need
to
be
sealed
and
available?
Great,
that's
a
great
question.
So
I'll
refer
to
the
to
the
falcon
discover
folks,
I
think
I'll.
We
can
ask
that
on
slack
they
that
data
is
also
waiting
for
for
the
verifiers
to
be
to
be
online
to
be
able
to
hand
out
verified
deals,
so
we're
working
on
that
right
now
to
to
line
up
some
some
good
verifiers.
B
Initially,
we
think
it's
really
important
that
the
discover
data
is
not
the
only
verified
data,
so
we
want
to
make
sure
that
when
the
discover
data
comes
online,
it
also
comes
online
with
other
verified
data.
B
That's
not
the
discover
data
and
so
we're
just
working
right
now
to
line
up
a
few
more
folks,
adding
data
if
you
can
find
some
extremely
useful
sources
of
data
and
storage
in
various
different
regions
of
the
world,
especially
especially
in
in
asia-pacific
regions
or
or
in
in
europe,
would
love
to
to
hear
from
from
your
suggestions,
especially
scientific
data.
B
So
if
you
have
any
groups
that
are
trying
to
store
a
lot
of
really
useful,
valuable
scientific
data,
reach
out
to
me
on
slack
and
and
I'll
I'll
incorporate
those
into
into
some
some
of
the
potential
first
verifiers.
B
Great,
thank
you.
Thank
you
very
much.
We're
super
excited
about
this
thanks
for
joining
us
for,
for
so
long
have
a
great
great
night.
Thank
you
so
much
zx
for
for
your
awesome,
awesome
explanation.
I'm
sure
a
lot
of
people
got
your
I'm
super
thankful
for
you
for
your
answer.
Keep
going
with
the
space
race,
everyone
there's
a
lot
of
stuff
going
on.
We
added
we
we're
so
blown
away
and
so
excited
about
reaching
100
petabytes.
So
we
added
another
goal
after
after
100..
So
there's
another.
B
You
know
500
000
file,
coin
up
for
grabs
in
the
in
the
global
to
reach
the
200
mark,
so
you
know
keep
going.
We
want
to
hit
as
much
storage
as
we
can
and
remember
that
we're
trying
to
find
a
good
way
to
to
use
all
of
that
all
of
the
storage
that
you're
sealing
so
again.
We
can't
guarantee
it
yet,
but
we're
finding
good
ways
for
them.
Good
luck
and
great
good
luck
with
the
competition
and
see
you
all
soon,
thanks
so
much.