►
From YouTube: ECDC - DAY 2: Rent roundtable moderated by Phil Daian
Description
No description was provided for this meeting.
If this is YOUR meeting, an easy way to fix this is to add a description to your video, wherever mtngs.io found it (probably YouTube).
A
This
place
is
actually
a
monastery
in
Taipei.
You
can't
rent
it
out,
unfortunately,
but
it's
a
nice
segue
into
the
talk
hi,
everyone,
I'm
Phil
I'm
a
second
year,
PhD
student
and
researcher
at
isee-3
and
Cornell
University
and
I
work
on
smart
contract
security
and
distributed
systems
security,
including
distributed
systems,
protocols
and
I'm
here,
to
talk
to
you
guys
about
incentive
compatibility
in
rent.
A
A
So,
let's
talk
about
blockchain,
dirty
little
secrets
and
it's
not
I
owe
des
turnery
or
a
OSS
block
reducer
cartel
it's.
Basically
the
fact
that
every
blockchain
right
now
is
consists
of
users
doing
transactions
on
a
network
and
what
this
essentially
is
a
two-sided
commodity
market
between
users
and
people
who
are
running
a
network
so
like
miners,
full
nodes,
etc.
Sometimes
these
categories
overlap,
but
not
always
so
this
model
is
kind
of
flawed,
because
right
now
miners
decide
what
transactions
are
included
and
they're,
not
the
only
ones
bearing
the
cost.
A
For
example,
full
nodes
are
not
incentivized
and
the
computational
model
as
charging
these
fees
is
overly
simplistic.
It's
sort
of
approximates
costs
at
a
rough
cut,
but
it
doesn't
really
have
any
deep
insight
about
what
these
research
should
cost
or
any
way
of
responding
to
sort
of
the
market
changes.
There's
a
huge
reliance
on
comments.
A
So
a
lot
of
things
people
are
doing
in
block
games
right
now
are
not
incentivized
they're,
relying
on
this
Commons
model,
where
people
are
being
nice
and
altruistic
and
are
doing
it
like
mostly
for
free
or
because
they
have
some
other
reason
to
do
it.
For
the
moment,
resources
as
a
result
are
currently
mass,
the
under
price
or
subsidize.
A
However,
you
want
to
see
it
so
when
you
store
something
on
the
network
today,
it
cost
you
a
few
cents
at
most
and
the
reason
it's
so
cheap
is
essentially
because
you're
externalizing
this
cost
on
to
everyone,
who's
running
a
network
now
and
everyone
who
will
be
running
the
network
forever.
So
we
saw
Alexie's
graphs
earlier
about
the
sink
with
like
the
do
s
attack
and
like
that
was
sort
of
a
temporary
blip
in
the
network
history.
A
But
it's
something
that
client
developers
and
especially
people
who
want
to
run
these
archive
on
those
potentially
may
have
to
deal
with
forever.
So
we
need
to
enable
more
efficient
marketplaces.
That's
my
theory
and
there's
a
project
Chicago
webpage
here.
If
you
want
to
read
more
about
the
like
overarching
project,
but
I
won't
get
too
into
it,
because
I
want
to
have
like
I,
said
more
of
a
discussion.
A
So
today
we're
gonna
talk
about
storage,
rent
storage,
rent
is
a
pretty
simple
idea.
That's
been
around
in
etherium
for
a
while,
so
there's
a
bunch
of
background
under
the
agenda,
it
was
originally
proposed
in
the
IP
35,
which
was,
let's
see
exactly
April
third
twenty
stuff,
or
no,
even
before
that
November
25th
2015
so
like
early
early
days
of
it
Syria.
So
this
isn't
a
new
issue,
but
it
also
hasn't
happened
yet
and
that's
one
of
the
things
we're
going
to
talk
about
today,
all
right.
A
A
Let's
all
agree
to
pay
together
rather
than
having
anyone
decide
all
right,
cool,
so
I
guess,
let's,
let's
start
with
rough
sentiment
of
the
room
like
how
many
people
here
are
aware
of
blockchain
state
and
what
it
is
all
right.
How
many
people
know
the
difference
between
state
and
history?
Okay,
most
people,
that's
great.
How
many
people
here
use
the
etherium
a
net
like
on
a
regular
basis
and
pay
gas
on
the
Syria
main
name?
A
Okay,
okay,
good
number,
all
right
cool,
so
I
think
we
can
skip
some
of
the
some
of
this
background
issue,
but
maybe
we
can
start
because
this
is
a
client
dev
meeting.
Are
there
any
client,
devs
or
anyone
else?
Any
researchers
want
to
talk
about
how
large
state
can
potentially
like
negatively
impact
users.
Other
theory
all
right,
this
whole
section
I
have
something
to
say.
B
The
state,
how
much
how
big
the
state
is,
and
at
the
moment
it's
not
that
big
if
you
were
actually
store
it.
You
know
it
was
about
10
gigabytes,
which
I
think
it
could
still
be
further
compressed,
because
the
state
has
got
a
lot
of
repeating
elements
like
the
somebody's
addresses
as
a
token
owners.
That
would
be
further
reduced.
So
it's
still,
it's
still
possible
to
keep
the
state
in
memory
from
most
modern
machines.
B
So
I
think
they
might
be
there.
That
was
that
when
I
was
at
who
know
then
and
I
welcome
to
this
guy,
he
was
what
he
was
doing.
He
was
trying
to
retrace
for
some
reason.
I
don't
know
he
was
trying
to
retrace
all
the
transaction
that
ever
happened
in
the
theory,
and
he
said
by
his
calculation
is
going
to
take
him
like
six
months
to
do
that
on,
go
Assyria,
you
know,
and
he
obviously
means
a
full
state.
I
mean
the
exchanges
need
some
of
the
states
to
monitor
the
deposit.
B
A
Yeah
and
to
just
comment
on
that
quickly,
I,
for
example,
would
need
to
do
something
like
what
you
said
retrace
every
transaction
in
some
range
of
aetherium
and
here's
like
a
concrete
reason
for
it.
So,
like
one
of
the
problems,
we're
studying
right
now
is
like
how
much
rent
our
user
is
paying
in
the
system
on
like
decentralize
exchanges
to
arbitrageurs,
for
example,
and
to
study
these
marketplaces.
A
You
also
need
to
know,
and
people
like
try
to
make
fail
transactions
to
them,
as
well
as
making
successful
transactions
and
because
failed
transactions
don't
have
any
side
effects
on
the
parts
of
the
state
that
are
easily
accessible,
usually,
and
because
these
transactions
can
be
like
intentionally,
complicated
and
obfuscated
through
contracts.
There's
there's
actually
no
way
to
know
like
a
full
picture
of
this
marketplace
without
rerunning,
every
single
transaction
and
checking
like
does
it
call
this
contract.
C
You
know
I
mean
if
you
need
the
history,
you
know
I
mean,
maybe
you
can
keep
it,
but
the
issue
with
with
we're
paying
rent
on
history
is
that
well
I
mean
then
obviously
we're
not
gonna
keep
it.
If
you
don't
pay
rent
and
that's
kind
of
contrary
I
mean
so
somehow
what
I'm
saying
is,
but
we
can't
really
charge
people
rent
to
keep
something
in
history.
A
Knows
so,
I
think
today,
we're
like
scoping
out
to
mostly
state
ranks,
but
it's
worth
like
thinking
about
history
and
how
that's
not
incentivized
and
maybe
also
making
it
somehow
incentivize
in
the
future
or
just
making
sure
as
an
as
a
norm
for
the
community
that
it
won't
ever
be
like
unscored
will
to
to
anyone.
So.
B
But
for
his
use
case
he
kind
of
needs
a
history,
but
without
the
nest
he
doesn't
actually
need
to
stay,
but
he
had
needs
a
history.
If
there
was
another
type
of
storage
in
aetherium,
which
is
essentially,
you
can
write
once
and
never
update,
which
is
kind
of
the
history
thing
and
it
was
cheaper
than
let's
say
the
usable
storage,
then
that
would
be
very
beneficial
for
that
use
case.
So
there
might
be
a
youth
they
might
get
benefits
of
having
like
another
type
of
storage,
which
is
immutable.
A
A
A
C
A
D
E
A
C
C
Thank
you
Doppler
to
adapt
that
you
know
lets
you
store
a
bunch
of
things
in
the
state,
theater,
contract
or
storage,
and
then
it's
clear
when
the
gas
price
is
higher
for
a
discount
and
so
like.
There
was
actually
incentives
because
the
cause
of
gas
price
to
store
junk
in
the
ethereum
blockchain,
and
you
can
do
it
today
and
it's
very
rational,
disclaimer.
E
E
E
Economic
security
y-you
said
in
the
area
in
an
absolute
sense.
Storage
is
underpriced
relative
to
an
externality
minute
imposes,
but
that's
only
one
thing
right.
That
was
the
only
problem
that
was
for
just
some
form
of
the
gas
cost
event
store
life
after
Tecna
call
it
a
day.
The
second
problem
is
that
there
is
insufficient
incentive
to
your
storage.
E
Basically,
the
volatility
of
the
price
and
volatility
and
and
the
externality
so
look
like
gas
prices
are
10
times
more
expensive
when
lots
of
other
people
are
setting
transactions
at
the
same
time,
but
and
you
could
clean
if
the
key
in
make
a
case,
possibly
that
there's
economic
model
under
which
that's
happening
because
of
that
particular
points.
The
cost
of
the
network
of
accepting
even
more
transactions,
actually
there's
10
times
higher,
but
with
respect
to
state
storage.
That's
that's
not
true
at
all.
Right,
adduction
doesn't
really
matter
by
those
who
stay
or
go
or.
E
Megabyte
next
year,
and
then
some
of
the
finer
issues
basically
are
that
the
gas
costs
for
storage
should
first
of
all
the
gas.
That's
for
storage
should
not
depend
on
the
level
of
usage
of
the
extent
to
which
other
people
are
filling
up,
storage
right
now.
The
other
thing
is
indeed
that
the
gas
box
of
storage
should
not
depend
on
the
extent
to
which
people
are
doing
non
storage
or
related
things
right
now,
so
I
can't
mean
all
of
that
put
together,
there's
basically
the
some
large
incentive
to.
E
C
C
G
H
A
D
I
Yeah
a
site
from
the
long-run
equilibrium
and
finer
issues.
There
is
this
gaping
underbelly,
you
know
storage
bomb
attack,
that
was
2015
least
authority
paper
and
they
always
take
his.
You
know
one
mining
pool
to
get
mad
at
a
theory.
Oh
and
you
know,
there's
nothing
to
stop
them
from
blowing
over
state.
Like
the
you
know,
Shanghai
POS
tax.
E
In
the
case
of
bandwidth
and
computation,
miners
have
an
additional
disincentive
against
including
transactions,
which
is
the
incremental
uncle
rate,
whereas
with
storage.
That
is
something
not
existing
at
all
and
this
first
of
all,
further
under
prices
storage.
But
second,
it
creates
the
incentive
that
miners
have
the
incentive
to
discriminate
in
favor
of
storage
storage,
gas
that
heavy
transactions
rather
than
commutation
heavy
transaction.
Because
that
way
they
reduce
their
own
Co
rate
and
to
this
basically
with
the
optimal
strategies
for
minors
or
weird
stuff.
I
A
A
Then
I
think
this
problem
actually
gets
even
worse,
because
storage
is
probably
gonna,
get
cheaper,
so
I
guess
I
have
two
questions
and
then
I
want
to
open
it
up
to
like
a
more
general
comment
period.
The
first
one
is:
let's
talk
about
Bitcoin
for
a
second
which
of
those
pathologies
that
we've
been
mentioning.
H
J
E
E
G
E
B
There's
another
kind
of
extreme,
more
extreme
point
of
view
which
I
don't
remember
where
I
heard
it.
It
was
basically
come.
I
ain't
seen
that
maybe
maybe
we're
worrying
too
much
about
those
the
growth
problems
because,
like
you
know,
I
remember
I've
been
promised
that
there
would
be
some
opt
fiber
optic
cables
with
a
limited
bandwidth
everywhere
in
like
five
years,
and
then
everybody
will
have
like
terabytes
and
terabytes
of
storage,
which
is
then
and
then
everybody
will
have
something
else
and
then
maybe
what
we
do.
If
you
worry
about
this
grow
storage
growth.
G
E
E
Like
a
storage
getting
filled
with
like
I
tell
intentions,
people
be
complaining.
Those
storage
getting
filled
with,
like
people
shoving
their
tweets
into
store
into
storage
because
they
want
ever.
Then
they
want
30
tweets
to
be
accessible
by
a
lot
by
the
network
more
easily
than
through
whatever
boys
for
grabbing
history,
and
also
with
sharding
like
basically
and
remember,
about
shorting
is
quadratic.
So
if
computers
better
than
the
shortest
and
potentially.
D
B
Reason
why
I
think
the
reason
why
something
like
electron
takes
that
much
memory
is
not
technological.
Well,
it's
sort
of
technological
progress,
but
here
because
the
computers
have
become
and
faster
than
the
programmers,
don't
care
too
much
about
the
performance
because
they
think
it's
like
we're.
Gonna
just
ride
on
a
Morse
code
and
that's
why
what
operating
systems
loaded
up
and
things
like
this?
But
if
you
do
kind
of
spend
time
on
optimizations
and
writing
a
proper
code,
then
you
can
go
over
anywhere
long
way
and
it
does
it
doesn't
stop.
B
B
Make
it
better
so
I
remember
when
when
I
was
at
when
I
was
at
like
the
University,
and
we
were
going
to
this
programming
competition
I
remember
vividly
this
moment
where
they
usually
put
up
the
quadratic
complexity,
algorithms
and
I
will
say
yeah
implemented
in
the
run.
Time
limit
is
like
10
seconds
if
it's
over
10
seconds
and
it's
fail,
fail
the
test
and
you
normally
realize
it.
B
But
if
you
have
a
quadratic
complexity
of
rhythm,
it
will
definitely
fail
because
linear
complexity
or
and
again,
then
it
will
be
fine,
but
then
three
years
later,
and
we
really
tried
to
do
linear
complexity.
Others,
three
years
later,
that
didn't
matter
anymore,
quadratic
complexity
flew
through
the
through
the
things,
and
it
was
actually
difficult
to
find
a
challenge
where
the
quadratic
complexity
algorithm
would
be
sufficiently
slower
than
a
dead
little
complexity,
and
that
was
kind
of
depressing
because
the
people
just
drove
whatever
they
they
feel
like,
and
then
it
won't
pass
the
test.
B
E
B
C
So
you
want
to
know:
is
it
like
there's
this
kind
of
assumption
around
that,
like
after
scaling
or
with
bigger
blocks
that,
like
fees
will
be
lower,
the
rent
will
be
cheaper.
This
not
like
fundamentally
economically
a
thing
like
it.
Just
because
you
increase
the
supply
of
something
doesn't
actually
necessarily
mean
that
the
demand
will
increase
by
at
least
as
much
I
thought.
It's
so
like
a
nice
example.
Is
that
like,
as
we
make
more
and
more
efficient
engines?
E
C
E
D
A
Lad
all
right,
so
all
the
people
that
lat
doesn't
trust,
I
kind
of
want
to
open
this
up
to
the
whole
room
and
maybe
get
some
opinions,
some
perspectives
and
question
anything
you
guys
might
want
to
say.
I
have
some
prompts
up
here.
So
if
you
like
any
of
these,
you
can
kind
of
answer
it
or
ask
it
again
or
run
with
it.
If
not
feel
free
to
also
say
viewing
things
don't
make
me
pass
the
mic
around.
Please.
I
G
D
E
E
Sitting
over
stage
so
like
it
doesn't
make
the
situation
more
barky
and
even
in
this
model
right,
if
you
wants
to
you,
could
so
we
could
totally
have
a
system
where
the
state
storage
is
because,
like
constricted
to
be
very
small
like
like
32
megabytes.
Oh
sorry,
and
if
you
do
that
like,
if
you
push
on
the
extremes,
then
like
you
could
encourage
a
very,
very
large
portion
of
users
to
like
basically
any
kind
of
cultural
contract
to
just
have
32
bytes
of
state
and
rely
for
groups.
Instead
of
doing
what
they're
doing
no.
A
Yes,
so
I
think
one
interesting
question
I
heard
earlier,
just
really
quickly
was
like
why
not
just
use
swarm
or
another
one
of
these
second
layer
markets
that
have
figured
this
out
and
I
would
say
the
same
thing
like
in
order
to
use
something
like
that.
If
the
etherium
community
would
have
to
ask
itself
like,
is
this
incentive
model
something
we're
okay
with
in
the
long
term?
Otherwise
people
will
lose
data.
L
I
B
Two
bytes
for
storage
and
the
blockchain
and
everything
is
obtained,
and
so
what
it
would
mean
in
terms
of
the
climate
mutation
or
it's
a
storage.
If
you
look
at
remember
the
diagrams,
where
today
I
showed
you
that
there's
forty
three
gigabytes,
all
local
bodies
and
the
blood
puddings,
actually
include
transaction,
which
included
a
so.
B
What
we're
doing
is
that
you're
shifting
the
lot
of
storage
into
that
part
of
the
pipe,
because
you
will
still
score
some
throw
some
of
the
plans
will
still
square
over
blocks
and
then,
basically,
all
that
stories
are
those
with
logs,
where
it's
not
very
beautiful,
but
it
will
still
be
shuffled
around
the
network.
All
these,
it
looks
like
North
approves.
B
Another
thing
I
was
gonna,
say
yeah
yesterday,
where
I
came
up
with
another
teacher
in
that
meeting
spoke
about
how
to
incentivize
the
knowledge
actually
to
store
the
data
not
through
the
rent,
but
actually
by
using
interactive
groups
instead
of
monitoring,
which
they
are
here
that
if
you
want
to
really
mistake
for
me,
I
will
give
you
the
state,
the
part
of
state
and
prove
that
you
cannot
be.
The
proof
will
be
interactive,
which
means
that
you
will
not
be
able
to
prove
to
somebody
else.
B
E
G
E
Yes,
exactly
so
I
know,
this
is
kind
of
surprising,
but
I
have
thought
about
this
and
there's
quite
a
I
have
kind
of
quite
a
bit
of
reasoning
for
it.
It's
so.
Basically,
one
important
thing
to
realize
is
that
like
transaction
is
in
any
city
may
be
realistic.
One
run
equilibrium
are
and
will
continue
to
be,
a
lot
more
volatile
than
that.
E
You
might
think
you
have
an
historical
experience
right
and
we
kind
of
saw
the
first
clip
of
this
at
the
end
of
2017,
but
you
know
there
will
be
more
and
part
of
the
reason
why
I
think
people
got
complacent
about
transaction
fees
before
2017.
Is
that
like,
if
you
look
at
the
carafe
of
the
quench
Bitcoin
transaction
fees
and
either
transaction
fees
in
USD,
it
actually
does
manage
to
be
reasonably
stable?
E
The
political
mechanism,
basically,
is
if
the
transaction
fee
gets
too
high,
people
organized
into
plane
and
the
car
like
it
just
goes
in
and
drops
the
number
of
it,
and
this
actually
happened
several
times
a
big
point.
It
actually
happened
several
times
in
aetherium
and
it
actually
did
and
because
it
is
a
political
mechanism
idiot
like
it,
did
end
up
reading
the
this
kind
of
pressure
toward
stabilization
right
now
in
the
long
run.
E
Challenge
the
the
problem,
basically,
is
that
so,
first
of
all,
the
transaction
fees
are
going
to
get
very
volatile
right
and
in
general,
like
in
the
case
where
the
two
what's
option
increases
by
a
factor
of
ten
like.
First
of
all,
you
can
expect
price
that,
like
as
a
first
approximation,
got
to
increase
by
a
factor
of
ten
Angie,
it
expects
that
he
had
to
be
incur
for
sending
transactions
to
go
up
by
a
factor
of
ten
so
like
in
this
situation.
The
impact
I.
E
E
E
Super
linear
right
so
like
the
cost
of
increasing
the
of
adding
4
million
gas
to
eat
million
gases
like
one
variable,
but
then
the
cost
of
grilling
from
8
million
to
save
32
million
is
not
four
times
worse,
it's
more
than
four
times
worse
and
the
reason
why
it's
more
than
four
times
worse,
that
it
has
substantial
probability
of
just
breaking
the
network,
so
the
basically
so
because
of
that
it
makes
sense
to
have
an
equilibrium
where,
like
basically
you're,
please
go
higher
as
a
more
existing
usage.
There
is,
and
so
the.
D
E
That
the
vertical
supply
is
a
supply
curve
that
the
protocol
presents
actually
has
been
of
choice
or
obstacle
in
storage
space
land.
Actually,
if
people
start
temporaries
using
four
times
the
storage,
it's
not
that
bad
right,
like
if
I
stuck
if
the
size
of
storage
goes
up
by
a
factor
of
four
or
even
ten,
nothing
brace
things
you
get
inconvenient
and
it's
really
not
nice,
but
nothing
breaks
in
the
same
way
that
things
break
if
the
uncle
rate
goes
to
80
percent,
and
so
because
of
that.
E
C
E
E
Thing
is
that
there
are
greater
obvious
Thai
fielder's
greater,
like
it's
it's
more
of
a
bad
thing
for
rent
supp
prices
denominated.
It
used
to
be
volatile,
that
it
is
for
transaction
fees
to
be
volatile,
and
the
reason,
basically
is
that,
like
rent
prices
are
something
that
there
are
a
lot
of
people
where
they
be
scary,
but
they're
pretty
much
stuck
and
they
have
to
pay
it
regardless
and
because
they
already
started
some
application.
They
already
have
some
contract.
E
L
C
Right
well,
I
hear
you,
but
if
we
do
want
to
bound
it
state
space,
there's
like
superpower,
states
make
you
and
we
that
was
a
yes.
There
are
contracts,
gonna,
be
there
because,
like
they're,
competing
with
everyone
else
today
actually
commit
against
wanting
the
bound
state
space.
But
you
know
decentralization
doubling.
E
Wait
well.
The
point
is,
though,
because
the
social,
but
basically
like
you,
have
to
think
sort
of
not
on
the
first
derivative.
All
we
have
to
think
on
this
on
the
second
derivative
level,
like
I,
know
it's
kind
of
hard
to
picture
your
head,
because
all
right
primates
are
involved
for
second
derivatives,
but
like
the
idea
is
right
that
let's
say
some
given
state
space
model
gives
us
some
level
of
decentralization.
J
C
E
D
E
This
is
obviously
the
demand
curve.
Just
suddenly
becomes
super
huge
then
like.
Basically,
the
level
of
these
people
will
be
really
really
high
right,
and
so
because
of
that,
the
supply
of
these
will
start
going
down,
and
so
the
price
of
people's
to
any
ll
start
going
up
in
anticipation
of
future
supply
decreases,
and
so
the
price,
if
it's
going
going
up,
will
by
itself
help
to
discourage
more
people
throughout
legend.
E
E
C
E
Saying
that,
like
basically
especially
any
kind
of
Enron
scenario
where
the
price
of
beef
is
like
I
knows
more,
based
on
like
direct
usage
to
the
end
and
whatever
it
is
people
eat
for
now,
these,
like,
basically
it
the
some
of
that,
like
we
get
some
kind
of
security
that
we
get
some
level
of
assurance
back
because
of
this
mechanism.
We
are,
if
transaction,
if.
E
E
A
A
One
of
several
questions
number
one:
if
you're
a
dev
or
a
researcher
like
how
urgent
is
this
problem
and
how
quickly
do
we
need
to
move
and
if
you're,
not
a
Deborah
researcher
like
what
would
make
you
feel
good
about
paying
rent
some
potential
ideas
include,
if
being
burned,
to
increase
the
price
of
your
model.
You.
E
A
A
M
M
M
Would
be
interesting
to
see
if
this
wouldn't
be
just
left
on,
like
the
air
store
costs,
but
rather
it
would
involve
an
interpretive
redesign
highest
or
works
in
terms
of
cost
and
everything
else.
I.
Don't
think
it
would
be
good
to
have
like
another
extra
rule,
just
left
on
top
of
it.
In
the
second
comment,
I
would
be
wondering
what
kind
of
patterns
this
could
spring
in
languages
how
languages
would
work
with
trims.
F
So
as
a
question,
you're
sort
of
second
category
of
questions
about
the
controversy
I,
like
at
least
in
the
etherium
world,
gas,
doesn't
seem
to
be
particularly
controversial.
So
I'm
kind
of
wondering
why
Brent
is
so
controversial
and
is
it
rent
as
a
concept
or
is
it
more?
How
do
we
add
rent
onto
a
system
that
doesn't
have
it
today.
E
Like
the
answer,
basically
I
think
is
that
the
main
thing
that
causes
friends
to
be
controversial
is
that
it
introduces
the
possibility
that
if
you
have
some
accounts
and
you
don't
pay
rent
for
it,
then
the
account
gets
deleted
and,
like
you,
can
mitigate
that
and
like
there's,
only
five
different
ways
of
mitigating
it,
one
being
like
charging
a
fixed
rent
price.
So
you
can
calculate
how
long
you
have
to
live
more
recently.
Another
is
the
hibernation
mechanism.
I
wrote
about
anything
research
bellick
still,
it
is
the
very
fundamental
change
to
guarantees.
E
E
E
Do
you
have
to
delete
the
state
if
that,
if
that
user
ends
up
not
paying
up
or
the
other
alternative,
is
that
you
have
to
kind
of
move
objects
into
other
contracts
than
you
and
you
have
to
create
an
abstraction
that
say
this:
is
the
state
object
owned
by
some
user
to
could
disappear
at
any
time?
So,
basically,.
H
J
F
One
more
quick
question:
why
not
put
the
model
around
and
like
make
storage
store
very,
very
expensive
as
if
something
was
going
to
be
stored
forever
right
to
sort
of
charge,
all
the
cost
upfront?
And
then
it's
edifice
people
to
more
aggressively
free
storage
to
get
that
money
back
if
they
don't
need
to
store
something
forever.
C
A
Think
one
of
the
complaints
I
have
with
this
scheme
and
a
lot
of
other
schemes
I've
seen
that
or
like
based
around,
like
locking
up
phones
or
something
similar
like
anything
where
you
sort
of
pay
upfront
forever.
Is
that
somehow
it
feels
wrong
to
me
to
assume
that
we
can
predict
future
costs,
so
we
may
be
like
over/under
selling
that
space.
A
So
if
we're,
if
we're
sort
of
over
selling
it,
then
the
user
would
actually
save
money
if
they
were
paying
rent
and
if
we're
under
selling
it
we're
creating
like
potentially
an
unsustainable
situation
for
from
later
and
and
the
way
I
kind
of
see.
These
systems
is
that
they
should
kind
of
be
dynamic
and
like
respond
to
capacity,
and
to
do
that,
you
need
to
sort
of
change
these
these
these,
like
dials
a
little
bit
and
at
some
point
you
will
be
like
over/under
pricing
I.
Don't.
A
E
All
these
refund
models,
there's
kind
of
us
to
be
part
trilemma.
We
are
there's
to
be
possibilities
right.
So
let's
say
you
have
any
Easton
all
for
Excel,
so
came
to
nominated
C
for
filling
mr.
works
a
lot.
Then
your
question
is:
what
is
the
refund
for
clearing
option?
One?
The
refund
for
clearing
is,
it
is
exactly
one
widget
percent.
If
you
do
that,
then,
like
basically,
everyone
cuddles
Luthor
has
no
incentive
not
to
handle
it
in
storage
slots
instead
of
hogging
I
get
immunity,
and
so.
E
K
So
basically,
if
you
let's
say
store
rice
in
and
you
want
to
trade
it
in
the
next
two
months,
if
you
pay
rent
on
that,
and
so
some
key
words
you
have
for
validation,
if
actually
is
the
rent
is
positive,
but
you
can
also
have
negative
rent
if
people
just
want
to
not
store
it
and
want
to
sell
it
anything
like
if
it's
you
have
expiration
date,
for
example-
and
this
is
called
the
back
validation.
So
if
you
want
to
find
some
idea
of
names,
look
into
the
finance
world
and
investopedia
I.
G
C
C
C
M
C
Then
the
application
will
have
to
have.
You
know,
be
online
and
envy
on
swarm
and
how
it
has
actions,
but
I
mean
I,
mean
I
kind
of
think
of
it
as
being
an
independent,
independent
thing.
But
if
we
were
to
have
stateless
clients,
then
I
think
that
it
would
potentially
help
or
potentially
could
be
a
big
part
of
the
solution.
N
So
I'm
most
concerned
I
suppose
with
how
the
users
expect
these
systems
to
behave,
as
is
mentioned
earlier,
and
a
lot
of
the
rhetoric
has
been
around
particularly
about
immutability,
and
so
when
we're
thinking
about
how
we
describe
this
stuff
I
tend
to
agree
that,
but
that
rent
is
perhaps
the
wrong
word,
because
it
has
the
wrong
kind
of
connotations
and
I've
been
sitting
yet
seriously,
trying
to
think
of
like
difficult
senators.
It's
quite
a
difficult
question:
they'll
be
yeah,
I!
Think
that
that's
most
likely
right.
It's
just
that.
It
has
me
thinking.
N
Actually,
you
know
this
shouldn't
be
free
as
in
beer,
because
you're,
actually,
you
you
need
to
pay
for
information,
and
we
need
to
somehow
find
a
way
of
getting
that
across
to
users,
and
so
it
seems
to
me
like
there's
that
there's
some
opportunity
to
be
like
this
is
a
freedom
costs
right
because,
like
what?
What
ultimately
is
the
point
of
having
immutability?
N
On
a
theory,
I
mean
it's
it's
so
that
you
can
assert
your
own
version
of
history
and
agree
on
that
history
deterministically
to
the
Clovis
set
of
kills,
which
is
kind
of
like
the
reasons.
I
teach
the
technical
kind
of
freedom
which
interests
everybody
in
this
room.
That's
the
like,
like
like
a
feedom
lunatic
freedom,
I,
don't
know
something
deep
them
yeah,
there's
something
they
out
of
it.
O
So
there's
a
lot
of
like
signaling
externalities
to
consider
and
everyone
like
Namor
frame,
something
but
like
one
direction.
So
when
you
want
to
frame
like
charging
people,
something
like
one
direction
is
to
frame
it
as
like
a
membership
or
like
a
club
like
kind
of
a
thing
and
there's
like
you
know,
there's
a
again
elitist
considerations,
other
considerations,
but
that
would
be
a
cat
Boreum
five.
If
you
and
I
was
kind
of
reversed,
the
thought
did
go
to
the
sentiment
around
it.
One.
M
O
O
Naturally,
then,
like
the
inherent
kind
of
forces,
will
lead
to
a
reasonable
outcome,
and
so
I'm
almost
like
as
interested
in
like
what
happens
if
there
is
like
a
storage
bomb
like
attack
sooner
or
like
right
now,
the
markets
don't
really
work
in
a
way
where
that
seems
to
be
incentivized
at
performance
and
those
kinds
of
things
don't
seem
to
really
affect
prices
and
other
things.
But
you
know,
as
we've
seen
in
the
blockchain
world
a
lot
of
times,
it
only
takes
one
event
for
the
entire
sentiment
to
change
around
that.
O
So
you
know
it's
kind
of
like
arguably
like
a
Black
Swan
thinking
where
it's
like.
Well,
how
much
weight
can
you
put
on
this
likelihood?
But
if
there
is
this
kind
of
storage
attack,
let's
say
in
three
months:
what
do
you
do
about
it
I
wonder
how
much
we
should
be
considering
that
possibility
and
that
category
of
possibilities-
and
maybe
it's
just
like-
maybe
this
is
just
so
we
hope
this
is
able
to
progress
organically.
So
we
don't
have
to
deal
with
I
mean
that's
certainly
happened,
a
lot
of
other
circumstances,
another.
E
Thing
worth
pointing
out
as
the
hibernation
mechanism,
so
look
with
the
hibernation
mechanism
bright,
like
a
buckle
on
track,
doesn't
pay
rents
for
long
enough.
Then
it's
not
like
it's
done.
It
dies
forever
right,
so
contract
can
be
revived
any
TV
revived
through
this
mechanism,
where
basically,
you
provide
more
proofs
of
the
original
state
and
immortal
gruesome.
You
did
not
revive
to
the
contract
earlier
and
that
basically,
what's
bring
the
contract
back,
and
so
it's
not
so
much
as
like.
You
have
to
pay
for
being
alive.
E
E
Like
first-class
storage
so
likes,
your
robot
like
addresses
was
four
zeros
in
front
of
them
like
you
have
to
be,
you
have
to
pay
more
for
them,
but
then,
at
the
same,
at
the
same
time,
what
clients
are
expected
to
already
have
those
branches
in
the
end,
you
have
to
be
much
gas
for
actually
calling
me
using
those
accounts.
So.
E
The
amounts
that
you
pay
for
to
access
a
storage
salat
is
proportional
to
the
number
of
luck.
Innumerable
chip.
You
know
it's
that
you're
close
to
the
cost
of
notary
notaries
would
have
to
pay.
Then
you
basically
get
this
kind
of
emergence
like
first-class
storage,
because
whatever
that
I
kind
of
storage
Islands,
is
that
our
most
popular?
That
said
those
are
the
ones
that
are
going
to
be
more
expensive
and
the
ones
where,
if
you're
located
right
there,
they
don't
have
the
fewest
more
branches
that
you
have
to
open
up
and.
K
C
M
G
L
So
to
me
the
idea
of
the
rent
sounds
sounds
natural
and
I
just
wanted
to
have
that
to
to
this
hibernation
mechanism,
I
mean,
wouldn't
it
be
also
possible
to
if
a
contract
runs
out
of
gas
tomorrow
friend
damn
instead
of
like
directly
hibernating
it,
it
could
have
some
kind
of
fabrication
period
so
that
I
don't
know
for
three
months
or
half
a
year
or
to
find
it
blocks.
It
keeps
going,
but
you
would
be
notified
earlier
on
that.
Okay,
if
no
one.