►
Description
## Agenda
We will be joined by Sam Macpherson(@hexonaut), an independent Smart Contract Developer at MakerDAO, and Bartek Kiepuszewski(@bartek) to talk about the recently announced [Dai Optimism Bridge with fast withdrawals!](https://forum.makerdao.com/t/announcing-the-optimism-dai-bridge-with-fast-withdrawals/6938)
## Information About Optimism and the Bridge
[Optimism](https://optimism.io/) is a popular Ethereum Layer 2 solution.
[Forum Post Announcement](https://forum.makerdao.com/t/announcing-the-optimism-dai-bridge-with-fast-withdrawals/6938)
[Tweet thread on how it works](https://twitter.com/sgmacpherson/status/1369315493245509632)
If you're curious to learn more about MakerDAO explore some of the resources below:
Awesome-MakerDAO Resource Repository: https://awesome.makerdao.com/
Technical Documentation: https://docs.makerdao.com/
Community Portal: https://community-development.makerdao.com/
Governance Forum: https://forum.makerdao.com/
Website: https://makerdao.com
Twitter: https://twitter.com/makerdao
Reddit: https://www.reddit.com/r/MakerDAO/
Chat: https://chat.makerdao.com/home
A
A
So
you
know
with
people's
time
coming
and
attending,
we
want
to
make
it
more
interesting.
We
want
to
make
it
more
diverse,
and
so
we've
been
focusing
a
lot
more
on
guests,
and
today
we
do
have
a
couple
of
really
really
great
guests.
We
have
sam
mcpherson
and
bartek
k.
I
will
not
pronounce
bartek's
last
name
because
it
is
polish
long
and
hard
to
pronounce,
and
so
they
are
going
to
be
talking
about
the
die
optimism
bridge
that
sam
announced
a
couple
weeks
ago.
A
So
I'm
going
to
invite
sam
on
screen
and
invite
bartek
on
screen
and
I'm
going
to
give
over
the
mic
to
them.
I
believe
they
prepared
a
really
nice
presentation
so
excited
for
that.
Hey,
sam
hello,
david,
hey!
What's
up,
what's
up
so
who's
who's,
gonna
be
sharing
the
screen.
Bartek.
Are
you
I'll
be
sharing
sweet,
sweet,
hey,
bartek,.
B
A
C
Sure
so
I'm
sam
mcpherson
I've
been
on
this
call
before
I'm
part
of
the
smart
contracts
team.
I
joined
into
maker
dow
from
the
community,
so
I'm
working
under
the
dao
as
opposed
under
the
foundation.
So
I'm
a
little
bit
unique
in
that
regard,
so
I've
been
working
under
the
dao
for
about
six
months
now
it's
been
great.
A
Nice
man,
nice.
Actually,
I
know
this
is
a
little
off
topic,
but
I
did
want
to
ask
you.
I
know
the
smart
contracts
team
is
doing
their
own
core
unit
and
I
know
previous
to
the
whole
core
unit
framework.
You
actually
got
voted
in
as
a
facilitator
for
your
own
smart
contracts,
domain
team.
Do
you
intend
to
do
your
own
like
corey
unit,
or
are
you
joining
that
smart
contract
square
unit?
I'm
kind
of
curious,
no.
A
B
I've
been
working
for
a
foundation
actually
for
the
last
three
years
and
been
focusing
on
all
sorts
of
aspects
of
the
general
liquidity
ecosystem.
If
you
like,
like
the
fringes
of
the
makeup
protocol,
doing
a
lot
of
data
analytics
analyzing
different
smart
contracts
as
well,
some
security
stuff,
and
lately
I
have
been
doing
a
lot
of
r
d
into
a
different
layer,
twos
working
close
with
optimus
team
with
ck
sync
metal,
apps
and
the
starkware,
and
just
you
know
watching
with
interest
this.
B
You
know
multi-train
environment,
that's
like
exploding
in
front
of
our
eyes
and
focusing
on
how
the
maker
protocol
will
have
to
expand
to
to
sort
of
chase.
You
know
the
liquidity,
that's
like
spreading
from
ethereum
to
to
other
chains
and
roll-ups
and
whatnot,
and
the
looking
forward
also
to
each
to
zero
and
sharding.
A
A
B
Yeah
so
from
my
perspective,
the
only
two
flavors
really
and
yeah
the
scaling
on
ethereum
that's
been
mostly
r
d.
For
a
long
time,
the
plasma
research
has
been
going
on
for
a
couple
of
years.
Now
I
think,
and
the
plasma
group
that
was
like
at
the
front
of
the
plasma
research.
You
know
they
rebranded
into
this
optimism,
team
right
same
people
essentially
and
and
the
optimistic
roll-ups.
B
So
this
is
the
flavor
that
they've
got
different
security
assumptions
than
zika
rollups
and-
and
I
think
like
in
my
opinion,
they
can
deliver
generalized
smart
contract
architecture
quite
fast
right,
and
this
is
like
fully
evm
compliant.
That's
gonna
make
it
very
easy
for
different
projects
to
to
sort
of
try
it
out.
Zika
roll-ups,
you
know
they've
got
much
harder
path
and
they
will
probably
come
a
little
bit
later.
A
B
A
B
A
C
A
C
C
So
optimism
in
case
people
aren't
aware,
is
a
fully
permissionless
optimistic,
roll-up
scaling
solution,
really,
there's
no
built-in,
like
bridge
mechanism.
Instead,
there's
a
like
a
generic
mechanism
of
message,
passing
between
l1
and
l2
and
from
l2
to
l1,
so
really
we'll
be
building
contracts
to
take
advantage
of
this
messaging.
C
The
the
bridge
is
built
on
top
of
that,
but
using
this
generic
mechanism,
so
we
can
implement
deposits
by
locking
erc20
die
into
a
deposit
contract
on
l1,
passing
the
message
from
l1
to
l2
and
then
melting
minting,
the
equivalent
token
on
l2
and
similarly
withdrawals
can
be
implemented
by
burning.
The
l2
die,
passing
the
message
from
l2
to
l1
and
then
releasing
previously
locked
tokens
in
the
deposit
contract
on
l1.
So
it
goes
full
cycle,
so
here's
kind
of
a
visual
representation
of
that.
C
The
tokens
will
be
transferred
to
the
bridge
and
held
in
escrow.
Then
we
call
this
messenger
message
passing
optimism
code
to
pass
a
message
to
optimism
and
it
does
this
it.
So
it
will
pack
another
message
on
top
of
the
what's
called
the
canonical
transaction
chain.
So
this
is
a
list
of
messages
such
as
you
know,
put
some
dye
into
l2
that
will
get
packed
in
ethereum
layer
one.
C
So
this
is
a
set
in
stone
once
this
is
packed
into
the
canonical
transaction
chain,
then
the
sequencer
will
relay
this
message
to
the
contract
on
l2,
which
will
then
forward
it
to
our
specific
optimism:
die
bridge
complete
the
deposit
and
voila.
You
are
minted,
100
optimism
die
and
then
send
it
to
the
l2
account
controlled
by
the
same
key.
C
So
for
the
reverse,
we
take
the
die.
The
optimism
die
on
l2.
We
send
it
to
the
l2
version
of
the
bridge,
initiate
withdrawal
which
will
burn
the
die.
The
optimism
die.
It
will
send
a
message
to
optimism's
message,
passing
contract,
which
will
then
initiate
a
withdrawal
in
the
canonical
transaction
chain
again.
So
keep
in
mind.
These
are
so
this
canonical
transaction
chain.
This
is
sort
of
a
list
of
messages
that
are
to
be
processed.
It
does
not
include
sort
of
what
the
end
state
result
of
what
this
will
be.
C
So
it's
up
to
the
the
sequencer
to
periodically
append
sort
of
state
updates
on
l1
and
these
state
updates,
because
this
is
an
optimistic
rollup.
So
these
state
updates
are
just
assumed
to
be
correct
unless
proven
not
to
be
correct.
So
this
is
where
we
get
the
one
week
withdrawal
sort
of
delay.
We
have
to
wait
a
week
to
ensure
that
there's
enough
time
that
any
single
actor
can
be
like.
Oh
wait,
no
wait
a
minute!
C
This
person
has
submitted
an
incorrect
state
update,
and
so
we
have
to
roll
back
the
state
batches
and
then
reapply
the
canonical
transaction
chain,
but
keep
in
mind
during
this
whole
process.
The
canonical
transaction
chain
is
set
in
stone.
It's
it's
just
a
list
of
these
are
transactions
that
are
going
through
so
yeah
and
again
it
passes
the
relay
message
after
one
week
and
then
completes
the
withdrawal
and
you
get
your
100
die
back
on
l1
yeah.
C
C
So,
but
we
still
need
that
week
to
wait
just
to
be
safe
before
confirming
things
on
l1.
So
what
we
can
do
here
is
we
can
institute
an
oracle
that,
instead
of
going
through
the
relayer
that
optimism
provides.
We
can
basically
just
say:
will
this
the
result
of
processing
this
canonical
transaction
chain?
C
C
C
And
if
the
oracle
is
wrong-
and
this
is
why
it's
important
to
have
this
fdi
token
and
not
just
pull
the
tokens
out
of
the
bridge
if,
in
the
event
of
a
failure,
we
need
the
maker
holders
to
ensure
the
losses
same
as
any
other
thing
right.
If
we
lose
the
collateral,
we
need
to
keep
the
accounting
correct,
and
so
we
have
to
take
a
loss,
and
so
this
is
an
important
case
to
take
into
account
as
well.
C
So
there
are
other
ideas
involved
with
this
still
kind
of
exploring
the
space,
so
we
could
have
a
sort
of
third-party
lps
instead
of
maker
protocol.
This
would
be
some
sort
of
like
lp
provider
that
would
enable
a
swapping
directly
between
l1
and
l2,
without
burning
the
tokens
which
would
be,
you
know,
more
efficient
for
small
players
and
stuff
like
that,
and
we
could
also
provide
fast
withdrawals
for
any
token
which
involves-
let's
say
you
know,
swap
v3s
out
on
l2.
C
C
Another
another
note
is
that
there
it's
kind
of
interesting
due
to
the
permissionless
nature
of
optimism.
We
we
really
need
social
consensus
around
the
one-two.
True
die
on
l2,
there's
all
these
other,
like
multi-chain
environments,
where
there
are
dye
showing
up
on
them,
but
like
without
sort
of
maker
endorsing
this.
Yes,
this
is
sort
of
we're
supporting
optimism,
and
this
is
the
true
die.
A
A
C
It's
it's
a
little
bit
arbitrary
optimism
has
come
up
with
this
number.
You
know
they
got
to
pick
a
number.
They
want
it
long
enough
that
it's
like
you
want
to
be
pretty
sure
that
everything
that's
going
to
l1
is
absolutely
correct,
so
they
just
pick
a
very,
very
safe
number
got
it,
but
it's
a
little
arbitrary
yeah.
B
But
I
can
also
add
to
this
that
there's
some
actually
interesting
crypto
economics
behind
it,
because
you
can
imagine
the
situation
where
somebody
like
an
honest
player.
They
actually
see
that
there's
a
fragile
state
being
committed
and
they
want
to
submit
the
throat
proof.
It's
just
that
you
know.
Suddenly
they
are
being
censored
by
miners
right.
B
So
picking
seven
days
would
literally
mean
that
the
minus
would
have
to
censor
my
transaction
for
the
whole
week,
which
optimism
team
kind
of
assumed
that
on
the
ethereum
network
is
virtually
impossible
right.
The
cost
of
such
an
attack
would
be
massive
to
censor
a
transaction
for
one
week
would
require
massive
hash
power
or
a
stake
in
in
the
future.
Proof
of
stake
and
and
seven
days
is
literally
saying
that
the
cost
of
such
an
attack
would
essentially
amount
to
censoring
a
transaction
for
the
full
seven
week
period.
B
A
C
B
But
eventually,
as
far
as
the
payment
on
optimism
on
l2
is
concerned,
because
optimism
is
actually
using
the
account
abstraction
for
all
the
accounts,
so
that
that
means
that
every
single
account
is
a
smart
contract
to
put
in
alignment
tenants
right.
That
means
that
they
can
do
whatever
they
want
with
how
the
gas
will
be
paid
in
the
future.
B
A
B
But
but
I
think
the
only
real
attack
is
an
oracle
attack
on
the
whole
construction
right.
So
if
I
were
to
attack
somehow,
which
essentially
means
that,
if
I
were
to
attempt
to
double
spend
my
tie,
I
would
try
to
send
my
tie
on
l2
to
somebody
right,
maybe
to
my
account
to
some
other
account
at
the
same
time
or
right
after
I
would
try
to
initiate
a
control
of
now
empty
account
right.
B
So
so
I
would
try
to
withdraw
money
that
I
do
not
have,
and
such
transaction
will
be
baked
into
this
canonical
chain
and
normally
under
normal
circumstances.
The
oracle
should
verify
it
and
say
that
you
know
that's
gonna
revert
right.
That's
gonna
fail
because
the
person
trying
to
withdraw
money-
they
just
don't
have
it.
B
However,
if
we
can
manage
somehow
to
bribe
particles
and
the
oracles
will
say
well,
look,
this
withdrawal
will
succeed.
Well,
then,
then
they
make
a
protocol
would
take
a
loss
right.
So
that's
one
attack
right
by
trying
to
bribe
oracles.
Another
type
of
attack
is
to
try
to
bribe
the
sequencer,
which
is
the
the
main
optimism
node.
However,
that
would
not
affect
make
at
all,
because
maker
does
not
rely
as
you've.
C
A
Yeah
the
way
that
I
was
thinking
about
it
because
this
that
a
part,
a
part
of
that
question
was
also
floating
around
in
my
head.
It
was
like
you
know
if,
if
they're,
if
we're
taking
like
risk,
not
just
in
like
you
know,
losing
to
fraud,
but
also
in
spending
money
on
like
running
the
oracles.
For
this
thing,
you
know
what
income
does
this
bring
and
my
my
thinking
was
okay
like
even
if
it
doesn't
bring
any
income
and
it's
like
purely
infrastructure.
A
It's
still
a
win
for
for
maker,
because
it
increases
dies,
usability
and
utility
because
it's
available
on
a
different
chain.
So,
although
it
might
not
directly
like
make
back,
you
know
the
money
that
will
be
spent
on
that
bridge
indirectly
it.
It
will
because
it's
ensuring
that
we
have
like
a
legitimacy
and
a
slice
of
the
layer,
two
pie
so
to
speak,
like
the
market
share
there,
but.
C
A
C
A
Yeah,
it's
interesting
that
there's
there's
options,
though
right
you
could
definitely
set
up
paying
for
immediate
withdrawals
if
you
needed
to
so
a
few
more
questions
in
the
chat.
So
guy
asks
so
we
have
die
layer,
one
die
o
die
optimism
die
fdi,
a
credit
for
dion,
l1
and
other
unofficial
dies
on
l2,
because
anyone
can
create
name.
A
token
is
that
right.
B
Okay,
actually,
when
we
think
about
fdi,
sometimes
we
call
it
the
forward
die,
but
sometimes
we
it's
better
to
think
about
it.
As
a
tokenized
withdrawal
request,
it's
almost
like
a
claim
to
die.
That's
going
to
be
released
in
one
week
right
and
the
guy
is
right.
We
will
have
die
on
l1.
We
will
have
optimism
die
on
l2,
which,
for
all
intents
and
purposes,
will
be
almost
the
same
token.
B
B
That's
actually
somebody
trying
to
withdraw
a
certain
amount
and
we
have
to
verify
that
you
know
that
can
be
okayed,
and
that
means
that
it's
got
value
because
after
a
week
it
will
actually
end
up
with
releasing
some
die
from
the
deposit
contract
and
it
may
be
worth
zero
right
because
you
know
when
we
verify
it
and
somebody's
trying
to
withdraw
tokens
that
they
don't
have.
You
know
that's
literally
worth
zero.
Now
the
other
unofficial
dies.
I
tend
to
think
about
it
as
a
wrap
die
right.
I
don't
think
about
it.
B
Then
these
tokens
just
die.
This
is
just
a
rap
token
that
is
created
by
putting
tai
into
some
other
bridge
and
on
the
other
side
of
the
bridge.
There'll
be
a
wrapped
token,
which
is
a
representation
of
the
dye
right.
It
can
be
redeemed
through
that
very
bridge
for
the
real
die
very
much
the
same
like
btc
and
wbtc
can
be
sort
of
interchanged
right,
but
nobody
thinks
about
wbtc
as
real
bitcoin
right.
It's
a
rap
bitcoin,
that's
on
under
some
management
of
of
some
custodian
in
the
middle
of
a
bridge
yeah.
C
Yeah,
but
the
interesting
thing,
that's
kind
of
new
that,
I
would
say,
is
different
from
say
btc
to
rap
btc
is
that
the
custodian
can
be
like
ethereum
layer,
one
which
is
like
non-custodial
right.
So
this
is
kind
of
a
new
space
where
we
can
get
these,
like
unofficial
dies
that
rap
dies
that
are
like
pretty
strong
guarantees
with
the
bridge,
because
it's
a
smart
contract,
so
there
may
be
more
competition
than
if,
like
some
random
unknown
person
was
providing
custodianship
on
die
to
some
other
place.
B
This
one
is
purely
permissionless
and
it's
it's
trustless,
and
because
anybody
can
deploy
any
type
of
smart
contract
and
optimism,
there
will
be
nothing
stopping
people
to
try
out
different
bridge
designs
and,
of
course,
any
single
contract
can
take
die
right.
I
mean
I'm
sure
that
you
know
people
have
put
data
into
all
sorts
of
smart
contracts.
You
know
with
different
expectations
and
different
results,
sometimes
perhaps
surprising
right
yeah,
sometimes
resulting
in
I
don't
know,
minting
some
pickles
and
whatnot.
A
Yeah,
so
just
before
I
ask
akasha's
question
that
came
earlier:
I'm
gonna
ask
guys,
because
it's
still
related
to
this
conversation,
so
the
unofficial
dies
would
still
be
permissionless.
The
real
difference
to
the
mkr.
Sorry,
the
real
difference
to
the
mkr
endorsed
version
is
brand
recognition
right
being
able
to
say
this
is
like
an
official
thing
from
the
maker
protocol
yeah
exactly.
A
Nice,
nice,
so
akash
asks,
is
maker
competing
with
anyone
else
to
provide
this
functionality
like
usdc
or
synthetics,
and
I
want
to
add
to
like
extend
that
question.
Is
it
like,
given
that
this
is
all
open
source
stuff?
Is
it
easy
for
people
to
just
like
copy
our
solution
and
create
their
own
bridges
to
optimism?
B
So
what
makes
maker
truly
unique
from
other
protocols
is
that
make
a
connection
mint
die
and
provide
unlimited
liquidity
by
taking
any
kind
of
a
collateral
which
will
be
accepted
by
the
the
governance
right.
So
in
this
particular
scheme
maker
will
use
that
tokenized
withdrawal
claim
as
a
collateral,
and
we
will
provide
this
extra
liquidity
regardless
of
the
amount.
B
Using
the
scheme
that
we've
discussed
there'll
be
no
limit
as
such
right,
we
might
put
initially.
The
debt
ceiling
to
sort
of
you
know
understand
better
the
risk.
The
oracle
attack
risk.
But
you
know,
if
make
a
governance
decides
that
the
oracle
attack
risk
is
negligent.
Then
I
think
this
limit
can
be
literally
anything
and
and
and
and
also
we
could
try
to
support
other
tokens
by
minting
die
and
using
this
diet
to
actually
purchase
liquidity
from
the
market
to
support
fast
withdrawals
for
other
tokens.
A
B
A
Nice
so
two
questions:
how
secure
is
the
code?
Where
is
the
code
right
now
for
this,
and
and
when
is
it
going
to
be
live
and-
and
I
understand
that
the
faster
the
draws
is
not
to
be
there
at
launch-
it's
something
that's
going
to
be
phased
in.
If
I
remember
reading
the
forum
post
correctly
so
yeah,
just
just
some
information
on
where
the
code
is
at
would
be
good.
C
C
A
A
B
B
B
You
know
they
will
be
definitely
more
interested
in
having
their
own
token,
with
much
more.
I
guess
sophisticated
functionality
if
you
like
right
than
just
a
simple
erc20
functions
and
obviously
maker,
will
be
looking
at
all
the
other
rollups
right.
We're
interested,
as
I
said
before,
in
having
die
available
everywhere,
and
this
is
really
about
expanding
usage
of
dyes.
B
So
that's
short
term
long
term
is
even
more
interesting
because
you
know
I
can
definitely
see
that
some
of
these
roll
ups
will
have
native
assets
still
quite
not
clear
how
these
native
assets
will
sort
of
be
created
on
on
these
roll-ups.
But
I
can
definitely
imagine
a
world
where
some
of
the
roll-ups
will
actually
have
tokens
or
whatever
asset
it
might
be
created
natively
on
l2
and
they
might
be
used
as
a
collateral
right.
B
So,
who
knows,
we
might
try
to
use
them
as
a
quadrant
for
dais,
somehow
either
directly
on
l2
or
we'll
be
going
the
other
way
right
and
and
sort
of
try
to
move
that
asset
into
the
l1
so
that
we
can
actually
use
it
inside
the
vault
so
make
a
kind
of
becomes
kind
of
a
hungry
protocol
right.
It
will
be
chasing
liquidity
everywhere,
whatever
the
form
or
shape
it
takes
because
you
know
the
more
liquidity
we
chase,
the
the
more
die
we
can
generate
right.
A
A
Nice-
and
I
think
that,
even
even
with
this
bridge
and
interacting
with
other
l2s
like
on
l1
at
the
base
right,
we're
still
totally
compatible
with
uidx
and
ave,
and
I
guess
it's
like
a
matter
of
in
the
future
like
what
they're
gonna
do
with
l2s,
because
I
imagine
if
they
also,
you
know,
recreate
their
protocols
or
you
know
it
really
depends
right.
There's
there's
just
so
much
unknown
from
this
from
the
place
that
we're
standing
now
right.
A
A
B
B
B
So
I
don't
know
I
mean,
but
but
again
I
mean
this
is
literally
something
that
we'll
be
looking
into
in
the
future,
because
this
space
is
evolving
right
and
it's
evolving
rapidly
and
we
need
to
take.
I
think
you
know
a
firm
position
as
a
makeup
protocol.
How
maker
will
support
this
multi-chain
multi-layer
ecosystem.
A
I
have
like
kind
of
a
a
general
question.
If,
if
I
have
an
ethics
both
on
l1
and
on
l2,
let's
say
I'm
gonna,
I'm
an
mkr
holder
is
it?
Is
there
a
way
to
like
create
and
airdrop
like
an
mk,
an
omkr
token
on
the
optimism
side
to
match
like
the
layer?
One
side
I
like,
I
guess
it's
a
question
of
like
are
eat,
addresses
like
one
to
one
between
optimism
and
l1.
C
Yeah,
so
the
your
externally
owned
account
that
is
one-to-one
with
optimism:
okay,
yeah!
So
if
you
you
can
it's
permissionless
sort
of
evm
setup,
so
you
can
do
anything
on
l2
that
you
can
do
on
l1,
so
you
can
make
a
contract
that
gives
out
what
you're
calling
omaker
or
whatever
and
give
it
to
people-
and
you
know
maybe
it
has
value.
Maybe
it
doesn't.
A
A
C
A
A
Yeah
I
mean
I
actually
like
the
governance
communications
core
unit
that
I'm
working
on
a
part
of
its
mandate
is
like
documentation.
Aid
to
gov
alpha
so
right
now,
like
the
governance
domain,
helps
a
lot
with
like
educational
resources
and
and
stuff
so
so
yeah.
Maybe
maybe
it
will
be
in
our
in
our
wheelhouse,
but
what
about
documentation
in
general?
Maybe
like
not
faqs,
but
what's
the
level
of
documentation
that
is
happening
as
you
guys
are
going
through
this
development
phase.
C
Yeah
so
there's
and
this
space
is
evolving
so
rapidly,
because
optimism
is
in
active
development,
so
we're
kind
of
in
close
tight
integration
with
them
they're
doing
updates.
We
pull
the
updates
in
and
do
changes,
and
so
there's
there's
documentation
in
the
repository
right
now,
but
we
need
to
kind
of
probably
solidify
things
a
little
bit
more
before
we
expand
that
out
to
other
places.
A
A
B
I
think
they're
already
supporting
it.
Actually
I
mean,
after
all,
optimism
has
launched
the
mainnet.
It's
it's
very
limited.
It's
not
permissionless,
only
synthetics
is
deployed
there,
but
I
think
you
can
play
with
this
today
if
you're,
the
owner
of
snx
tokens
nice
and
my
understanding
is
that
you
can
definitely
use
metamask
the
same
way
as
you
can
use
metamask
for
phantom
and
for
binance
for
avalanche
and
whatnot
right.
A
B
Zika
roll-ups
will
have
to
post
the
proof
and
the
smart
contract
will
have
to
verify
that
proof,
so
that
will
incur
a
quite
significant
gas
cost
optimistic
roll-ups.
They
just
literally
post
a
state
commitment,
which
is
this
tiny
little
piece
of
data
and
transactions
which
are
not
put
in
the
state
they
are
inside.
What
we
call
a
call
data,
it's
a
very,
very
cheap
operation.
B
In
my
opinion,
optimistic
roll-ups,
the
gas
cost
point
of
view.
They
will
be
actually
cheaper
at
least
initially
it
all
kind
of
depends
how
many
transactions
you
want
to
pack.
However,
there
will
be
a
significant
improvement,
I
think
probably
will
looking
100x
when
the
data
is
compressed
and
and
all
potential
optimizations
are
implemented.
But
again
I
I
don't
think
that's
going
to
be
the
case
initially
if
we
can
make
10x20x
initially,
you
know
that
it's
going
to
be
pretty
rad.
B
B
Being
able
to
submit
the
throat
proof
is
almost
like.
You
know
this
red
big
button,
which
you
can
press
when
evil
nation
will
attack
you
with
atomic
bomb
right,
but
you
know
that
you
know,
there's
just
I
mean
literally
years
can
pass
before
you
actually
have
a
need
to
press
this
button,
so
that
may
be
actually
the
case
as
well.
I
mean
you
have
to
remember
that
l1
ethereum
today
minus
when
they
publish
block,
they
also
publish
state
commitment-
and
this
is
you
know
very
much
optimistic.
B
B
I
mean
there
is
literally
nothing
stopping
mine
to
put
an
incorrect
state
route
and
try
to
mint
if
out
of
thin
air.
The
only
thing
that's
actually
stopping
this
miner
is
that
others
are
watching
right.
They
can
verify
themselves
and
if
they
see
the
liar,
they
will
immediately
punish
the
layer
by
rejecting
the
block
right.
So
all
the
hash
power
that's
actually
spent
mining.
This
block
will
be
wasted,
and
here
we
will
have
a
very,
very
similar
situation
there.
There's
this.
A
C
B
But
I
guess
you
can't
get
it
right.
The
first
time
can
you
write,
so
there
must
be
some
way
to
upgrade
the
bridge.
You
know
just
to
allow
for
some
iterations
until
we
hit
the
final
design
and
then
I
guess
you
know
when
the
final
design
is
sort
of
steady,
then
you'll
be
able
to
I
mean
the
governance
obviously
will
be
able
to
decide
what
to
do
with
the
keys
and
potentially
burn
them.
B
A
Nice,
nice,
I
know
so
I
know
like
when
multi-collateral
dye
was
still
in
development.
There
was
like
this
huge
emphasis
on
formal
verification
and
like
really
nailing
down
like
the
core
contracts,
and
I
know
like
in
this
environment
where
we
have
to
you
know,
move
fast
as
a
protocol
to
innovate
and
produce
like
code.
That
does
really
cool
things.
I
imagine
like
a
bit
of
the
formal
verification
goal
might
like
actually
yeah.
Let
me
rephrase
the
question
yeah.
A
C
Yeah
so
there's
kind
of
two
pieces,
there's
the
bridge
piece
and
then
there's
the
die.
Erc20
die
contract,
so
the
rc20
die
contract
is
basically
the
same
as
the
one
on
l1
we've
just
sort
of
modernized
a
little
bit
of
it
to
bring
it
up
to
modern
standards,
but
other
than
that.
It's
basically
the
same
thing,
so
that
one
is
immutable
and
because
the
contract
is
previously
formally
verified,
we
can
be
fairly
confident
that
one's
gonna
be
okay.
C
C
A
C
B
Yeah,
which
is
about
to
say
that
your
question
david,
is
actually
something
that
I
was
like
waiting
for
a
long
time
for
somebody
to
ask,
because
you
just
nailed,
you
know
the
the
the
I
mean
people
tend
to
sort
of
forget
that
the
form
of
verification
is
actually
done
on
evm
level.
We
don't
trust
compiler
right.
We
account
for
bugs
in
solidity
compiler.
B
So
when
maker
says
that
there
is
verification
done
that
actually
means
that
we've
got
the
byte
code,
that's
formally
verified
that's
how
it
works
right,
at
least
for
maker
and
the
other
tools
that
have
been
used
for
formal
verifications
and
with
your
scale
app
for
that
now.
The
optimism
to
my
knowledge
is
actually
the
only
rollup
out
of
every
other
role
that
I
know
which
is
pure
evm,
all
the
other
roll-ups
they
take
solidity
code.
B
They
say
that
you
know
you
can
like
take
your
certainty
contracts,
but
not
compile
solid
contracts
right.
They
take
your
solidity
code
and
they
actually
compile
it
to
their
own
internal
language.
So
starkware
has
their
own
language
right.
The
other
zika
sync
has
their
own
language,
even
arbitrarium.
Interestingly
enough
has
their
own
language,
so
they
do
not
execute
vm,
they
execute.
B
You
know
something
else,
and
I
think
I
don't
know
about
our
vitrum,
but
I
think
formally
verifying
whatever
comes
out
from
their
compiler,
which
is
by
the
way,
not
a
bytecode.
Even
it's
a
circuit.
It's
a
mathematical
circuit
right
for
zika
roll-ups.
That
is
going
to
be
an
interesting
proposition,
an
interesting
challenge
as
well
right,
so
I'm
very
curious
how
this
space
is
going
to
evolve
and
what
the
people
who
are
doing
formal
verification
for
a
living.
You
know
what
is
they
approach
to
to
all
new
zika
rollups,
for
example,.
A
B
A
A
But
I
don't
know
I
guess
this
makes
a
really
good
case
for
why
we're
doing
an
optimism
bridge
as
opposed
to
others.
First
but
but
yeah.
It
sounds
like
a
kind
of
a
secondary
thing
that
the
highly
talented
engineers
are
thinking
about.
So
I
dig
it
and
yeah.
It
sounds
like
it,
I'm,
echoing
what
ian's
saying
in
the
chat
it
does
sound
like
that.
That
feature
is
a
feature.
A
C
So
I
think
we
kind
of
have
to
ask
ourselves
a
few
questions.
First,
before
we
even
begin
to
scope
out
what
that
looks
like
and
so
like
it,
there
may
be
the
case
that
we
do
want
to
put
the
most
of
the
protocol
onto
l2,
but
I'm
not
entirely
convinced
of
that.
Yet,
because
different
protocols
have
different
sort
of
usage
usages
so
like
unit
swap,
for
example,
is
high.
High
traffic
people
are
swapping
all
the
time.
It's
a
perfect
candidate
to
go
on
l2
right
away,
but
maker.
C
You
know
people
don't
like
wind
and
unwind
their
leverage
like
like
many
many
times
per
second
kind
of
thing:
it's
not
that
sort
of
throughput,
so
it
I'm
not
sure
if
it's
even
necessary.
It
may
be
that
once
all
the
high
traffic
gas
usage
stuff
goes
to
l2,
we
get
sort
of
a
breather
with
gas
prices
and
then
it's
okay
to
use
maker
on
l1
for
a
while
again,
but
I'm
not
saying
that's
for
sure.
We
just
have
to
wait
and
see.
A
Yeah,
I
guess
like
the
the
idea,
is
we
don't
want
to
miss
out
on
like
a
potentially
large
customer
base
like
like
at
the
moment,
basically
only
whales,
it
only
makes
sense
for
whales
to
use
maker
and
because
of
gas
prices.
A
lot
of
the
smaller
users
with
lower
capital
amounts
got
priced
out,
and
if
that's
going
to
persist
to
be
the
case
for
another
one
or
two
or
however
many
years
probably
like,
I
would
hope
maximum
like
two
or
three
years.
A
A
B
So
let
me
provide
my
take
on
akash's
questions
because
they
are
super
interesting
as
well.
I
kind
of
think
that
the
current
approach
for
small
users
for
protocols
like
maker
is
going
through
intermediary,
like
let's
say
year
and
vault,
for
example,
that
allowed
me
to
actually
put
some
ease
whatever
this.
The
amount
is,
and
they
will
create
one
giant
vault
right,
but
it
doesn't
have
to
be
permissionless.
It
doesn't
have
to
be
here-
and
I
think,
there's
a
lot
of
scope
for
liquidity
providers
to
actually
provide
such
service
for
small
users.
B
They
will
just
operate
on
one
big
fault:
they
will
take
small
portions
of
collateral
and
they
will
pack
it
up
right
bundle.
It
together
and
just
you
know,
stick
it
in
into
into
a
vault,
because
I
mean
it
may
be
expensive
to
use
if
you're
in
for
small
amounts,
but
actually
for
bigger
amounts.
It's
actually
cheaper
than
centralized
exchange.
Interestingly
enough
right,
if
your
trade
is,
I
think,
above
maybe
ten
thousand
dollars
which,
for
a
while,
is
not
such
a
huge
amount.
B
It's
actually
cheaper
to
trade,
on
uni,
swap
than
on
the
centralized
exchange
and
for
the
oasis
tax.
I
think
it's
kind
of
right
now
I
mean
there
is
absolutely
no
reason
to
not
deploy
ac
stacks
on
every
single
rollup
right,
which
is
evm
compliant,
including
optimism
as
there
is
no
reason
for
say,
unisop
or
sushi
not
to
do
the
same
right.
There
is
text
after
all,
is
just
you
know,
permissionless
of
the
book
type
of
decentralized
exchange.
B
A
Yeah,
nice
great
answers
and
great
questions
from
everybody,
and
we
are
about
30
seconds
from
the
top
of
the
hour,
and
so
I'm
gonna
open
up
to
the
chat.
Any
final
questions
I'm
happy
to
stay
like
a
couple
two
three
minutes
over,
but
otherwise
this
was
an
extremely
interesting
conversation
and
and
yeah.
Thank
you,
sam
and
thank
you
bartek
for
coming
always
a
pleasure
to
speak
with
both
of
you
thanks
david
yeah
cheers
cool
yeah
and
I
don't
see
any
more
questions.
So
I'm
happy
to
end
it
here.